23

European Forest Institute Research Report 23
The approach taken in the Fire Paradox project was based on the paradox
that fire can be “a bad master but a good servant”, thus requiring the
consideration of the negative impacts of current wildfire regimes
(understanding fire initiation and propagation) and the beneficial impacts
of managed fires in vegetation management and as a planned mitigation
practice (prescribed burning together with some traditional fire uses)
and for combating wildfires (suppression fire). These were the four
integration pillars of the project.
This Research Report reflects the structure of the project, corresponding
to its integration pillars – initiation, propagation, prescribed burning and
suppression fires – and including a closing chapter which synthesizes
and combines the main project outcomes. The book provides science
based knowledge that can assist policy makers to develop the necessary
‘common strategies’ to elaborate and implement integrated fire
management policies. It makes extensive use of the science and technology
findings from the Fire Paradox project, focusing on policies and best
management practices, as well as providing guidelines for the future.
The Fire Paradox project (2006–2010) was funded by the European
Commission Research and Development 6
th
Framework Programme.
The project included 30 partners from eleven European countries and
six partners from Africa, South America and Asia, with close support
from an International Advisory Committee formed by nine specialists
from the USA, Canada and Australia. Fire problems and solutions are
found all over the world, and we see the knowledge exchange and benefits
of Fire Paradox will extend far beyond Europe.
Joaquim Sande Silva
Francisco Rego
Paulo Fernandes
Eric Rigolot
(editors)
The European Forest Institute – EFI – is an international organisation established
by European States. EFI strengthens and mobilises European forest research
and expertise to address policy-relevant needs.
www.efi.int
ISSN: 1238-8785
ISBN: 970-952-5453-48-5
Towards Integrated Fire Management –
Outcomes of the European Project
Fire Paradox
S
i
l
v
a

e
t

a
l
.



T
o
w
a
r
d
s

I
n
t
e
g
r
a
t
e
d

F
i
r
e

M
a
n
a
g
e
m
e
n
t

–

O
u
t
c
o
m
e
s

o
f

t
h
e

E
u
r
o
p
e
a
n

P
r
o
j
e
c
t

F
i
r
e

P
a
r
a
d
o
x
Towards Integrated Fire Management –
Outcomes of the European Project
Fire Paradox
Joaquim Sande Silva, Francisco Rego,
Paulo Fernandes and Eric Rigolot (editors)
European Forest Institute Research Report 23, 2010
Scientifc Advisory Board of the European Forest Institute 2009–2010
Prof. Dr. Bas Arts, the Netherlands
Prof. Kristina Blennow, Sweden (as of 1 January 2010)
Prof. Americo Carvalho Mendes, Portugal
Dr. Emil Cienciala, Czech Republic
Prof. Dr. Carlos Gracia, Spain (as of 1 January 2010)
Prof. Dr. Hubert Hasenauer, Austria, Chairman of the SAB
Dr. Antoine Kremer, France (until 31 December 2009)
Prof. Michael Köhl, Germany
Prof. Marco Marchetti, Italy
Prof. Leena Paavilainen, Finland
Prof. Jean-Luc Peyron, France
Dr. Victor Teplyakov, Russian Federation (until 31 December 2009)
Editorial Offce
Prof. Dr Hubert Hasenauer, Editor-in-Chief
Ms Minna Korhonen, Managing Editor
Contact: publications@ef.int
European Forest Institute
Torikatu 34, 80100 Joensuu, Finland
www.ef.int
Cover photo: Pedro Palheiro
Layout: Kopijyvä Oy
Printed at WS Bookwell Oy, Porvoo, Finland
Disclaimer: This report was prepared for the project Fire Paradox under EU Sixth Framework
Programme (contract no. FP6-018505). The content of this publication is the sole responsibility
of the authors and does not necessarily refect the views of the European Union or the European
Forest Institute.
© European Forest Institute 2010
ISSN: 1238-8785
ISBN: 970-952-5453-48-5
To José Moreira da Silva, Pau Costa, and all those
who dedicated their lives to fre management.
Table of Contents
Foreword ...............................................................................................................vii
Acknowledgements ................................................................................................ ix
1. Introducing the Fire Paradox .............................................................................. 3
2. Wildfre Initiation: Understanding and Managing Ignitions

2.1 Fire Starts and Human Activities ................................................................. 9
2.2 Flammability: Infuence of Fuel on Fire Initiation .................................... 23
2.3 An Overview of Policies and Practices Related to Fire Ignitions at the
European Union Level .............................................................................. 35
3. Wildfre Propagation: Knowledge and Search for Solutions
3.1 Humans, Climate and Land Cover as Controls on European Fire
Regimes ..................................................................................................... 49
3.2 Fire Modelling and Simulation Tools ........................................................ 61
3.3 Wildland Urban Interfaces, Fire Behaviour and Vulnerability:
Characterization, Mapping and Assessment ............................................. 71
3.4 The Importance of Economics in Fire Management Programmes
Analysis ..................................................................................................... 93
3.5 Improving Fire Management Success through Fire Behaviour
Specialists ............................................................................................... 105
3.6 Wildfre Scenarios: Learning from Experience ....................................... 121
4. Prescribed Burning: Preventing Wildfres and More
4.1 Overview of Prescribed Burning Policies and Practices in Europe and
Other Countries........................................................................................ 137
4.2 Scientifc Knowledge and Operational Tools to Support Prescribed
Burning: Recent Developments ............................................................... 151
4.3 Learning and Training on the Use of Prescribed Burning Techniques ... 161
5. Suppression Fire: a Tool to Control Wildfres

5.1 Overview of Suppression Fire Policies and Practices in Europe ............ 177
5.2 Suppression Fire Use in Learning Organizations ................................... 189
5.3 Improving Suppression Fire Capacity ..................................................... 203
6. Solving the Fire Paradox – Regulating the Wildfre Problem by the Wise
Use of Fire ...................................................................................................... 219
Important Concepts and Terms ........................................................................... 229
Foreword
One of the most important challenges for European forests, especially in Southern
Europe, is fre. Fire has been affecting on average half a million hectares of forests,
shrublands and crops every year. The European Forest Institute is committed to
support initiatives which help generation and dissemination of scientifc knowledge
on forest fires. This knowledge is fundamental to support adequate political
decisions and to implement the most appropriate technical solutions in order to
mitigate the problem. This was the justifcation for the establishment in 2005 of
the EFI Project Centre PHOENIX concentrating on fire ecology and post-fire
management.
Another cornerstone addressing the importance of the fre issues in the EFI’s
activities was the EFI Discussion Paper 15 “Living with wildfres: what science can
tell us”. This publication by Dr. Yves Birot and his co-authors from the Regional
Offce EFIMED, provided the perspective that societies should learn to live with
wildfres by controlling the circumstances causing forest fres. This general concept
of Fire Management was also the leading principle of the EC-funded project “Fire
Paradox”, 2006–2010, in which EFI was part of the consortium.
In terms of knowledge dissemination, key researches in the Fire Paradox project
have been compiled into a report, with the aim of gathering the main results
achieved by the project in one publication.
The result of this effort is the present EFI Research Report. Besides the necessary
acknowledgements to all experts involved, I would also like to thank the European
Commission for supporting this publication through Fire Paradox. I am convinced
that it will be a useful contribution to the implementation of fire management
policies and measures in Europe.
Risto Päivinen
Director, European Forest Institute
Acknowledgements
The editors would like to thank all those who have contributed to the present
publication. In particular we have to thank three persons who gave precious
assistance to the publication of this Research Report: Minna Korhonen, Tim Green
and Luís Queirós. The editors appreciate the recommendations sent by Prof. Dr.
Hubert Hasenauer, Editor-in-Chief of EFI's Research Report series. We would also
like to thank all of the experts who participated in the review process:
Martin Alexander
Margarita Arianoutsou
Mark Beighley
Yves Birot
David Christenson
Miguel Cruz
Neels De Ronde
Mark Finney
Armando Gonzalez-Caban
Jim Gould
Chad Hoffman
Justin Leonard
Vittorio Leone
Rodman Linn
Peter Moore
José Manuel Moreno
Penelope Morgan
Jeffrey Prestemon
Douglas Rideout
Kevin Ryan
Jim Smalley
Scott Stephens
Ramon Vallejo
Gavriil Xanthopoulos
The editors are also grateful to all authors who accepted the challenge to participate
in this Research Report. Finally we would like to thank the whole Fire Paradox
Consortium, since this book is the direct result of the work developed during the
four years duration of the project.
The Editors
Joensuu, January 2010
1. Introducing the Fire Paradox
1. Introducing the Fire Paradox
Joaquim S. Silva
1
, Francisco C. Rego
1
, Paulo Fernandes
2
and Eric Rigolot
3

(the editors)
1
Centre of Applied Ecology “Prof. Baeta Neves”, Technical University of Lisbon,
Portugal
2
University of Trás-os-Montes e Alto Douro, Vila Real, Portugal
3
National Institute for Agricultural Research, Avignon, France
The Fire Paradox project (2006–2010) was funded by the European Commission
Research and Development program. The main objective of the Fire Paradox
project was to produce and use science to inform decisions, practices and policies
under integrated fre management in Europe. The project included 30 partners from
eleven European countries and six partners from Africa, South America and Asia,
with close support from ten specialists from the USA, Canada and Australia which
formed the International Advisory Committee. Fire problems and solutions are
found all over the world, and we see the knowledge exchange and benefts of Fire
Paradox will extend far beyond Europe.
The approach taken by Fire Paradox includes the development of the science
to inform the use of fre as a main component of the fre management solutions
proposed, and the development of strategies for their implementation at the regional,
national, and European levels.
The Wildland Fire Management Terminology produced by the Global Fire
Monitoring Centre for the FAO
1
defnes Fire Management as including “all activities
required for the protection of burnable forest and other vegetation values from
fre and the use of fre to meet land management goals and objectives”, involving
“the strategic integration of such factors as knowledge of fre regimes, probable
fre effects, values-at-risk, level of forest protection required, cost of fre-related
activities, and prescribed fre technology”.
Integrated Fire Management involves further integrating the different components
of fire management (fire prevention, detection, suppression, and use) with
consideration of fre ecology relationships, being defned by FAO/GFMC as “fre
management systems”. These include one or both of the following concepts of
integration:
1. Integration of natural or human-caused wildfres within prescription and/
or planned application of fre in forestry and other land-use systems in
accordance with the objectives of prescribed burning;
2. Integration of the activities and the use of the capabilities of the rural
populations (communities, individual land users) to meet the overall
objectives of land management, vegetation (forest) protection, and smoke
management (community-based fre management).
1 1999 Wildland Fire Management Terminology produced by the Global Fire Monitoring Centre for the FAO
4 Towards Integrated Fire Management
In other documents, such as those produced by the Global Fire Initiative, integrated
fre management is viewed as “an approach to addressing the problems and issues
posed by both detrimental and benefcial fres”, by “evaluating and balancing the
relative risks posed by fre with the benefcial or necessary ecological or economical
roles that it may play”, and by facilitating the implementation of “cost-effective
approaches of preventing detrimental fres and maintaining desirable fre regimes”,
recognizing that “managing benefcial aspects of fres may involve various forms of
fre use”.
The concept of integrated fre management was never intended for Europe, and
the philosophy of Fire Paradox offers an innovate approach for its implementation.
In fact, Fire Paradox proposes the full integration of fre use in the prevention and
suppression strategies by promoting the possibilities of the benefcial use of fre
by prescribed burning, and the traditional fre use; and by using fre to suppress
wildfres (see Figure 1).
When compared to other less populated parts of the world, the European situation
offers a very special context to the use of the Fire Paradox philosophy, as in many
regions of Europe fre has been traditionally used and continues to be used in rural
practices whereas, at the same time, the vast majority of damaging wildfres have
human causes. This apparent paradox is more evident in Europe than in many other
regions of the world; because of this, there may also be more potential solutions to
solve the paradox.
The approach taken in Fire Paradox was based on the paradox that fre can be
“a bad master but a good servant”, thus requiring the consideration of the negative
impacts of current wildfre regimes (understanding initiation and propagation) and
the benefcial impacts of managed fres in vegetation management and as a planned
mitigation practice (prescribed burning together with some traditional fre uses) and
for combating wildfres (suppression fre). These were the four integration pillars of
the project.
This Research Report refects the structure of the project, corresponding to its
integration pillars – initiation, propagation, prescribed burning and suppression fres
– and including a closing chapter which synthesizes and combines the main project
outcomes. The book provides science based knowledge that can assist policy makers
to develop the necessary ‘common strategies’ to elaborate and implement integrated
fre management policies. It makes extensive use of the science and technology
fndings from the Fire Paradox project, focusing on policies and best management
practices, as well as providing guidelines for the future.
The chapters were written by leading authors recruited from key researchers from
Fire Paradox Modules and Work Packages. Within each chapter there are a number
of text boxes, many of which describe specifc project activities, and outputs in
order to illustrate key messages and recommendations.
In order to complement the information provided in the different chapters, the
present publication includes a CD containing the deliverables and the internal reports
from Fire Paradox Modules and Work Packages. These documents are directly or
indirectly referred to in the chapters and were included in order to provide the outputs
of the project on which the conclusions and recommendations were essentially based.
The book has four different topic sections, containing 15 chapters:
Section 1. Wildfire initiation: understanding and managing ignitions. In this
section authors present the current knowledge and the project findings on fire
Introducing the Fire Paradox 5
ignitions related to human activities, fuel fammability and fre occurrence, and an
assessment of policies and practices related to fre ignitions.
Section 2. Wildfre Propagation: knowledge and search for solutions. This section
addresses issues related with different factors infuencing European fre regimes, fre
modelling and simulation, wildland urban interfaces, fre economics, the knowledge
on fre behaviour for fre fghting operations and the use of wildfre scenarios for fre
suppression purposes.
Section 3. Prescribed burning: preventing wildfires and more. This section
reviews recent developments to evaluate prescribed burning policies and practices
in Europe and other regions, the scientific knowledge and operational tools to
support prescribed burning, and discusses the learning and the training on the use of
prescribed burning.
Section 4. Suppression fre: a tool to control wildfres. This section presents a
review on the policies and practices of suppression fre in Europe and other regions,
Figure 1. Integrated Fire Management, from a fre use perspective.
6 Towards Integrated Fire Management
the application and practice of suppression fre as an operational tool in Europe and
discusses the requirements of training for improving suppression fre capacity.
As a concluding text, there is a standalone chapter presenting some of the project
developments which may contribute to solve the Fire Paradox in the context of
integrated fre management.
Overall, the present Research Report provides a comprehensive synopsis of
science, technology, products and services resulting from the Fire Paradox project.
In addition, this book presents science-based knowledge and recommendations for
changing the paradigm of fre management in Europe and elsewhere.
2. Wildfre Initiation:
Understanding and Managing Ignitions
2.1 Fire Starts and Human Activities
Filipe X. Catry
1
, Francisco C. Rego
1
, Joaquim S. Silva
1
, Francisco Moreira
1
,
Andrea Camia
2
, Carlo Ricotta
3
and Marco Conedera
4

1
Centre of Applied Ecology “Prof. Baeta Neves”, Technical University of Lisbon,
Portugal
2
Institute for Environment and Sustainability, Joint Research Centre of the
European Commission, Italy
3
Department of Plant Biology, University of Rome “La Sapienza”, Italy
4
Ecosystem Boundaries Research Unit, Swiss Federal Research Institute WSL,
Switzerland
2.1.1 Introduction
The knowledge about where, when and why do fres start is essential to assure
appropriate fre policy and management. The ability to understand and predict the
patterns of fre ignitions will help managers and decision makers to improve the
effectiveness of fre prevention, detection and fre fghting resources allocation.
This chapter starts by giving an overview of the patterns of fire occurrence
in Europe. Then we present a more detailed analysis of key factors driving fre
ignitions based on some case studies, and we conclude by making some remarks
concerning implications for management.
2.1.2 Overview of fre ignitions occurrence in Europe
Unless otherwise specified the information presented in this introductory
section results from the analysis of data extracted from the European Forest Fire
Information System (EFFIS) wildfre database concerning the period 1998–2007
(21 countries – not all the countries have data for the whole period). The EFFIS
database only includes wildfres that affected, at least partially, forest and other
wooded land areas.
Anthropogenic and natural causes
There is evidence that the earliest use of fre by hominids occurred more than one
million years ago (Pausas and Keeley 2009), although solid evidence date from
less than 300 000 to 400 000 years (Weiner et al. 1998). Natural fre regimes have
been increasingly altered by humans for many thousands of years at least, so that
in most regions of the world, human activities have become much more important
than natural sources of ignition (Goldammer and Crutzen 1993). This is especially
true for Europe and particularly for the Mediterranean Region with its long history
of human intervention in the ecosystems (Thirgood 1981).
10 Towards Integrated Fire Management
Currently, 95% of the fires in Europe are directly or indirectly caused by
human behaviour and activities. Among the fres with known causes, 51% were
intentionally caused, 44% started by negligence or accident, and only 5% had
natural causes (mainly lightning). However, there are enormous differences between
countries, with intentional causes representing 2% to 79% of all fres, negligent and
accidental causes representing 10% to 98%, and natural causes representing 0% to
32%. Figure 1 shows the main general causes of wildfres for the 13 countries that
had at least 500 fres that were successfully investigated.
Independently of the general designations (intentional, negligent, accident), which
can vary among countries, there are some common aspects which characterize
the human behaviour and activities which result in fre starts. Some of the major
anthropogenic causes of wildfres in Europe are associated with land management,
such as the burning of agricultural and forestry residues, land burning for pasture
renovation or the use of machinery. However, there are many other factors that are
also known to cause fres, including arson, accidents with electric power lines and
railways or the fre use associated with forest recreation (e.g. Bassi and Kettunen
2008; Vélez 2009).
Concerning the question of how do fres start, the investigation of ignition causes
is crucial if we want to understand the important role of humans in fre regimes.
As we saw there are big differences among countries with regard to what are the
main fre causes. However there is a considerable difference among the different
Figure 1. Main general causes of wildfres in Europe. Data extracted from the fre database
of the European Forest Fire Information System (EFFIS) (Period 1998–2007; only countries
with a minimum of 500 fres successfully investigated are included; countries are ordered by
decreasing proportion of deliberately caused fres).
0 °
10 °
20 °
30 °
40 °
50 °
60 °
70 °
80 °
90 °
100 °
I
t
a
l
y
F
i
n
l
a
n
d
S
p
a
i
n
P
o
l
a
n
d
P
o
r
t
u
g
a
l
T
u
r
k
e
y
F
r
a
n
c
e
S
w
e
d
e
n
L
i
t
h
u
a
n
i
a
B
u
l
g
a
r
i
a
C
z
e
c
h
L
a
t
v
i
a
C
r
o
a
t
i
a
Natural Negligence/Accident Deliberate
Fire Starts and Human Activities 11
European countries with respect to the percentage of fres for which the cause of the
fre is determined. These percentages range from nearly 100% in Latvia to only 5%
in Portugal. Moreover the criteria for investigating and classifying the cause of a
fre are not uniform among European countries, precluding the possibility of reliable
comparisons. Although the distinction between natural and human-caused fres may
be more or less straightforward, it is much more diffcult to classify the different
human activities responsible for a fre ignition. Even at the very coarse level of
classifcation which only separates negligent/accidental from intentional (arson)
causes, the different countries have not used the same criteria, making international
comparisons diffcult. For example, in Portugal uncontrolled agricultural burnings
are classifed as negligence, whereas in Spain this cause for fre starts is classifed as
intentional (APAS 2004).
Spatial and temporal patterns of fre ignitions
The higher densities of wildfres in Europe are mainly concentrated in the southern
regions with the Mediterranean countries accounting for most of the fres reported.
However in some central and northern countries fres are also frequent (Figure 2).
There are several factors that can potentially explain the spatial variation of ignition
density, such as climate, population density, main economic activities and land use.
Several studies worldwide have identifed numerous factors infuencing the spatial
patterns of fre ignitions (e.g. Mercer and Prestemon 2005; Loboda and Csiszar
Figure 2. Fire density (number of fres/year/10 km
2
) in Europe: average annual distribution in
the period 1998–2007; based on province-level data (NUTS 3).
12 Towards Integrated Fire Management
2007). However, the effects of different factors on fre ignition occurrence can
vary a lot among ecosystems and across spatial scales (e.g. Yang et al. 2007), and
until recent years scarce information was available for Europe. Some additional
information is presented in this chapter (see sections 2.1.3 and 2.1.4).
The number of wildfres in Europe has been increasing in recent decades (UN
2002; EC 2008). UN (2002) reported an increase from about 40 000 fires in
the 1970s to 95 000 in the 1990s (decade averages for a group of 31 European
countries). Several socio-economic factors, including the increasing recreational use
of forests and land accessibility contribute to this trend. However, the improvements
on fre detection methods and in the communication/reporting systems are likely
to have also contributed to the detection and reporting of more fres, and we do not
know to what extent each factor has infuenced the observed trend.
In the last decade, despite the large annual variability, the number of fres has
apparently stabilized or is even beginning to decrease. For example the EFFIS
database shows an annual average of 89 000 fres in the last decade (1998–2007)
which is lower than the average reported for the 1990s in a UN report (UN 2002);
however the EFFIS database refers to a group of 21 European countries, while the
UN report refers to 31 European countries, thus limiting direct comparisons. On
the other hand, looking at the trends of the fve largest EU Mediterranean countries
(those having a longer time-series fre database), and even if they present very
different situations, we can also see that the overall tendency in the current decade
is towards a stabilization or to a decrease in the number of fres (Figure 3), but some
more time is necessary to clearly confrm this trend.
Figure 3. Evolution of the number of fres in fve southern European countries in the period
1980–2007, using moving averages (n = 5 years) (based on country-level data; source: EC
2008).
0
5000
10000
15000
20000
25000
30000
1980 1985 1990 1995 2000 2005
N
u
m
b
e
r

o
I

I
i
r
e
s
Years
Portugal
Spain
Italy
France
Greece
Fire Starts and Human Activities 13
Fire occurrence in the European region is mostly concentrated in summer and
spring (see Figure 4). The majority (51%) of all wildfres occurs during the third
trimester of the year, corresponding to the summer months (July to September).
On average, 27% of the wildfres occur during the spring months (April to June),
14% occur during winter (January to March), and only 7% in autumn (October to
December). However the monthly distribution of wildfres is not the same for all
the countries. In the Mediterranean countries the majority of fres occur during the
summer, while in most of the remaining countries the majority of fres occur in spring.
On average the hourly distribution of wildfires peaks between 14.00h and
17.00h, with 30% of all fres starting in this period (Figure 4). The majority of
fres start between 13.00h and 19.00h (53% of all ignitions), and the period of the
day when fewest fre starts is between 03.00h and 07.00h (only 3%). There are
few differences among European countries with regard to the distribution of fre
ignitions during the day.
Figure 4. Average monthly (top) and hourly (bottom) distribution of wildfres in Europe.
Data extracted from the fire database of the European Forest Fire Information System
(EFFIS) (Period 1998–2007; 21 countries).
0
4000
8000
12000
16000
20000
J F M A M J J A S O N D
N
u
m
b
e
r

o
I

F
i
r
e
s
Months
Mediterranean countries
Other countries
0
2 000
4 000
6 000
8 000
10 000
0 2 4 6 8 10 12 14 16 18 20 22
N
u
m
b
e
r

o
I

F
i
r
e
s
Hours
14 Towards Integrated Fire Management
Fire ignitions and resulting burned area
In the European region, the vast majority of wildfres are small, and only a few have
major consequences. On average 70% of all European wildfres burn less than 1 ha
and 95% burn less than 10 ha. In total, less than 1% of all fres are larger than 100
ha, and less than 0.2% are considered to be large fres (more than 500 ha). Although
in general these patterns are quite similar in all countries, the larger fres are almost
exclusively concentrated in the Mediterranean countries.
The Mediterranean countries have the highest number of fres and the largest
area burned in Europe. On average about 500 000 ha are burned every year in these
countries.
2.1.3 Analysis of key factors driving fre ignitions
The Portuguese case study
Portugal is the European country with the highest density of wildfres, both in
terms of ignitions and of burned area, and thus it was selected as a case study for
a more detailed analysis. Recently a research team studied about 130 000 fres that
occurred in Portugal during a fve-year period (Catry et al. 2007, 2008, 2009). The
work was focused on characterizing and identifying the main factors driving the
spatial occurrence of wildfre ignitions. Several socio-economic and environmental
variables hypothesized to be related to the spatial distribution of fre ignitions were
selected and analyzed. Several logistical models were developed and an ignition
risk map was produced for the entire Portuguese mainland, demonstrating that it
is possible to predict the likelihood of ignition occurrence with good accuracy and
using a small number of easily obtainable variables.
Modelling fre ignitions
Human presence and activity are the key drivers of fre ignitions in Portugal. The
selected explanatory variables were all highly signifcantly related with the spatial
distribution of ignitions.
Population density was found to be the most important variable in the multivariate
models. This variable was positively correlated with ignition occurrence, meaning
a higher probability of ignition in the more populated areas. In several other studies
worldwide population density was also found to be positively related to wildfre
ignitions (e.g. Cardille et al. 2001). In Portugal, 70% of all fre ignitions occurred in
municipalities with more than 100 persons/km
2
although they only represent 21% of
the territory.
Land cover was also hypothesized to be a determinant factor causing ignitions
because different kinds of human activities (land uses) lead to different levels of
risk, like those implicating an active of fre (e.g. traditional burning to eliminate
agricultural residues), and because different land covers have different fuel
characteristics (e.g. moisture, fammability), which can determine fre ignition and
initial spread. In our studies land cover showed a strong infuence on the probability
of fre ignition, similar to other author’s fndings (e.g. Cardille et al. 2001). The vast
Fire Starts and Human Activities 15
majority of fres started in agricultural and mixed urban-rural areas (85%), and only
15% started in forested or uncultivated areas although they cover 50% of the country.
Results indicate that agriculture is a very important factor infuencing fre starts. This
is in accordance with previous investigations of fre causes, which concluded that a
large proportion of wildfres is due to agricultural activities (e.g. MMA 2007). Also
contributing to this high ignition incidence are the higher population densities usually
present in these areas (in comparison to forested and uncultivated areas), and that the
herbaceous vegetation in many Mediterranean agricultural areas is easier to ignite and
the fres propagate more easily than in other fuel types, especially during the summer
when fuel moisture is very low. The high ignition incidence in urban-rural areas may
also be explained by the co-occurrence of agricultural activities and a high human
presence. Forests, shrublands and sparsely vegetated areas also showed a positive
infuence on the multivariate models, but their infuence was considerably lower.
Probability of fre ignition occurrence decreased with increasing distance to main
roads; this result has also been found in other studies (e.g. Vega-Garcia et al. 1996).
According to our results, 98% of all ignitions occurred at less than 2 km from the
nearest road and 85% were within a distance of 500 m.
Elevation positively infuenced ignition distribution. This effect may be because
some human activities are more likely to occur at higher elevations, such as the
renovation of pastures for livestock using traditional burnings, which are known to
cause frequent wildfres in the Iberian Peninsula (see Box 1). Badia-Perpinyà and
Pallares-Barbera (2006) also found a higher ignition frequency at higher elevations
in a rural area of north-eastern Spain. Additionally, lightning-caused ignitions have
been described to be more likely to occur at higher elevations (e.g. Vazquez and
Moreno 1998).
Fire ignitions and resulting burned area
The spatial patterns of ignition are not the same for all fire sizes (Catry et al.
2008). This conclusion is based on both the comparisons between the frequency
of ignitions resulting in fres of different sizes, and the modelling work performed
taking into account the fnal burned area.
Although most fres start in areas of higher population density, there are important
differences in the origin of ignitions that resulted in smaller versus larger burned
areas, with most larger fres starting in areas with comparatively lower population
density (see Figure 5). The strong positive infuence of population density observed
in the global multivariate predictive models rapidly decreases for increasing fre
sizes, becoming negative when predicting the occurrence of fres larger than 500 ha.
The infuence of land cover type is different for ignitions resulting in small or
large fres. Considering all fre sizes, the class of mixed urban-rural areas had nine
times more ignitions than would be expected if they occurred randomly in the
territory. Forests and shrublands registered three times less fre ignitions than if they
were randomly distributed. However, results also show that the probability of large
fre occurrence is signifcantly higher in woodlands and lower in urban-rural areas,
when compared to small fres.
The frequency of ignitions also depends on distance to roads, and ignitions are
more likely to occur closer to the main roads independent of the resulting fre size.
However we verifed that the frequency of larger fres starting at longer distances
16 Towards Integrated Fire Management
Figure 5. Comparison between fre ignitions that resulted in burned areas smaller or larger
than 500 ha (signifcant differences * p<0.05, ** p<0.001) in relation to population density,
land cover, distance to roads and elevation. Expected fre frequencies (corresponding to a
random occurrence) are shown as reference (adapted from Catry et al. 2008).
0
10
20
30
40
50
60
70
0-10 10-20 20-40 40-100 ~ 100
F
r
e
q
u
e
n
c
y

(
°
)
Population density (persons/km
2
)
**
**
** *
**
0
10
20
30
40
50
60
70
Agriculture Forest Shrubland Urban-rural Others
F
r
e
q
u
e
n
c
y

(
°
)
Land cover
* **
**
0
10
20
30
40
50
60
70
80
0-500 500-1000 1000-1500 1500-2000 ~ 2000
F
r
e
q
u
e
n
c
y

(
°
)
Distance to roads (m)
**
**
*
0
10
20
30
40
50
60
0-250 250-500 500-750 750-1000 ~ 1000
F
r
e
q
u
e
n
c
y

(
°
)
Elevation (m)
**
**
**
Expected (random) Ignitions ·500ha Ignitions ~÷500ha
Fire Starts and Human Activities 17
is signifcantly higher than for smaller fres. Concerning elevation, larger fres are
signifcantly more likely to start at higher elevations than smaller fres (Figure 5).
Results of this study clearly showed that fre ignitions are strongly related with
human presence and activity, but also that the spatial patterns of ignitions are
different for larger or smaller wildfres. The models produced had reasonable to
good predictive ability, with ROC (Receiver Operating Characteristics) analysis
indicating 74% to 87% agreement between predicted probabilities and observed
outcomes. Models built for different years performed similarly, which indicates little
spatial variability in a fve-year period.
2.1.4 European comparisons – Relationships between ignitions and
land cover
Due to increasing anthropogenic fre activity, understanding the role of land cover in
shaping coarse-scale wildfre regimes has become a major concern. This is because
human presence and impact (e.g. population density, agricultural practices, grazing
pressure) together with the amount of fuel and its spatial distribution are basic
factors in explaining fre ignition and propagation (Lloret et al. 2002). Land cover
and fuel characteristics are characteristics that are often closely related (Turner and
Romme 1994), such that the probability of fre ignition will be affected to a large
extent by the nature of the different land cover types.
In a different study a research team quantifed coarse-scale patterns of fre ignition
in three selected European study areas in order to determine to what extent ignition
occurrence is dependent on land cover in different European regions (see Bajocco and
Ricotta 2007). Authors identifed land cover types where fre ignition probability is
higher or lower than expected by a random null model assuming that, if several land
cover types were equally fre-prone, fres would occur randomly in the landscape.
Methods
The analyzed data sets are composed of (i) 13 377 fre records in Sardinia (Italy)
during the period 2000–2004, (ii) 3023 fre records in the Coimbra district (Portugal)
during the period 2001–2005, and (iii) 1331 fre records in the Cantons of Ticino,
Graubuenden and Uri (Switzerland) during the period 1982–2005. Land cover maps
with a common legend were used for all selected study sites derived from CORINE
Land Cover data, where the original classes were aggregated into 12 coarser macro-
classes, the composition of which slightly varies across the study areas (see Table
1). The urban class includes mixed urban-rural areas, as stated in the Portuguese
study case.
To determine whether the number of fires in the examined cover classes is
signifcantly different from random, Monte Carlo simulations were used. The null
hypothesis is that fires occur randomly across the landscape so no differences
between the relative abundance of fres in each land cover class and the relative
extent of each class within the analyzed region are expected. At each study site, the
actual number of fres in each land cover class was compared with the results of 999
random simulations.
18 Towards Integrated Fire Management
Results
The results of the analysis on fre ignitions in selected European study areas are
shown in Table 1. According to the Monte-Carlo simulations, fre incidence in terms
of number of fres by land cover class is not random (p < 0.01).
Overall, though the results obtained slightly vary across the study sites, there is
a clear positive association between the number of fres and urban and agricultural
land cover types. More specifcally, in Sardinia, the number of fres is higher than
expected by a random null model in urban areas and in all agricultural classes.
By contrast, the number of fres is lower than expected in forests, grasslands and
shrublands. A similar result is obtained for Portugal, where, apart from ‘irrigated
arable land’ and ‘vineyards’, all land cover types with high human pressure are
characterized by very high fre ignition probabilities. Forests, natural grasslands and
shrublands exhibit the opposite behavior, all being characterized by low ignition
probabilities.
In Switzerland, the situation is slightly more complex, due to the presence of
two different fre seasons (spring and summer). Most artifcial and agricultural land
cover classes are characterized by high ignition probabilities. This is also the case
for two forest classes, ‘broad-leaved forests’ and ‘mixed forests’ that are strongly
affected by anthropogenic induced fres because of their proximity to settlements.
On the other hand, in grasslands, shrublands and coniferous forests, the number of
fres is lower than expected by chance alone.
From a statistical viewpoint, this analysis is easy to conduct at any spatial scale and
with any meaningful land-use classifcation scheme providing valuable information
for the development of strategies for fre risk assessment and fre prevention.
Table 1. Results of the analysis on fre ignitions occurrence by land cover class in different
European regions (Italy, Portugal and Switzerland). The symbols + and – indicate, respective-
ly, whether fre ignition occurrence probability is higher or lower than expected by a random
null model on a given land cover class at p < 0.01. () = not signifcant at p = 0.05; NP = the
land cover class is not present in the study area.
Land Cover Types Italy Portugal Switzerland
(Sardinia) (Coimbra) (TI- GR-UR)
Urban surfaces + + +
Non-irrigated arable land + + (-)
Irrigated arable land (+) - NP
Vineyards + - +
Fruit trees and olive groves + + +
Heterogeneous agricultural areas + + +
Broad-leaved forests - - +
Coniferous forests - - -
Mixed forests - - +
Pastures - NP (+)
Natural grasslands - - -
Transitional woodland-shrub
and/or sclerophyllous vegetation - - -

Fire Starts and Human Activities 19
2.1.5 Conclusions – Implications for management
The ignition risk (the chance of a fre starting as determined by the presence and
activity of any causative agent, also defned as ‘fre risk’; FAO 1986; NWCG 2006)
is an essential element in assessing fre danger (e.g. Finney 2005; Vasilakos et al.
2007). Nowadays, many fire management decisions are still exclusively based
on the factors infuencing the fre spread and suppression diffculty. However, as
resources are limited, it is important to defne priorities among areas. Under similar
fuel, topographic or weather conditions, the areas with higher ignition risk should be
given priority for surveillance (e.g. Vasconcelos et al. 2001; Chuvieco et al. 2003).
The ability to understand and predict the patterns of fre ignitions is essential for
managers. It will help improve the effectiveness of fre prevention, detection and
allocation of fre fghting resources. Knowledge about where, when and why do fres
start improve that ability.
Fire incidence varies greatly among land cover types. Some types of land cover
burn far more often than would be the case if fres occurred randomly. Further,
fire ignition probability is usually connected to social, economic and cultural
drivers. The disproportionately high fre occurrence in the urban and agricultural
classes is largely due to the high human presence that represents the major source
of fre ignition in most European regions. Population density, distance to roads
and elevation were also found to be important predictors of the occurrence of fre
ignitions. The present results may strongly contribute to solving the Fire Paradox by
providing to the managers information on priority areas with high fre hazard and
risk. In these areas, optimizing prescribed burning applications and the ecological
benefits of the wise use of fire may be very effective in achieving the wildfire
reduction goals.
The temporal patterns of fire starts seem to be mostly related with the
meteorological conditions, which affects the fre ignition and initial spread (e.g.
affecting fuel moisture and fammability). The knowledge of these patterns can also
have important consequences for fre management, namely in terms of vigilance and
fre fghting preparedness.
The knowledge on fre causes is also of crucial importance and should have a
high priority in terms of investing resources for minimizing the wildfre problem.
Nevertheless, there are important differences among European countries concerning
the process of fre cause investigation. These differences result in different rates of
investigated fres and especially in different classifcation systems, which preclude
reliable, quantitative comparisons between countries. Given these discrepancies,
the European Commission through its Joint Research Centre has recently launched
a call for proposals concerning the “Determination of forest fire causes and
harmonisation of methods for reporting them”. However it is necessary that results
coming from this initiative are translated into policies aimed at harmonizing the
different methods and criteria for fre investigation and fre cause classifcation.
Although local actions targeted at well identifed groups are an essential strategy
for preventing fre ignitions, Europe should have common initiatives and support
mechanisms on this topic. The existence of common criteria will help identify
which countries and regions need the most attention and where resources should
be preferably allocated to prevent fre ignitions. On the other hand, we also saw
20 Towards Integrated Fire Management
Box 1. Wildfres and pastoralism – the Portuguese case study
The relationships between wildfres and pastoralism were studied in Portugal using a
national database of about 135 000 fres that occurred during a six-year period (2002–
2007). A sample of 7337 fres with a known cause was used to investigate the particular
characteristics that differentiate the fres that were caused by shepherds (renovation of
pastures for livestock) from all the other causes.
On average, pastoral activity was responsible for 20% of all wildfres and for 11%
of the area burned. Shepherd-caused fres burned mainly shrublands (78%) and forests
(18%). In contrast the other fres burned more forests (56%) than shrublands (37%).
Logistic regression methods were used to analyze the relative importance of 26
potential explanatory variables and to develop fire ignition predictive models. The
factors having the greatest influence on the probability of pastoral fire occurrence
were the presence of forest, active population, recurrence of fres in the past, altitude
and slope. Presence of forests and active population had a negative infuence and the
remaining variables had a positive infuence. The tests performed to evaluate the models
showed good accuracy on predicting fre occurrence (82–85%, ROC).
Concerning topography, results showed that fres associated with pastoral activity
occurred at higher elevation and in more irregular terrain than the other fres. These fres
also started more often in areas with lower human population density/activity, more
distant to the major roads and where fres burned more frequently in the past. They were
also more frequent in areas with more shrublands and less forests when compared to the
other fres.
Fires due to pastoral activities also showed a different temporal pattern, occurring
mostly from mid-summer to mid-autumn (Figure 6). However large annual variations
in the fre season were observed, possibly being related with the annual meteorological
conditions and food availability for cattle.
Figure 6. Comparison between fres that were started for pastures renovation (6-year
averages): main land cover classes affected (left); and annual distribution of fres (right).
Reproduced from Catry and Rego (2008); (EU Forest Focus Programme: UTAD, AFN).
0
5
10
15
20
25
30
J F M A M J J A S O N D
F
i
r
e
s

(
°
)
Months
Pastoralism
Others
0 °
20 °
40 °
60 °
80 °
100 °
Pastoralism Others
B
u
r
n
e
d

a
r
e
a

(
°
)
Agriculture Forests Shrublands
Fire Starts and Human Activities 21
that fres of different causes can have different spatial and temporal patterns. Thus
having more information about fres investigated (with known causes) and analysing
them separately by cause type (e.g. as we did for shepherd-caused fres), will also
contribute to improving existing knowledge.
Further important research topics should also include measures of temporal fre
activity, such as predictions on number of fres per day in relation to natural and
anthropogenic variables. Additionally, as only a low proportion of ignitions result in
large burned areas, from a management perspective it would be important to further
understand the characteristics of ignitions that result in large fres, as these should
become a priority for initial attack and suppression. This was addressed in the case
of Portugal, but more research is needed on this important topic.
References
APAS 2004. Estado del Conocimiento sobre las Causas de los Incendios Forestales en
España. Asociación para Promoción de actividades Socioculturales, Madrid. 54 p.
Badia-Perpinyà, A. and Pallares-Barbera, M. 2006. Spatial distribution of ignitions in
Mediterranean periurban and rural areas: the case of Catalonia. International Journal of
Wildland Fire 15: 187–196.
Bassi, S. and Kettunen, M. 2008. Forest Fires: Causes and Contributing Factors in Europe.
European Parliament, Brussels. 56 p.
Cardille, J.A. and Ventura, S.J., Turner, M.G. 2001. Environmental and social factors
infuencing wildfres in the Upper Midwest, USA. Ecological Applications 11(1):
111–127.
Catry, F.X., Damasceno, P., Silva, J.S., Galante, M. and Moreira, F. 2007. Spatial
distribution patterns of wildfre ignitions in Portugal. Proceedings of the 4
th
International
Wildland Fire Conference. CD. Seville, Spain.
Catry, F.X., Rego, F.C., Bação, F.L. and Moreira, F. 2009. Modelling and mapping the
occurrence of wildfre ignitions in Portugal. International Journal of Wildland Fire 18:
921–931.
Catry, F.X., Rego, F.C., Moreira, F. and Bação, F. 2008. Characterizing and modelling the
spatial patterns of wildfre ignitions in Portugal: fre initiation and resulting burned area.
In: de las Heras J., Brebbia C.A., Viegas D. and Leone V. (eds.). International Conference
on Modelling, Monitoring and Management of Forest Fires. WIT Press, Vol. 119:
213–221, Toledo, Spain.
Catry, F.X. and Rego, F.C. 2008. Caracterização e análise dos padrões temporais e espaciais
dos fogos relacionados com o pastoreio. In: Botelho, H.S., Bento, J.M., Manso, F.T. (Co-
ords.) A relação entre o pastoreio e os incêndios forestais. UE Forest Focus Programme.
UTAD, AFN.
Chuvieco, E., Allgöwer, B. and Salas, J. 2003. Integration of physical and human factors in
fre danger assessment. In ‘Wildland Fire Danger Estimation and Mapping. The Role of
Remote Sensing Data’. Chuvieco, E. (ed.) Vol. 4: 197–218.
EC 2008. Forest fres in Europe 2007. Report No 8. European Commission, Joint Research
Centre, Institute for Environment and Sustainability. Ispra, Italy. 80 p.
FAO 1986. Wildland fre management terminology. FAO Forestry Paper 70, Food and
Agriculture Organization of the United Nations.
Finney, M.A. 2005. The challenge of quantitative risk analysis for wildland fre. Forest
Ecology and Management 211: 97–108.
22 Towards Integrated Fire Management
Goldammer, J.G. and Crutzen, P.J. 1993. Fire in the environment: Scientifc rationale and
summary of results of the Dahlem Workshop. In: Goldammer, J.G. and Crutzen, P.J.
(eds.). Fire in the environment: The ecological, atmospheric and climatic importance of
vegetation fre. John Wiley & Sons, Toronto. Pp. 1–14.
Lloret, F., Calvo, E., Pons, X. and Dìaz-Delgado, R. 2002. Wildfres and landscape patterns
in the Eastern Iberian Peninsula. Landscape Ecology 17: 745–759.
Loboda, T.V. and Csiszar, I.A. 2007. Assessing the risk of ignition in the Russian Far East
within a modelling framework of fre threat. Ecological Applications 17 (3): 791–805.
Mazzoleni, S. 2007. Spatial and temporal statistics on wildfre regimes across Europe.
Deliverable D4.1-1a of the Integrated Project “Fire Paradox”, Project no. FP6-018505.
European Commission. 14 p.
Mercer, D.E. and Prestemon, J.P. 2005. Comparing production function models for wildfre
risk analysis in the wildland–urban interface. Forest Policy and Economics 7: 782–795.
MMA 2007. Los incendios forestales en España. Decenio 1996–2005. Ministerio de
Medio Ambiente, Centro de Coordinación de la Información Nacional sobre Incendios
Forestales. Madrid, Spain. 106 p.
NWCG 2006. Glossary of wildland fre terminology. National Wildfre Coordinating Group,
PMS 205 (Boise, ID).
Pausas, J. and Keeley, J. 2009. A Burning Story: The Role of Fire in the History of Life.
BioScience 59: 593–601.
Thirgood, J.V. 1981. Man and the Mediterranean Forest a History of Resource Depletion.
Academic Press, San Diego 194 p.
Turner, M.G. and Romme, W.H. 1994. Landscape dynamics in crown fre ecosystems.
Landscape Ecology 9: 59–77.
UN 2002. Forest fres in Europe 1961–1998. International Forest Fire News 27: 76–80.
FAO/UNECE, United Nations (UN). Available at: www.fre.uni-freiburg.de/iffn/iffn_27/
content27.htm
Vasconcelos, M.J.P., Silva, S., Tomé, M., Alvim, M. and Pereira, J.M.C. 2001. Spatial
prediction of fre ignition probabilities: comparing logistic regression and neural
networks. Photogrametric Engineering & Remote Sensing 67(1): 73–81.
Vasilakos, C., Kalabokidis, K., Hatzopoulos, J., Kallos, G. and Matsinos, Y. 2007.
Integrating new methods and tools in fre danger rating. International Journal of Wildland
Fire 16(3): 306–316.
Vazquez, A., Moreno, J.M. 1998. Patterns of lightning- and human-caused fres in peninsular
Spain. International Journal of Wildland Fire 8(2): 103–115.
Vega-Garcia, C., Woodard, P.M., Titus, S.J., Adamowicz, W.L. and Lee, B.S. 1995. A logit
model for predicting the daily occurrence of human caused forest fres. International
Journal of Wildland Fire 5 (2): 101–111.
Vélez, R. 2009. The causing factors: a focus on economic and social driving forces. In:
Birot, Y. (ed.). Living with wildfres: what science can tell us. EFI Discussion paper 15.
European Forest Institute, Joensuu. Pp. 21–25.
Weiner, S., Xu, Q.Q., Goldberg, P., Liu, J.Y. and Bar-Yosef, O. 1998. Evidence for the use of
fre at Zhoukoudian, China. Science 281: 251–253.
Yang, J., Healy, H.S., Shifey, S.R. and Gustafson, E.J. 2007. Spatial patterns of modern
period human-caused fre occurrence in the Missouri Ozark Highlands. Forest Science 53
(1): 1–15.
2.2 Flammability: Infuence of Fuel on Fire
Initiation
Barbara Ubysz
1
and Jean-Charles Valette
2
1
Forest Research Institute, Raszyn, Poland
2
National Institute for Agricultural Research, Avignon, France
2.2.1 The fre initiation phase
Fire initiation consists of two phases in a wildfre growing process: (i) fre ignition
and (ii) fre initial extension (until the arrival of the fre fghters).
Fire ignition regroups fuel reaction to energy inputs (drying and thermal
degradation as endothermic phases, and combustion of flammable gases and
charcoal as an exothermic phase); energy is input either by conduction (contact),
radiation and convection.
Fire initial extension is piloted, step-by-step, by the combustion of the fuel
particles located close to the ignition point. During this initial phase, if the burning
conditions are stable and if the fuel is homogeneous, the pattern of the propagation
of the fre front is circular; the centre of the circle is the ignition point and the radius
of the circle is related to the spread rate of the fre. The front shape will be modifed
by changes in the burning conditions (gusts of winds, local relief) and local fuel
heterogeneities.
The initial phase is common to all fres whatever the fnal size of the burned area,
whatever the problems and diffculties faced by the fre fghters to control and fnally
to extinguish it, and whatever the loss of human life and the associated damage.
Each catastrophic wildfre is always a fre, ignited at a given point or on a given
small area, which grows in accordance with its own dynamic during the initial phase
and may continue to spread despite the existing prevention means and the fghting
means used to reduce its intensity and its rate of spread.
Ignition is the consequence of the mix of fuel particles, oxygen and heat input at a
given location and at a given moment.
The phase of the thermal degradation of a fuel particle is endothermic (i.e.
absorbing energy ), the combustion phase either of fammable gases or of charcoals
is exothermic (i.e. releasing energy). The basic phenomenon can be split into three
sequential steps:
1. water is vaporised from the fuel particle;
2. (phase of pyrolysis) fammable gases associated with the oxygen of the air
produce carbon-monoxide and carbon-dioxide and water;
3. charcoals and oxygen produce ashes and carbon-monoxide and carbon-dioxide.
24 Towards Integrated Fire Management
In general, industrial and structural fres ignite in closed spaces, where the frst step
of the combustion phase consumes the available oxygen very rapidly and produces
large amounts of black and toxic smoke. Wildfres are open fres where oxygen is
not the limiting factor; nevertheless, such smoke can also be produced during large
and intense wildfres.
Consequently, fre fghters and wildland managers cannot act directly on these
combustion phases and must act on other parameters to limit the energy released and
to reduce the fre extension.
2.2.2 The role of fuel particles characteristics
2.2.2.1 Chemical, thermal and physical characteristics
Given their role in the basic mechanisms, chemical, thermal and physical
characteristics of fuel particles have been largely studied specifcally to provide
input data to the models predicting the behaviour of fires. Eufirelab teams
promoted common methods to determine these characteristics (Allgöwer et al.
2007, Fernandes et al. 2006), and Fire Paradox partners have improved the cone
calorimeter method for these types of fuel particles (Rein et al. 2008).
Wildland and forest fuel is basically constituted of living material (needles and
leaves, twigs of various diameters, branches, trunks, and parts of the root system)
and of dead material, partially decomposed (litter and duff) whose characteristics
are altered during the decomposition process.
Fuel material is essentially composed of carbon, nitrogen, oxygen and hydrogen.
In addition, and depending on the species, their environment, their stage of
development, and their physiological state, this material also contained oligo-
elements which intervene in the photosynthesis and growth processes but which
very rarely plays a role in the initiation processes of a fre. Fuel material, particularly
issued from Mediterranean tree and shrubs species, is rich in organic oils and in
organic volatile compounds and terpenes (Pinaceae and Cupressaceae) whose fash
points are low; this characteristic largely contributes to the ignition and spread of
fre (Alessio et al. 2008).
Heat capacity (the amount of energy needed for heating the material) and heat
content (the amount of energy released by the combustion of the material) are two
important parameters for characterising wildland and forest fuel. Heat capacity
varies between 0.4 and 1.4 kJ/kg/K and heat content of most Mediterranean species,
either shrubs and trees, varies from 18 500 kJ/kg (oak leaves) to 24 000 kJ/kg (pine
needles and Ericacea foliage) (Dimitrakopoulos and Panov 2001): these species are
considered as highly energetic. Heat content depends on the richness in organic oils
and in other organic volatile compounds. Baker (1983) indicates a heat content from
18 600 kJ/kg to 23 200 kJ/kg for cellulose and hemicelluloses, and 25 600 kJ/kg
for lignin; consequently, heat content of wildland and forest fuel decreases with the
richness of the tissues in cellulose and increases with their richness in lignin.
Szczygieł et al. (1992) established the relationships between heat and ash contents
and fammability parameters and confrmed that ash content is fairly stable and does
Flammability: Infuence of Fuel on Fire Initiation 25
not signifcantly intervene in the combustion process as long as it remains lower
than 5%.
Physical characteristics of the wildland and forest fuel contribute to the ignition
and initial propagation processes. The larger the surface exposed to heat, the easier
the heat will be intercepted; this heat is released by the initial source or by the
combustion of fuel particles and transmitted either by radiation or by convection.
Surface-to-volume and mass-to-volume ratios are two of the most important
parameters for characterising the fuel particles. Surface and volume of leaves,
needles and twigs can be determined through geometrical approximations detailed
by Eufrelab members (Allgöwer et al. 2007; Fernandes et al. 2006). For more
complicated shapes, Fernandes and Rego (1998) proposed to evaluate the surface
by water immersion and weighing the mass of water which constitutes a flm around
the fuel particle. This water immersion method also allows the user to determine the
volume of the fuel particle by adapting the pycnometer method to the specifcities
of the fuel particles.
Fuel particles have been classifed according to their thickness (leaves, bark)
or their diameter (twigs and branches). Eufrelab members proposed to classify
Mediterranean fuel particles in four classes: coarse (>25 mm), medium (>6 mm to
25 mm), fne(>2 mm to 6 mm), and very fne (≤2 mm). The limits of the three frst
classes are the same as the American standards. The subdivision of the fnest class
in two classes (more than 2 mm and less than 2 mm) allows the user to separate
leaves, needles and fne twigs from larger fuel particles and to determine more
accurately the mass of the fuel consumed at the fre front of a prescribed burning,
given that very fne, fne and medium fuels are generally consumed at the fre front
of a wildfre.
2.2.2.2 Fuel moisture content
Fuel moisture content (FMC) also plays a paramount role in the reaction of wildland
and forest fuel to heat exposure. Water vaporisation uses an important amount of
energy due to the high heat capacity of water. It increases the concentration of non-
fammable gases in the mixture of gases of combustion, and reduces their ability to
initiate and sustain the combustion process by increasing the endothermic part of the
combustion reaction.
For live fuels, a given FMC is the balance between the water inputs throughout
foliage and roots system and the water outputs by plant evapo-transpiration. Each
species adopts specific strategies to resist drought, and to limit its own water
reserves at a level which permits the survival of the plant, or at least part of it. In
contrast, dead material is not regulated by physiological mechanisms; its moisture
content depends only on the water exchanges regulated by physical laws.
The moisture content of a given fuel is obtained by drying the fuel and calculated
with the commonly used formula FMC = 100 (Mi –Mf) / Mf
where Mi is the initial mass or wet mass of fuel and Mf is the fnal mass or its dry
mass.
26 Towards Integrated Fire Management
Unfortunately, some end-users prefer to express the moisture content using a
different formula:
WFMC = 100 (Mi – Mf) / Mi
FMC can easily be deduced from WFMC by this third formula
FMC = 100 WFMC / (100 – WFMC)
It is assumed that the mass loss (Mi-Mf) is exclusively the mass of water which was
inside the material and vaporised during the drying period. The international standard
drying temperature is 105°C, applied for inert materials like sand, cement or plaster.
Using this temperature level for drying wildland and forest fuel over-estimates the
mass loss, and consequently the moisture content, because at this high temperature
proteins are destroyed and oils and other organic compounds are either decomposed
or volatilised. Therefore, many Mediterranean, and more broadly European, teams
adopt a lower drying temperature. Eufrelab members proposed and adopted 60°C
for live and dead fuels; this temperature threshold respects the proteins and limits the
vaporisation of the largest part of the organic compounds. This lower temperature
induces a longer duration of exposure to reach the stabilised mass. This duration
depends on the type of oven, the structure of the fuel, the mass of the sample, the size
of the containers: in order to be sure to reach the oven-dry mass and for simplifying
the manipulation, Eufrelab members recommended a duration of 24 hours.
Other teams prefer to use a moisture “analyser” or “determination balance”
constituted by an infra-red dryer placed above the plate of the electronic balance:
this analyser displays the initial current and fnal weights during the drying process
and determines in real time the FMC. This device provides quicker results but
does not allow the users to carry out simultaneous replications or to compare
simultaneously several fuel types, or several species.
Because these two methods cannot be easily applied outside, Australian teams have
designed and developed a device based on the measurement of the electric resistance
of a pellet of a pellet made from ground-up fuel (Chatoo and Tolhurst 1997). Through
calibration tables, the FMC is directly displayed on the screen of the device. Fire
managers may adapt immediately their tactics and strategies as soon as the value is
displayed, for example the practitioners may stop the prescribed burning when the
moisture content of the litter or of the duff falls below the prescription range.
In general, Fire Danger Systems are based on fuel characteristics and on
topographical (slope and aspects) and meteorological data (local or regional
temperature, rainfall and wind regime). They combine their monitoring with the
monitoring of the moisture content of dead fuels (litter) and/or of living fuels
(leaves of shrubs and trees). Meteorological data can be continuously monitored,
and data are displayed as frequently as required. Moisture content of 1 hour, and
more commonly, 10 hour time-lag fuels might be acquired at a “high” frequency, but
generally only during the driest and hottest period of the day in order to catch the
relative minimum which corresponds to the peak of the daily fre danger. Depending
on the use of the parameter, and of the density of the network of plots where the fuel
is collected, FMC can be available daily or every two or three days.
Szczygieł et al. (2009) and Ubysz and Szczygieł (2009) present, as an example,
the Polish network of observation sites and the deduced map of fre danger (Figures
1 and 2).
Flammability: Infuence of Fuel on Fire Initiation 27
Figure 1. Polish network of fuel moisture content measurements.
Figure 2. Polish deduced fre danger information map.
28 Towards Integrated Fire Management
Figure 3 synthesises the evolution of the FMC of Erica arborea samples collected
during six dry summer periods (2001 to 2006) on one plot of the French network
which covers the French Mediterranean region (Forest Focus project).
During the summers 2002 and 2005, the drought was not too severe and Erica
arborea was able to limit the decrease of its FMC: the frst third of the variation was
reached after 25 days in 2005 and 30 days in 2002, and the frst half after 40 days in
2005 and 47 days in 2002.
During the other summer periods, these thresholds were respectively reached after
8 to 17 days and 17 to 28 days.
This kind of monitoring of several species in a given plot and several plots at
a given date improves the accuracy of the fire danger prediction by including
biological data. In a recent synthesis paper, Chuvieco et al. (2009) demonstrate that
live fuel moisture content can be used to predict the behaviour and the occurrences
of wildfres in specifc Mediterranean ecosystems.
2.2.2.3 Fuel fammability
To update maps of fre hazard, the maps need to be deduced from meteorological
and fuel moisture content data. This is generally based on the knowledge acquired
through previous events, logical evolutions and also fammability data.
Methods and main results of other studies on the fammability of wildland and
forest fuel have been described (Valette and Moro 1990; Moro 2006). Weise et al.
(2005) in preliminary tests indicate that the cone calorimeter can also be used for
Figure 3. Summer variations of Erica arborea FMC, 2001–2006.
0.45
0.50
0.55
0.60
0.65
0.70
0.75
0.80
0.85
0.90
0.95
1.00
1.05
1.10
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95
FMC/FMCi
Number oI days oI the monitoring period
Summer variations oI Erica arborea FMC
Var Department. Laquina D plot
2001
2002
2003
2004
2005
2006
2007
Flammability: Infuence of Fuel on Fire Initiation 29
relating time to ignition and moisture content. Dibble et al. (2007) indicate that the
cone calorimeter provides data that enable comparison of combustion properties
among many species. Fire Paradox partners improved the procedure in order to
characterise wildland and forest fuel and to provide the fre behaviour modellers
with accurate values of required input (Rein et al. 2008; Madrigal et al. 2009).
Guijarro et al. (2002) described an adapted method for fuel beds and concluded
that an increase of moisture content and of bulk density of litters implies an increase
of the time to ignition. It also implies a decrease of the rates of initial fre spread and
of combustion, and of the resulting fame height and fuel consumption ratio. Finally,
it implies a decrease of the risk of fre initiation. Jappiot et al. (2007) proposed a
classifcation of wildland urban interface areas, based on the fammability of the
dead litter fuels.
On www.eufirelab.org/specific.php, Moro offers results on Erica arborea;
measurements have been carried out during the summers from 1989 to 2007;
the graphs illustrate how fuel fammability is largely dependent on fuel moisture
content. Figure 4 summarises these results and a general equation, either linear or
exponential, linking FMC to the Mean Ignition Time (MIT):
MIT = 0.173*FMC + 2.117 R² = 0.84 MIT = 6.767*e
0.010FMC
R² = 0.82
Measurements of flammability parameters permit the classification of species,
which is of highest interest during the fres seasons. The following table summarises
the classification proposed by Valette and Moro (1990), Velez (1990) and
Dimitrakopoulos (2001). Even though the approaches and methods are different, the
common species are classed at the same level of fammability.
Figure 4. Mean Ignition Time of Erica arborea versus Fuel Moisture Content.
MIT ÷ 0.173 FMC ¹ 2.117
R
2
÷ 0.84
8
10
12
14
16
18
20
22
24
26
28
30
40 50 60 70 80 90 100 110 120 130 140 150
MIT (s)
FMC (°)
Erica arborea. MIT I (FMC)
1989 to 2007
30 Towards Integrated Fire Management
Flammability France Greece Spain
Extremely fammable Pinus halepensis, Pinus halepensis, Pinus halepensis
Quercus ilex, Pinus brutia, Quercus ilex
Erica arborea Quercus ilex
Highly fammable Arbutus unedo, Arbutus unedo, Arbutus unedo,
Quercus coccifera, Quercus coccifera, Quercus coccifera
Pistacia lentiscus Pistacia lentiscus
Moderately fammable Cistus salvaefolius Phlomis fruticosa, Cistus salvaefolius
Cistus salvaefolius.
Many teams have studied in laboratory conditions the mechanisms of wildland
initiation by incandescent objects like sparks and cigarettes. Xanthopoulos et al.
(2006) demonstrate that the probability that fre may be initiated by a cigarette
thrown from a car window is very small but not null, particularly when the car speed
is low and the road is narrow; this is the case for many European roads crossing
wildland areas and forests.
2.2.2.4 Conclusions
The above overview illustrates the role of the chemical, thermal and physical
characteristics of the particles of wildland and forest fuel on fire initiation. It
underlines the paramount role of the variations of fuel moisture content on the
variations of their fammability and consequently on the fre hazard they present, at
least during the driest and hottest seasons.
It is quite impossible to modify the chemical, thermal and physical characteristics
of the fuel particles of the wildland and forest fuel at a reasonable cost. However,
concerning the fuel bed, many treatments can reduce the fuel load and the fuel
porosity.
Likewise, fuel moisture content cannot be easily maintained at a low risk level
over large areas; the existing means are dedicated for very local application.
So, the recommendations towards wildland areas and forests managers, fire
fghters, end-users, owners, local or regional authorities that can be deduced, will
concern only limited areas, mainly wildland urban interfaces and vicinities of
structures (houses, workshops, factories, etc.) located in or close to wildland areas
and forests (Vélez 2009).
These recommendations will be related to the identifed wildfre causes. Martin et
al. (2005) describe the existing databases, analyse the content of three of them (from
France, Italy and Spain) and summarise the procedure followed by the Spanish
Nature Protection Service SEPRONA for wildfre investigation and prosecution.
Alexandrian (2008) analyses the wildfre causes following the new classifcation
used by the French Prométhée system.
Flammability: Infuence of Fuel on Fire Initiation 31
2.2.3 Coping with fuel fammability – practical recommendations
Except the areas concerned by volcanism which are well known, monitored and
managed, lightning wildfres occur during dry storms. Their frequency seems to
increase (Alexandrian 2008); this might be a consequence of the global warming.
This type of wildfre is typically ignited at a given point, where an enormous amount
of energy is released and the initiation phase is very dynamic. Recommendations
should be in the sense of:
• reinforcing research to identify the conditions under which such phenomenon
occur;
• associating the remote monitoring of lightning impacts to the radar monitoring
of rain to predict the localisation of fre initiation;
• using historical data to identify lightning-prone areas, develop a lightning
detection system and improve the accuracy of the predictive model of lightning
fres;
• reinforcing fre monitoring on lightning-prone areas and when the weather
conditions are favourable to lightning and dry storms.
Many fire initiations are related to works carried out in wildland areas and
forests during the periods of high fire danger (drought, wind). Generally, the
initial energy released is very weak and the initial phase is long. At that step, if
the weather conditions are not too severe, the extinction is easy as soon as the
workers have some tools and equipments to control and extinguish the fre. But,
this fre can smoulder for a long time and be detectable later after the departure of
the workers. To reduce the number of fre initiations related to “forest” activities,
recommendations should be in the sense of:
• equipping professional workers with fre fghting tools;
• integrating the prevention of fre initiations in the work organisation and
schedule;
• making aware the workers to the specifc fre risks of the area;
• excluding burnings during the summer period;
• including the days or periods of high fre danger in the “bad weather” clause
of the working contracts and, consequently, to exclude any penalty in case of
absence;
• forbidding any activity in wildland areas and forests during days or periods of
high fre danger, except for fre prevention and fghting.
Other fre initiations are related to unintentional and imprudent acts committed
either by local inhabitants or by tourists coming mainly from regions or countries
where the fre danger is generally low. In both cases, local or regional authorities
have to make them aware of the risk and to encourage them to change their personal
behaviour. In general the initial energy is weak and the fre initiation is detected
early. Recommendations should be in the sense of:
• reinforcing information on fre danger;
• improving information to common citizens;
• managing recreation areas to minimise the fre initiations (no car parks in
shrub or herbaceous strata, effcient ash trays, active information panels with
clear and simple recommendations towards the visitors, presence of specialised
personnel);
32 Towards Integrated Fire Management
• prohibiting any introduction of “hotspots” in these areas like barbecues or
ovens;
• installing garbage bins and encouraging visitors to take the garbage back with
them;
• closing and suppressing refuse areas, and in a frst step to banish any access to
them, and develop adapted refuse centres.
Other wildfre initiations are related to accidents or events which occur along or at
the vicinities of equipments like electric power lines, highways, roads and opened
paths, railways. Initial energy can either be very weak (a spark from an engine, a
cigarette discarded from a car) or larger in case of car, van or lorry crashes. To limit
the risk of fre initiation, recommendations should be in the sense of:
• managing along the highways a physical barrier to maintain cigarettes
discarded from vehicles;
• promoting on both sides of the highway the planting of native shrubs and tree
species with low fammability;
• making the highways users aware of the local fre danger through active
luminous signs;
• managing the borders of open roads and paths by reducing the existing live and
dead fuels;
• encouraging railway companies to implement brakes that do not emit sparks
and windows that prevent passengers throwing cigarettes outside;
• managing the borders of the railways by reducing the existing live and dead
fuels;
• encouraging the burying of electric power lines (high, medium and low
voltage) in order to reduce the risks of power discharges between the line and
the ground vegetation;
• reinforcing the efforts to reduce the amount of fuels below the electric power
lines which can not be buried.
Some management measures can be recommended in wildland urban interfaces and
in the vicinities of structures built in or close to wildland areas or forests, such as:
• promoting the selection, development or introduction of shrub and tree species
presenting lower fammability;
• encourage the reduction of fuels on the interface and in the vicinities of the
structures;
• making the inhabitants aware of the specifc risks they will face during the
high danger period;
• encouraging the inhabitants to clean the immediate border of the houses,
including roof, rain gutters and woody shutters;
• collecting rain water and store it in a tank for fre fghting purposes;
• promoting less fammable building materials for houses.
Finally, the last category of wildland initiation causes is related to the acts of
pyromaniacs and arsonists. The first sub-category concerns persons who need
psychiatric treatments. On the contrary, arsonists are completely responsible of their
acts; all the above recommendations are ineffcient.
Flammability: Infuence of Fuel on Fire Initiation 33
These acts may be based on neighbours’ disputes or related to hunting dissensions;
mediator’s interventions may be effcient to solve problems.
When the arsonists’ acts are based on soil occupation, a dissuasive decision is to
prohibit the building of houses or any other structures into wildland areas or forests,
which have been partially or totally burned; this prohibition can be applied during
the following ten, twenty or thirty years. In Mediterranean areas where the pressure
of the landed property is important, such a regulation must be strongly supported by
the population and the public authorities at local, regional and national levels.
References
Alessio, G.A., Peñuelas, J., Llusià, J., Ogaya, R., Estiarte, M. and De Lillis, M. 2008.
Infuence of water and terpenes on fammability in some dominant Mediterranean species.
International Journal of Wildland Fire 17: 274–286.
Alexandrian, D. 2008. Les statistiques “feux de forêt” de ces trente dernières années. Forêt
Méditerranéenne XXIX(4): 377–384.
Allgöwer, B., Calogine, D., Camia, A., Cuiñas, P., Fernandes, P., Francesetti, A., Hernando, C.,
Koetz, B., Koutsias, N., Lindberg, H., Marzano, R., Molina, D., Morsdorf, F., Ribeiro, LM.,
Rigolot, E. and Séro-Guillaume, O. 2007. Eufrelab: Methods for Wildland Fuel Description
and Modelling: Final Version of the State of the Art. Deliverable D-02-06. 57 p.
Baker, A.J. 1983. Wood fuel properties and fuel products from woods. In: Proc. Fuel Wood
Management and Utilization Seminar, Nov. 9–11, 1982, Michigan State Univ., East
Lansing, MI. Pp 14–25.
Chatto, K. and Tolhurst, K. 1997. The development and Testing of the Willtronics T-H Fine
Fuel Moisture Meter. CFTT, Centre for Forest Tree Technology, Fire Management Branch,
Department of Natural Resources and Environment, Research Report No. 46. 44 p.
Chuvieco, E., González, I., Verdú, F., Aguado, I. and Yebra, M. 2009. Prediction of fre
occurrence from live fuel moisture content measurements in a Mediterranean ecosystem.
International Journal of Wildland Fire 18: 430–441.
Dibble, A.C., White, R.H. and Lebow, P.K. 2007. Combustion characteristics of north-
eastern USA vegetation tested in the cone calorimeter: invasive versus non-invasive
plants. International Journal of Wildland Fire 16: 426–443.
Dimitrakopoulos, A.P. 2001. A statistical classifcation of Mediterranean species based on
their fammability components. International Journal of Wildland Fire 10: 113–118.
Dimitrakopoulos, A.P and Panov, P.I. 2001. Pyric properties of some dominant
Mediterranean vegetation species. International Journal of Wildland Fire 10: 23–27.
Fernandes, P.M. and Rego, F.C. 1998. A New Method to Estimate Fuel Surface Area-to-
Volume Ratio Using Water Immersion.. International Journal of Wildland Fire 8: 121–128.
Fernandes, P., Allgöwer, B., Guijarro, M., Cuiñas, P., Hernando, C., Mendes-Lopes, J.,
Morsdorf, F., Ribeiro, L.M. and Valette, J.C. 2006. Eufrelab: Methods for wildland fuel
description and modelling: Commonly used methods. Deliverable D-02-05. 7p
Guijarro, M., Hernando, C., Díez, C., Martínez, E., Madrigal, J., Lampin-Cabaret, C., Blanc,
L., Colin, P.Y., Pérez-Gorostiaga, P., Vega, J.A. and Fonturbel, M.T. 2002. Flammability
of some fuel beds common in the South-European ecosystems. IV International
Conference Forest Fire Research; Luso; 18–23 November 2002
Jappiot, M., Curt, T., Lampin, C., Borgniet, L., Vinet, O., Luois, S., Chandioux, O. and Estève,
R. 2007. Characteristics and Flammability of French Mediterranean Dead Litter Fuels.
Poster at the International Wildland Fire Conference, Seville, Spain, 13–17 May 2007.
34 Towards Integrated Fire Management
Madrigal, J., Hernando, C., Guijarro, M., Diez, C., Marino, E. and De Castro, A.J. 2009.
Evaluation of forest fuel fammability and combustion properties with an adapted mass
loss calorimeter device. Journal of Fire Sciences 27:323–342.
Martin, P., Bonora, L., Conese, C., Lampin, C., Martinez, J., Molina, D. and Salas, J. 2005.
Eufrelab: Towards methods for investigating on wildland fre causes. Deliverable D-05-
02. 54 p
Moro, C. 2006. Méthode de mesure de l'infammabilité du combustible forestier
Méditerranéen. Internal technical report. 14 p.
Rein, G., Torero, J., Guijarro, M., Esko, M. and de Castro, A. 2008. Methods for the
experimental study and recommendations for the modelling of pyrolysis and combustion
of forest fuels. Deliverable D2.1-1 of the Integrated Project “Fire Paradox”, Project no.
FP6-018505. European Commission. 83 p.
Szczygieł, R., Valette, J.C. and Moro, C. 1992. Teneur en cendres et pouvoir calorifque
supérieur d'espèces forestières méditerranéennes : relations avec les paramètres
d'infammabilité. PIF1992-01. 15 p + annexes.
Szczygieł, R., Ubysz, B., Piwnicki, J. and Kwiatkowski, M. 2009. Fire fammability of
forest plant formations in Poland. Deliverable D4.2-3 of the Integrated Project “Fire
Paradox”, Project no. FP6-018505. European Commission. 94 p.
Ubysz, B. and Szczygieł, R. 2009. Fire Situation in Poland. Instytut Badawczy Lesnictwa,
Internal Report. 34 p.
Velez, R. 1990. Preventing forest fres through silviculture. Unasylva 41.
Vélez, R. 2009. The causing factors: a focus on economic and social driving forces. In:
Birot, Y. (ed.). Living with wildfres: what science can tell us. EFI Discussion paper 15.
European Forest Institute, Joensuu. Pp. 21–25.
Valette, J.C. and Moro, C. 1990. Infammabilités des espèces forestières méditerranéenne,
conséquences sur la combustibilité des formations forestières. Revue Forestière Française
XLII numéro spécial 1990. Pp. 76–92
Weise, D.R., White, R.H., Beall, F.C. and Etlinger, M. 2005. Use of the cone calorimeter
to detect seasonal differences in selected combustion characteristics of ornamental
vegetation. International Journal of Wildland Fire 14: 321–338.
Xanthopoulos, G., Ghosn, D. and Kazakis, G. 2006. Investigation of the wind speed
threshold above which discarded cigarettes are likely to be moved by the wind.
International Journal of Wildland Fire 15: 567–576.
2.3 An Overview of Policies and Practices Related
to Fire Ignitions at the European Union Level
Cristina Montiel and Gema Herrero
Research Group Forest Policy and Socioeconomics, University Complutense of
Madrid, Spain
2.3.1 Introduction
Recent socio-economic changes, some forest management actions and other policy
measures outside the forest sector (e.g., environmental and nature protection
policies) have contributed to change the structure of some ecosystems enabling
the build-up of fuels, which makes ecosystems more fre prone and increases the
associated risk of large wildfres. Thus, the forest sector is facing new realities that
are causing changes in the European wildfre context.
Until now, the process to induce changes in wildfre policies or the adoption
of new political measures has been a fast ad hoc reaction to a past catastrophic
situation, rather than proactive mitigation before an emergency arises (FAO 1999).
This situation requires a shift in paradigm from a short-term fre policy, based
mainly on technological investments for emergency suppression measures, to a
longer-term preventive policy of removing the structural causes of wildfres and
reducing, as much as possible, potential damage (Birot 2009).
However, the shift from suppression-oriented policies to preventive and
integrative policies requires also an investigation of the human causes contributing
to wildfres in Europe
1
. Established fre policies provide the framework for social
prevention actions, including new governance methods and pre-extinction measures
(Aguilar and Montiel 2009a).
This holistic and proactive approach to wildfire management (Morugera and
Cirelli 2009) needs different formulations in Europe. In fact, wildfre issues receive
different consideration within the national policy instruments depending on the
risk severity of wildfres in the national contexts and the different political and
administrative systems existing in each country. In countries along the northern rim
of the Mediterranean Basin, wildfres are a subject of growing concern; in Northern
and Central European countries wildfres have been historically far less serious.
In general, the wildfre problem has become worse during the second half of
the 20
th
century. For example, land-use change dynamics have aggravated fire
hazards and disaster potential, especially in Southern Europe, in part because
the abandonment of rural areas, the prolonged protection of forest lands (and
consequently fre exclusion from the ecosystem), and the expansion of wildland
1 Workshop on Forest Fires in the Mediterranean Region: Prevention and Regional Cooperation. Report. Corpo
Forestale dello Stato, Rome, Italy – Joint Research Centre, European Commission – FAO. Sabaudia, Italy, 13–15
May 2008
36 Towards Integrated Fire Management
urban interface areas (Galiana et al. 2009; Martínez 2004; Vélez 2002, Vilar del
Hoyo et al. 2008).
The best instruments to introduce innovative approaches and recommendations
for the necessary paradigm shift towards a solution to the current wildfre problem
are a combination of legislation together with policy measures. However, the pattern
and spatial distribution of the fre problem in Europe are different in the different
regions and countries, and should be tackled differently at the regional and national
levels, taking into account wildfre incidence and socio-economic conditions.
2.3.2 Description and characterization of policy instruments
Depending on the decentralization level of European countries, issues of wildfre
management are handled principally within forest and civil protection policies at
the national and regional level. National and Regional Forest Plans usually include
preventive and curative actions to mitigate wildfre hazard, while civil protection
policies aim to protect human lives and goods.
The European Union (EU) has also considered the wildfres problem in their
political agenda since the 1980s through various regulations
2
that have encouraged
the Member States (MS) to reinforce prevention of wildfres. Recently, the European
Commission (EC) has adopted a Communication on the prevention of natural and
man-made disasters (COM(2009)82) that focuses on subjects where a common
approach is more effective than separate national approaches; for example,
developing knowledge, linking actors and policies, and efficient targeting of
community funds to prevention.
In agreement with national administrations, the main initiatives undertaken by the
EC regarding fre prevention include:
• The European Forest Fire Information System (EFFIS) that collects
information on forest fres (e.g. causes, risk levels) to create a common EU
Fire Database.
• Financial provisions to MS. Regarding support to the national prevention
policies, the LIFE+ instrument
3
co-fnance communication and awareness
campaigns for fre prevention; while the Rural Development Regulation
constitutes the main source of investment in infrastructure for wildfre
prevention.
• Support for scientifc research on various wildfre topics.
Information derived from the EFFIS database suggests that the pattern of wildfres
is not related only to climatic conditions, but also to socio-economic causes that
affect fre ignitions (Vélez 2009). In Europe, about 95% of wildfres are directly or
indirectly human caused.
2 Regulation EEC Nº 3529/86 on forest protection against fre; EEC Nº 2158/92 on protection of the Community´s
forests against fres; Forest Focus Regulation EC Nº 2152/2003 concerning monitoring of forests and environmental
interactions in the Community.
3 The LIFE+ Regulation (EC) No 614/2007 is the new EU fnancial instrument for the environment for the period
2007–2013 and includes provisions for funding forest fre prevention projects.
An Overview of Policies and Practices Related to Fire Ignitions at the European Union Level 37
Policy measures in each country are influenced by the perceived level of
threat, which varies with the intensity and scale of fire-related problems. The
Mediterranean Basin countries have by far the most and signifcant policies, and
implementation of regulations about fre prevention. In the rest of Europe (especially
in Northern countries) forest production goals are more important than forest fres
in forestry management. Fire occurrence and incidence are different from region
to region; therefore, national legal and policy frameworks should be adapted to
existing ecological and socio-economic characteristics of each European region (see
Figure 1).
Regulations are complemented and translated through planning instruments
outlining specific actions. To prevent and reduce fire ignitions, the key issues
considered within the national and regional policy frameworks are the following:

Figure 1. Wildfre legislation and zoning policy in Europe (by Montiel, C., UCM).
38 Towards Integrated Fire Management
• Causes investigation: although the investigation of wildfre causes needs further
development in many countries, it is an essential activity to reduce fre ignitions,
and it is becoming a more important priority in the Mediterranean region.
• Detection systems: detection also plays an important role in reducing fre
ignitions and fre propagation.
• Fire risk zoning: this means the demarcation of areas that are potentially
hazardous and where there is an increased fre risk. These areas normally
receive a special legal status (e.g. restrictions, special measures). In determining
fre risk zones new fre prone areas, like the wildland urban interface, require
special attention because specifc preventive measures might be needed.
• Determination of a ‘high-risk’ season: the declaration of high-risk fre seasons
is a decision taken by individual Member States. The duration of a high-risk
fre risk season differs from country to country and is a function of the climate
and the seasonality of fre causes. Mediterranean countries defne longer high-
risk fre seasons than Northern regions.
• Regulation of hazardous activities: in countries where these kinds of measures
are established, it is common to prohibit the use of fre in agro-silvo-pastoral
and leisure activities outside of the designated areas, and/or to prohibit the
use of fre at specifc times of the year. Likewise, to avoid fre starts there are
measures that regulate existing hazardous installations or their location.
• Placement of silvicultural treatments: (removal of brush, thinning or pruning)
are carried out adjacent to hazardous infrastructures areas either to prevent fre
ignitions in the area, or to prevent wildfres spreading from these areas.
• Use of fre as a tool: when used for achieving prevention objectives, prescribed
burning is directly related with the strategic reduction of fuel accumulation and
the avoidance of fre starts because of uncontrolled rural burnings.
• Existence of a legal framework for burning: the legal frameworks in the
Mediterranean Basin range from countries that do not contemplate the use of
fre and even prohibit it (e.g. Greece) to those that have developed regulations
and basic criteria and conditions for the use of fre (e.g. France, Portugal and
some regions in Spain and in Italy).
• Fire prevention campaigns: although public information and social awareness
about fre prevention receives relatively little attention in national legislation, it
is a common and relevant issue in national fre plans.
In addition to forest and civil protection policies, there are also other public policies
– spatial planning, agricultural and rural development policies, energy policy and
environmental policy – that infuence the structural causes affecting the wildfre
problem (Lázaro et al. 2009).
• Spatial planning has the potential to adequately coordinate all interventions
needed to respond to natural hazards (foods, wildfres, etc.) to achieve a more
sensible use of the territory. However, alongside urban policies, this policy
may affect one of the most important causes increasing wildfre risk: the
scattered settlements in forest areas.
• The evolution of the Common Agricultural Policy towards a real Rural
Development Policy brings about signifcant opportunities for intervention on
wildfre management such as: (i) the linking of wildfre prevention objectives
to local development processes; (ii) the recognition of agriculture’s territorial
An Overview of Policies and Practices Related to Fire Ignitions at the European Union Level 39
role in wildfre prevention; and (iii) the possibility to take advantage of the
new energy priorities to increase the use of forest biomass as a source of
energy.
• The growing perception of wildfres as one of the main environmental
problems for Europe, especially in the Mediterranean area, favours its
consideration as a priority for environmental management.
2.3.3 Territorial policies and underlying wildfre causes
Wildfres, especially in the Mediterranean region, are more than just a consequence
of periods of drought; they can also be considered an indicator of the socio-
economic differences between areas and their respective levels of development.
Some of the factors infuencing fre regimes in recent decades are closely related to
changing socioeconomic conditions. The most relevant are
4
:
• Rural depopulation, leading to the abandonment of rural areas and changes in
landscape because the invasion by natural and pioneering vegetation, which
may lead to increases in build-up of fuels, and therefore in fre risk. A subset
of this factor is the aging of rural population, which also increases wildfre risk
because of the loss of know-how in traditional burning practices.
• Concentration of population in urban areas, which increases the wildland urban
interface. The construction of new residences (frst or second homes) near
surrounding vegetation increases fre risk.
• Shifts in forestry policy priorities from the production of wood and other raw
materials to focusing on nature conservation, landscape management and
recreation.
Other factors influencing the temporal variation of fire regimes during recent
decades, such as climate change, changes in landscape confguration, changes in
ignition causes and the success of predominant fre suppression policies should
be also considered. Therefore, an important segment of the causes contributing to
the wildfre problem is left out of the traditional policy domains related to wildfre
management (Forest and Civil Protection policies) and must be supported by other
sectoral policies, which address the multiple dimensions of forest ecosystems such
as those related to the economy, natural resources and the environment (see Figure 2)
2.3.4 Risk practices in new fre scenarios: the need of social
prevention adapted measures
The main components of the wildfre problem in Europe – fre and territory – have
substantially evolved in the last decades, increasing vulnerability to wildfres. Fire
behaviour is evolving because there are new fuel patterns, which are the result of
4 Self Assessment, conclusions and recommendations of the Regional Session: Europe, Southeast Europe,
Mediterranean North Africa and Caucasus, held in 4
th
International Wildland Fire Conference (Seville, Spain. 2007).
40 Towards Integrated Fire Management
uncontrolled development and biomass accumulation. Spatial and human concerns
are also defning new territorial models, which create emerging fre prone areas and
new territories at risk (Vélez and Montiel 2003). The multiple combinations of fuel
patterns, new trends in spatial-planning and new social realities, have resulted in
new fre scenarios, which require new approaches to fre management.
At the regional scale, it is possible to recognize three territorial scenarios (see
Figures 3 and 4), showing different fre ignition patterns and different vulnerability
to wildfres, and therefore needing different fre management strategies:
• Scenario 1: Disadvantaged rural areas. These are rural areas in crisis,
suffering from rural abandonment, decline or abandonment of agrarian
activities, and without signifcant increase in new urban pressures, in which
forests have reclaimed the abandoned lands. The aging and declining
population has also meant the loss of the fre culture, hence turning its use into
a risk factor. Today the use of fre in traditional rural activities is set in different
spatial conditions in which continuity of woodlands and fuel load results in
an increase in wildfre risk. However, the use of fre by a neo-rural population
or by users of wildland recreational areas becomes a new ignition risk factor
because they are not able to guarantee fre control.
• Scenario 2: Dynamic rural areas. These areas are characterized by a
consolidated socio-economic context and productive forests, with no signifcant
problems of rural abandonment. The use of fre as a management tool in rural
activities has been generally maintained. Furthermore, the existing fre culture
can determine a type of ‘fre-dependent rural communities’, where fre use
regulation and governance mechanisms for fre management are strongly
needed.
Figure 2. Territorial policies in wildfre management (by Lázaro, A., UCM).
An Overview of Policies and Practices Related to Fire Ignitions at the European Union Level 41
Figure 3. Regional distribution of rural and urban areas in Europe (by Lázaro, A., UCM).
ND: no data; IR: intermediate rural areas; PR: predominantly rural areas; PU: predominantly
urban areas.
Figure 4. Rural development index in European regions (by Lázaro, A., UCM).
42 Towards Integrated Fire Management
• Scenario 3: Suburban rural areas and wildland urban interfaces. These are
areas with a predominance of urban developments, where agrarian activities
have been abandoned. The transitional areas between wildlands and urbanized
spaces – the wildland urban interface – represents an increasing fre risk factor
and are highly vulnerable.
The effectiveness of interventions on structural causes affecting wildfres depends,
to a large extent, on their adaptation to the spatial and socio-economic characteristics
of the area under consideration. The three different types of fre territorial scenarios
demand different policy and management measures regarding the biophysical and
social components of the fre problem.
Adapting the national and regional programmes of social prevention measures to
the fre territorial scenarios is very effective for reducing fre ignitions (Vélez 2010).
Social preventive measures refers to public information, social awareness and new
governance mechanisms allowing the existing stakeholders in each scenario to
participate in the integrated fre management scheme.
In the disadvantaged rural areas territorial scenario, the use of prescribed
burning techniques by professional teams to carry out fuel management tasks is
advisable. This strategy permits facing the main challenge of fuels management at
lower costs, and also adapting to the situation of low population levels and scarce
human resources. Nevertheless, in the dynamic rural areas scenario, it is convenient
to incentivize the culture of fire in rural communities by the enhancement of
best practices of fre use and, most importantly, to promote regulations and new
governance mechanisms to resolve conficts that are at the origin of fre ignitions.
Finally, raising awareness of wildfre risk should be a priority in metropolitan rural
areas and wildland urban interfaces.
2.3.5 New demands and challenges for wildfre management
Fire has always been present in Europe. It has been widely used as a tool in rural
practices. Nevertheless the wildfre policies adopted by most European countries
over the last century have been based on total fre exclusion. In the context of global
change and new policy paradigms, the negative effects of these national policies
and the loss of traditional fre use knowledge has become evident by analyzing
fre ignition frequency. At present, the main political challenges to overcome are
the adoption of long-term prevention actions, the development of new governance
mechanisms and the improvement of decision support tools, always guaranteeing
the safety of fre fghters and the inhabitants of burned areas.
Governance in wildfre management needs to be participatory at all stages of
the process including agenda setting and problem definition, elaboration and
formulation, implementation and enforcement, and evaluation and proposal of
policy changes. In addition, learning processes related to the appropriate use of fre
by rural populations and the reduction of human-caused wildfres can effectively
contribute to reducing fre occurrence through the management of social conficts
and the development of new cooperation opportunities. Alongside the need
for empowerment of local groups, measures should be taken with the intent to
An Overview of Policies and Practices Related to Fire Ignitions at the European Union Level 43
improve, frst, accountability when adopting preventive and remedial actions, and,
secondly, multi-sectoral coordination and the multi-level implementation of wildfre
management programs, especially in decentralized countries.
All stakeholders in the different fre scenarios should have the opportunity to
participate in the different stages of fire management, from information and
planning to cooperation in pre-extinction measures, according to their knowledge,
interests and concerns (Morguera and Cirelli 2009). Governments should defne
and implement the corresponding legal frameworks and platforms to make public
participation in fre management possible and effective.
Further, confict resolution systems concerning fre use in the Mediterranean
countries are greatly needed. Recognizing the benefts of some traditional fre uses
and accepting them instead of systematically suppressing them, would contribute
to effective fuel management strategies and to sustainable fuels load control. This
approach would most likely result in wildfre management cost reductions; for
example, by reducing the need for some silviculture treatments and in fghting fres
caused by rural activities. In fact, prescribed burning is an effective tool for fuels
management, at least in the defned territorial scenarios 1 and 2. However, this
technique, as well as the use of suppression fre, entails a social, policy and technical
challenge, which is the social acceptance and regularization of traditional fre use, as
well as cooperation between stakeholders and public administrations.
Therefore, participatory mechanisms, learning and lessons-drawing, know-how
transfer and training schemes are greatly needed. We need to move from a one-
dimensional perception of the negative impacts of fre to a more sophisticated one
that also stresses its positive effects (Aguilar and Montiel 2009b). In addition,
interagency and multi-level coordination is essential for wildfire management,
considering the complexity of the fre problem in Europe, where different physical
and human factors are interacting at different socio-spatial and temporal scales.
Finally, considerations must be given to the emergence of new fre prone areas in
Eastern, Central and Northern countries, and to development of new decision support
tools to evaluate new fre scenarios, namely the large and continuous areas with fuel
models of high risk and the wildland urban interface territories (Galiana et al., 2009).
2.3.6 Conclusions and recommendations
The national and regional wildfre policies in Europe should consider all aspects
of wildfire management as a whole and with a long-term view. An integrated
approach through the connection of the wildfres policies and the territorial policies
(e.g. spatial planning, rural development, energy policy) is needed to address
the structural causes of wildfre events. In fact, the starting point to set proactive
and effective wildfire management policies is the identification of the factors
driving fre ignitions, framed within their territorial contexts. The success of the
fre risk reduction actions adopted depends, to a large extent, on this preliminary
identifcation of causes.
Preventive measures should also be given more emphasis in fre policies as the
only way to avoid emergency situations and costly suppression actions. However,
although prevention is slowly gaining importance within national wildfre legislation
44 Towards Integrated Fire Management
Box 1. Wildland urban interface (WUI) as
new fre prone areas – the Spanish study case
Fire risk is higher in the WUI zone, where human presence and its related activities near
forest fuels could ignite fres. In Spain, WUI areas spread over more than 2% of the
national surface and account for around 1 100 000 hectares.
Galicia (9.1%), Asturias (8.9%), Canarias (7.3%) and Madrid (6.4%) have the highest
proportion of WUI areas; while the proportion is lowest in Castilla-La Mancha (0.5%)
and Aragón (0.6%).
During the last decade, there has been a growing concern about the WUI in the context
of wildfre management due to the rapid expansion of houses near or inside forest areas
and several catastrophic fre events.
Based on the Corine Land Cover data, during the period 1987–2000, the WUI surface
in Spain grew 6.8 %, resulting in 55 100 hectares of new interface areas.
Spatial data on WUI extent, location and their evolution over time provide key
information to develop effective fre planning in order to both avoid fre initiation and
prevent negative impacts for the population.
The WUI zone is considered in the Spanish forest and fire regulations, but its
consideration in planning documents is still very scarce. The general measures
considered towards the prevention of ignitions and possible propagation of fres from
urbanized areas to the forest masses include: fuel treatments around the structures,
awareness and information campaigns to WUI residents, and the regulation of potential
fre risk activities in urban areas.
Figure 5. Distribution of the WUI in Spain,
in 2000. Source: G. Herrero Doctoral Thesis,
in progress.
Figure 6. WUI spread in
Madrid from 1987 to 2000.
An Overview of Policies and Practices Related to Fire Ignitions at the European Union Level 45
and policies, it is still undervalued when compared to fre extinction. One of the
reasons for this unbalanced situation is that preventive actions are a constant task
over the year that does not have the same visibility as extinction investments and
that receives less public recognition.
On the other hand, wildfre prevention covers various felds of action (forest
fuel management, public awareness and education campaigns, fre risk analysis,
conciliation of interests, detection, etc.) and concerns different stakeholders (forest
land owners, building enterprises, town councils, etc.). Basic prevention regulations
should address the following aspects: (a) forest management to avoid fres; (b)
preventive silviculture measures in defensive and preventive infrastructures; (c) risk
zoning to delimit areas regarding their fre risk and regulated land uses and activities
according to fxed levels of risk; (d) establishment of risk periods in each country;
(e) basic regulation on traditional and prescribed burnings; (f) social prevention
measures (public awareness, governance mechanisms, etc.). New territories at risk,
such as wildland urban interfaces, deserve special attention and specifc preventive
measures should be considered as well.
Furthermore, prevention strategies should evolve according to spatial, socio-
economic and natural changes, and they should also adapt to different socio-economic
and territorial contexts, considering infuencing factors like spatial planning issues.
More coordination at the national, Mediterranean, and European level is also
needed to face the new fre scenarios. To do that, new governance mechanisms,
adapted to the diverse regional and local conditions, should develop participatory
and community-based tools for integrated fre management, based on a dialogue
between the different stakeholders and the public authorities.
Legal aspects should also be clarifed and enhanced through the development of
incentives and obligations concerning preventive actions. The regulation of fre use
practices for different land management purposes should be promoted through the
national and regional legal systems, in order to prevent and reduce fre ignitions as
a consequence of negligence, or arson related to grazing and agricultural activities.
In fact, prescribed burning could be an alternative or complementary technique for
fuel management, but carefully adapted to the different contexts and according to
the existing territorial patterns (rural abandoned areas, wildland urban interface,
productive rural regions, etc.).
Finally, existing EU funds provide basic financial support for national, sub-
regional and regional prevention measures, and must therefore be made available for
integrated wildfre management, particularly in Mediterranean countries, including
non-EU states.

References
Aguilar, S. and Montiel, C. 2009a. Pays méditerranéens. Prévenir les feux de forêt. Courrier
de la Planète 88: 19.
Aguilar, S. and Montiel, C. 2009b. New governance strategies for wildland fre management
in Mediterranean countries. In: Buttoud, G. (ed.): Change in Governance as Collective
Learning Process: Management, Politics and Ethics in Forestry, International Symposium.
Nancy, France, 21–24 June 2009. Pre-conference Proceedings. Pp. 9–18.
46 Towards Integrated Fire Management
Birot, Y. 2009. Living with wildfres: what science can tell us. EFI Discussion Paper 15.
European Forest Institute. 82 p.
FAO 1999. Report of meeting on public policies affecting forest fres, 14
th
session, Rome,
Italy, 28–30 October 1998, I. Part 1.
Galiana, L., Herrero, G. and Solana, J. 2009. A wildland-urban interface typology for forest
fre risk management in Mediterranean areas. Landscape Research Journal (under review).
Herrero, G. Las interfaces urbano-forestales como nuevos territorios de riesgo: análisis y
caracterización regional en España. Doctoral Thesis, in progress.
Lázaro, A., Galiana, L., Aguilar, S. and Sesbou, A. 2009. Assessment document on the main
strengths and weaknesses of the legislation and spatial policy instruments concerning
wildfre integrated management in the EU, in European member states and in North
African countries. Deliverable D7.3-1of the Integrated Project “Fire Paradox”, Project no.
FP6-018505. European Commission. 81 p.
Martínez, J. 2004. Análisis, estimación y cartografía del riesgo humano en incendios
forestales. Doctorate Thesis, Departament of Geography, University of Alcalá, Madrid.
Morguera, E. and Cirelli, M.T. 2009. Forest fres and the law. A guide for national drafters
based on the Fire Management Voluntary Guidelines, FAO Legislative Study 99, Rome.
Vilar del Hoyo, L., Martín, Mª P. and Martínez, J. 2008. Empleo de técnicas de regresión
logística para la obtención de modelos de riesgo humano de incendio forestal a escala
regional. Boletín de la AGE 47: 5–29.
Vélez, R. 2002. Causes of forest fres in the Mediterranean Basin. In: Arbez, M., Birot, Y.
and Carnus, J.M. (eds). Risk management and sustainable forestry. EFI Proceedings 42.
European Forest Institute. Pp. 35–42
Vélez, R. and Montiel, C. 2003. La problemática del monte mediterráneo. Investigaciones
Geográfcas 31: 121–137.
Vélez, R. 2009. The causing factors: a focus on economic and social driving forces. In:
Birot, Y. (ed.). Living with wildfres: what science can tell us. EFI Discussion Paper 15.
European Forest Institute. Pp. 21–25.
Vélez, R. 2010. Prescribed burning for grazing improvement and fre social prevention: the
Spanish EPRIF Programme. In: Montiel, C. (ed.): Best Practices of fre use. Prescribed
burning and suppression fre programmes in selected case studies regions. EFI Research
Report 24. European Forest Institute. In press.
3. Wildfre Propagation:
Knowledge and Search for Solutions
3.1 Humans, Climate and Land Cover as Controls
on European Fire Regimes
Joaquim S. Silva
1
and Sandy P. Harrison
2
1
Centre of Applied Ecology “Prof. Baeta Neves”, Technical University of Lisbon,
Portugal
2
School of Geographical Sciences, University of Bristol, United Kingdom
3.1.1 The driving factors of fre regimes
The expression ‘fre regime’ has become a rather loosely used term since it was frst
introduced by A.M. Gill more than three decades ago. Gill proposed a description
of a fre regime as including: the time between fres, the time since last fre, the
intensity of the fre, the type of fuel burned, and the season of occurrence (Gill
(1975). Since then, the expression has been widely used, and other authors have
proposed modifcations to the original defnition (e.g. Bond and Keeley 2005). Some
authors have used the expression to describe a particular fre, but more commonly
the expression has been used to summarize the characteristics of the fires that
typically occur at a site (Whelan 1995). Within the present text, ‘fre regime’ is used
in the latter sense, without particular regard to the precise set of fre characteristics
to be included.
Both climate and fuel characteristics infuence fre regimes, and there has been
considerable debate about which is the most important driving factor infuencing
the fre regime of a specifc region (Minnich et al. 1993; Keeley and Fotheringham
2001; Gillett et al. 2004; Archibald et al. 2009). It is not a straightforward exercise
to assess whether one factor or another factor is the most important, since that
depends on the type of fuels and climate (Bessie and Johnson 1995; Agee 1997). In
many regions this assessment is even less straightforward because of the infuence
of humans (Pausas and Keeley 2009); this raises another unsettled issue of whether
humans or natural factors are more important in determining fre patterns (Bowman
et al. 2009). Humans infuence fre regimes in three basic ways: by starting new
ignitions, by changing fuel characteristics, and by actively suppressing fires.
Given its long history of human occupation, including the traditional use of fre for
agriculture and slash and burn practices, these aspects have assumed a particular
importance in Europe as reported by different local studies (e.g. Carrion 2002;
Andric and Willis 2003; Tinner et al. 2005).
Therefore, in order to understand the mechanisms responsible for fre regimes it is
important to look at historical data. Charcoal preserved in lake or bog sedimentary
sequences provides a long-term record of changing fre regimes at a catchment scale.
The Global Palaeofire Working Group (Power et al. 2008) has compiled nearly
800 such records from around the world, and analyzed these to examine changes
in fre regimes on decadal to millennial timescales. Fire regimes show a strong
50 Towards Integrated Fire Management
response to temperature changes on centennial to millennial time scales. During
the last glacial (ca 80 000–20 000 years ago), fre increased during the repeated
abrupt warming events registered in the Greenland ice core and decreased during
the following cool periods (Daniau et al. in press). This trend is also seen during
the last 21 000 years (Power et al. 2008). Because of the increased number of sites
available, it has been possible to show that there have been opposing trends in
temperature changes between the northern and southern hemisphere (with warming
in the southern hemisphere corresponding to cooling in the northern hemisphere and
vice versa); this has resulted in asynchronous changes in fre regimes between the
two hemispheres. Records for the past 2000 years also show strong correspondence
between temperature changes and fre, both at global and regional scales (Marlon et
al. 2008). Temperature changes infuence fre regimes primarily through vegetation
productivity and hence fuel availability. In addition, cold climates are characterized
by a reduction in the vigour of the hydrological cycle, and thus by reduced lightning
and thus reduced ignitions. Temperature changes may also be important through
influencing the rate of curing of fuels. Marlon et al. (2008) also examined the
potential role of humans on recent fre regimes, and have argued that the major role
of humans in global fre regimes during the past 2000 years has been to suppress fres
during the twentieth century. The timing of the downturn in fre in several regions,
most obviously in the Americas, Australia and Asia, pre-dates the introduction of
active fre fghting, but corresponds with the expansion of large-scale agriculture
and ranching (Bowman et al. 2009). In regions that have not experienced a recent
expansion of large-scale agriculture and ranching, such as the northern hemisphere
boreal zone or Europe, there is no downturn in fre during the last century. Thus,
Marlon et al. (2008) argue that the decrease of fire activity was an accidental
consequence of human activities and resulted mainly from landscape fragmentation
and/or the reduction of fuel loads in intensively managed landscapes. The downturn
in fre activity in Europe during the frst part of the twentieth century has been seen,
though expressed less strongly, in inventory data (see e.g. Mouillot and Field 2005)
and is incorporated in the representative concentration pathway (RCP) scenarios
developed for the next International Panel on Climate Change assessment.
In this chapter we discuss the factors determining the fre regime in Europe, and
in particular the relationship between fre occurrence and land cover, by presenting
recent studies developed in different European regions: island of Sardinia (Italy);
Macedonia (Greece); Portugal; Ticino (Switzerland); and Poland (Figure 1).
3.1.2 Factors determining the fre regime in different European
regions
In many areas of Europe, and particularly in the Mediterranean Region, the last
decades have been characterized by drastic changes in land cover, especially
grazing pattern and land abandonment, combined with reforestation. This recent
development emphasized the need of addressing the problem of fre proneness,
in order to understand which land cover types are ‘preferred’ by wildfres. Fire
proneness assessment of land cover types has received recent attention from
Humans, Climate and Land Cover as Controls on European Fire Regimes 51
researchers (e.g. Cumming 2001; Nunes et al. 2005; Moreira et al. 2009). The
analysis of fre proneness within a certain region depends on the method used and on
the available data sets. In this context we must have in mind that typically, data sets
are short and may not necessarily capture a comprehensive range of land use types
and patterns. On the other hand, it is very important to have a suffcient number of
geo-referenced fre perimeters in order to allocate the burned areas to the different
land cover types. This is not always possible because not all countries and regions
have gathered this information over the years. The use of ignition points instead of
burned areas may be a way to overcome this diffculty, although it has to be taken
into account that in the European context, the mechanisms infuencing the start of
a fre are strongly correlated with human activities leading to ignitions (Catry et
al. 2009), whereas the mechanisms responsible for the distribution of burned areas
across the landscape, are essentially climate and fuel-driven.
Recent studies were developed aiming at assessing the infuence of these factors on
fre regime. The methodologies varied according to the data available, but also to the
approach chosen by each research team. Besides the basic aspect of fre proneness,
Figure 1. Map of burned areas in the European Union by region (NUTS3) and location of the
study regions. Adapted from San-Miguel and Camia (2009).
52 Towards Integrated Fire Management
other aspects like fre seasonality and fre size were addressed by the different studies,
aimed at characterizing the fre regime according to different types of land cover.
Sardinia (Italy)
On the basis of the available fre history data for Sardinia for the period 2000–
2004, Bajocco and Ricotta (2007) analyzed the fre proneness of given land cover
classes both in terms of fre frequency and mean fre size, but keeping both variables
separate. The results obtained from the analysis of 13 377 fres showed that for
most land cover classes, fre behaved selectively, in terms of both fre frequency
and fre size. Fire frequency was higher than expected by chance alone in urban and
agricultural areas whereas in forests, grasslands, and shrublands, fre frequency was
lower than expected. In grasslands and shrublands mean fre size was signifcantly
larger than expected from a random null model whereas in urban areas, permanent
crops, and heterogeneous agricultural areas, there was a signifcant resistance to fre
spread. Additionally, in terms of the mean fre size, forests and arable land did not
appear to be particularly fre prone, since both land use classes burned in proportion
to their availability.
A complementary study by Bajocco et al. (2009) was performed, aiming at
assessing the seasonality of wildfres for each land cover class. More specifcally, the
aim of the study was to analyze the temporal patterns of fre occurrence in Sardinia
during the period 2000–2006, in order to identify land cover classes where wildfres
occur earlier or later than expected from a random null model. The results showed
that the timing of fre ignition was selective for all land cover classes analyzed,
with very high signifcance levels. Wildfres occurred earlier than expected from
a random null model, in urban areas and in all agricultural classes. By contrast,
wildfres occurred later than expected, in forests, grasslands, and shrublands. There
was also a strong association between land cover classes and the Climatic Regions
of Sardinia. Therefore the study highlighted a close relationship between the timing
of fre occurrence and land cover, governed by two complementary factors: climatic
factors that determine the spatial distribution of land cover classes according to
climatic regions; and the human population that directly infuences fre ignition.
Using the same wildfire data set, Bajocco et al. (2008) have evaluated the
remotely-sensed phenological uniqueness of the Sardinian phytoclimatic regions
(Mediterranean, Transitional Mediterranean, and Transitional Temperate) following
the classifcation by Blasi et al. (2000) and to what extent these regions differed in
the temporal properties of their wildfre time series. The results obtained provided
a clear and coherent segregation of the major phytoclimatic units in terms of
phenological dynamics and temporal characteristics of wildfre regimes. The results
also emphasized a gradient in the mean starting Julian day of fres. For instance, the
earliest mean time of wildfre occurrence was usually related to the Mediterranean
phytoclimatic region, which was also the region that experienced the highest
number of fres, while the latest mean time of wildifre occurrence was associated
with the Transitional Temperate region. The authors concluded that there was a
strong climatic control on the temporal characteristics of wildfres, even in regions
of high human pressure.
Humans, Climate and Land Cover as Controls on European Fire Regimes 53
Macedonia (Greece)
Bajocco et al. (2008) assessed fre proneness by using fre ignition locations across
the region of Macedonia in northern Greece over a period of 13 years (1985–1997).
The resource selection ratio (Manly et al. 1993; Moreira et al. 2001) was used to
derive a ranking of land cover classes in terms of fre proneness. This ranking was
further analyzed across vegetation zones. The results were discussed in relation
to the moisture and temperature regime of vegetation zones (Quercion ilicis,
Ostryo-Carpinion, Quercion confertae, Fagion-Abietion cephalonicae, Pinetalia
nigrae, Vaccinio Picetalia) across the elevation gradient, population density and the
relative abundance of different land cover classes. It appeared that fre frequency
was conditioned by two opposing trends. On the one hand fre frequency decreased
in higher elevations where climate becomes cool and moist. On the other hand,
human population density is higher in lowlands and in warmer, drier zones
where agriculture is the dominant land cover. Here climate conditions were more
conducive to fre ignition but the landscape was characterized by land cover classes
which are generally less fre prone. Especially, in the Ostryo-Carpinion vegetation
zone, semi-natural land cover classes were so scarce that, overall, this zone had
fewer fres than expected by its area, indicating that fres have a lower likelihood
of starting in intensively managed agricultural landscapes, where fuel load is
kept at low levels. The authors mention the existence of a dynamic ‘tension zone’
corresponding to a higher fre incidence. This zone was characterized by a gradual
shift from warm and dry conditions to cool and humid conditions, but also a gradual
shift from intensively managed landscapes (relatively abundant, fire-resistant,
predominantly agricultural land cover classes) to less densely populated landscapes
(fammable semi-natural and forest land cover classes).
Portugal
Patterns of wildfire occurrence at the landscape level were characterized by
Moreira et al. (2009) during the period 1990–1994 in Portugal. Based on land cover
information within 5591 burned patches (larger than 5 ha) and in the surrounding
landscape, resource selection ratios were used to measure fre proneness for different
land cover classes in 12 regions of the country. Shrublands were the most fre-prone
land cover, whereas annual crops, permanent crops and agro-forestry systems
were the least fre-prone. In terms of forest types, conifer plantations were more
fre-prone than eucalyptus plantations, and broadleaved forests were the least fre-
prone. There were regional variations in susceptibility of land cover classes to fre,
which were discussed on the basis of differences in climate, management, ignition
patterns, fre fghting strategies, and the regional availability of each land cover
class. A cluster analysis of regional variations in selection ratios for all land cover
classes, allowed the identifcation of three main geographical areas with similar fre
selection patterns. The study also showed a signifcant positive correlation between
land cover availability and the resource selection ratio.
Based on the study by Moreira et al. (2009), Silva et al. (2009) restricted the
assessment of fre proneness to the main forest types in Portugal classifed according
54 Towards Integrated Fire Management
to the main species, using three different approaches: the use of resource selection
ratios applied to burned patches; the proportion of randomly located plots that
were burned; and the proportion of burned National Forest Inventory plots. The
results allowed ranking fre proneness according to the following decreasing order:
maritime pine (Pinus pinaster) forests, eucalyptus (Eucalyptus globulus) forests,
unspecifed broadleaf forests, unspecifed conifer forests, cork oak (Quercus suber)
forests, chestnut (Castanea sativa) forests, holm oak (Quercus rotundifolia) forests
and stone pine (Pinus pinea) forests. Silva et al. (2009) also used a different data
set (1998–2005 fire maps) to explore the influence of structural variables plus
stand composition on fre probability, by using univariate logistic models. From all
variables, stand composition was the most important for explaining fre probability.
Stepwise regression procedures were used to build a multivariate logistic model
which showed that stand structure plays a different role for different types of forest,
in terms of fre probability (see text box).
Ticino (Switzerland)
Fire proneness of given forest vegetation classes was analyzed both in terms of fre
frequency and fre size (mean and median) for the period 1982–2005 in Canton
Ticino (southern Switzerland), by Bajocco et al. (2008). With this purpose, the
authors investigated the data set in four fre groupings (all fres, anthropogenic
winter fres, anthropogenic summer fres, and natural summer fres) and performed
1000 random Monte Carlo simulations on frequency and size. Anthropogenic
winter and summer fres occurred mostly at low elevation in chestnut (Castanea
sativa) stands, broadleaved forests and in the frst 50 m from the forest edge. In
winter, half of the fres in chestnut stands were signifcantly larger than 1.0 ha and
the average burned area in some coniferous forests tended to be high. Lightning
fres seemed to prefer spruce (Picea abies) stands and to avoid the summer-humid
chestnut and beech (Fagus sylvatica) stands and the 50 to 100 m buffer area.
In beech forests, in mixed forests and in the spruce stands hit by natural fre in
summer, the fre size tended to be small. The obtained fre occurrence pattern and
especially the occurrence of anthropogenic fres in terms of fre frequency, seemed
to be also related with geographical parameters such as altitude and aspect, and to
anthropogenic characteristics such as the closeness of roads or buildings.
Poland
A database of the fres that occurred between 1996–2006 in Poland was used by
Szczygieł et al. (2009) to analyze the relationship between fre frequency and land
cover. This study revealed that the vast majority of fres occurred in non-forest
lands. Using the Corine land cover classifcation (Büttner et al. 2004), the study
concluded that 31.9% of fres occurred in non-irrigated arable land (37.3% of burned
area); 16.2% occurred in continuous urban fabric (8.6% of burned area); 11.2%
occurred in complex cultivation patterns (8.5% of surface area); 9.5% occurred in
land principally occupied by agriculture (17.4% of burned area); 8.8% occurred
Humans, Climate and Land Cover as Controls on European Fire Regimes 55
in coniferous forests (5.8% of burned area), 6.6% occurred in broadleaved forests
(8.3% of burned area) and 3.6% occurred in mixed forests (2.9% of burned area).
The authors refined the results for the Polish State Forests, using data from
2002–2008, including a total of 30 494 fres. The analysis covered the relationship
between the number of fres and the forest site types, the dominant species and the
stand age classes, using the resource selection ratio. The analysis confrmed the
existence of different levels of fre proneness according to site and stand conditions.
The sites with higher fre proneness were conifer forests, starting from the driest
sites to the moister ones. The lowest values (less than 1) corresponded to broadleaf
forests and to mountain forests. In terms of dominant stand species, the highest
Box 1. Case study of fre proneness: Portugal
Fire proneness was studied in Portugal by using 1998–2005 fre maps. These maps were
overlapped with the locations of the 1997–1998 Forest Inventory plots, in order to relate
forest characteristics (composition and vegetation density) with fre occurrence. With this
data it was possible to build a fre probability model using logistic regression procedures
(Figure 2). Results suggest that fre probability is higher in eucalyptus, maritime pine
and unspecifed broadleaf stands, whereas it is lower in holm oak and cork oak stands.
Moreover it seems that fre probability increases for higher vegetation density (given
by a Cover Index) for all species, except for unspecifed broadleaf stands (a decreasing
trend) and for eucalyptus (a mixed trend).
Figure 2. Probability of fre occurrence during years 1998–2005 as a function of a cover
index, for fve types of forest according to a logistic regression model. Adapted from
Silva et al. (2009).
56 Towards Integrated Fire Management
value was obtained by Scots pine (Pinus sylvestris) stands (more than 1). All other
stands had values less than 1, the lowest values being obtained by broadleaf species
and fr (Abies spp.) stands. Finally, in terms of age classes, the results showed that
fre proneness generally decreased for older stands, with resource selection ratio
values greater than 1 only for stands younger than 40-years-old. The exception
to this trend was class V, corresponding to stands older than 80-years-old, which
showed higher values than class IV (60–80-years-old).
3.1.3 Conclusions
Despite the wide range of methodologies used in the reported studies, and the
differences in spatial and temporal scales, some coherent conclusions emerge that
have practical consequences for management policies.
There is a fundamental distinction to be drawn between number of fre starts
and the impact, as measured by area burned, of fres. Individual case studies show
that the number of fres is strongly infuenced by human activities, being higher in
agricultural areas and wildland urban interfaces. However, many of these fres are
small. Although the importance of controlling human-caused ignitions should not
be underestimated, climate factors and vegetation type appear to be more important
drivers of large fires, as seen during the extreme weather conditions occurred
in 2003 in Europe (Trigo et al. 2006). The palaeo-record of biomass burning
essentially confrms these conclusions, showing a strong response of burned area
from decadal to millennial-scale climate changes over most of the last 21 000 years
with signifcant increases in fre during warmer intervals.
The individual case studies indicate that land cover, which is partially
infuenced by climate but also a consequence of human management, is important
in determining the availability of fuel and hence fre regimes. Humans can play
an important role in deliberately or inadvertently modifying the land cover and
therefore the fuel characteristics of a given region. The results on fre proneness
at local scales clearly emphasize the importance of adequate land cover planning
and proper fuel management for wildfre prevention. These analyses suggest that
policies on land-use planning should be driven to minimize fre hazard, taking
into consideration the high costs of fre suppression and fre prevention. Societies
should refect on the models of land-use transformation, which have been followed
in the last century, namely that which concerns the generalized abandonment of
agricultural felds and the establishment of forest plantations (Nunes et al. 2005;
Vélez 2009). In this latter case, prescribed burning as a fuel management technique
has been suggested to be an economically and ecologically advisable solution (De
Ronde et al. 1990; Fernandes and Rigolot 2007). Therefore, new solutions for land-
use planning and management should be envisaged, if societies are really committed
to minimize the problems caused by wildfres.
The diffculties reported in some of the above described research points to a
clear need for European countries to use common criteria for wildfre statistics, as
well as a common data set on land cover occupation. Specifcally, to improve fre
prevention, fre managers need information concerning the spatial and temporal
Humans, Climate and Land Cover as Controls on European Fire Regimes 57
distribution of fres across the landscape. In this framework, ecologically meaningful
landscape classifcations may provide geographical units with great potential for the
development of strategies for fre risk assessment and fre prevention. While land
cover directly controls what is actually burned by fres in terms of fuel load and fuel
continuity, it also infuences the seasonality of fres through its strong association
with the bioclimatic characteristics of the landscape. Therefore, ecologically suitable
land cover maps may be helpful for the development of strategies for fre hazard
assessment – for example, for optimizing the location of water points and fire
brigades, thus rendering fre fghting strategies more effective. On the other hand,
it is of paramount importance that individual fre reports include both a map of
the burned area and the corresponding ignition point location. This would provide
valuable information which can be used to better understand the different role of
humans, fuel and climate as controls on fre regimes. At the European level, the
European Commission Joint Research Centre, which manages the European Forest
Fire Information System, would be the right agency to support and lead this task,
aiming at achieving a sound, coherent and comprehensive fre database.
References
Agee, J.K. 1997. The severe weather wildfre – Too hot to handle? Northwest Science
71:153–156.
Andric, M. and Willis, K.J. 2003. The phytogeographical regions of Slovenia: a consequence
of natural environmental variation or prehistoric human activity? Journal of Ecology
91:807–821.
Archibald, S.A., Roy, D.P., Van Wilgen, B.W. and Scholes, R.J. 2009. What limits fre? An
examination of drivers of burnt area in Southern Africa. Global Change Biology 2008:1–18.
Bajocco, S., Angelis, A.D., Rosati, L., Ricotta, C., Moufis, G.D., Gitas, I.Z., Silva, J.S.,
Moreira, F., Vaz, P., Catry, F., Ferreira, P.G., Pezzatti, G.B., Torriani, D. and Conedera,
M. 2008. Publication describing wildfre regimes distribution in selected European study
areas. Deliverable D4.1-1b of the Integrated Project “Fire Paradox”, Project no. FP6-
018505. European Commission. 79 p.
Bajocco, S., Ricotta, C., Pezzatti, G.B. and Mazzoleni, S. 2009. Quantitative evaluation of
fre proneness of different land use classes. Internal Report IR4.1-2a of the Integrated
Project “Fire Paradox”, Project no. FP6-018505. European Commission. 33 p.
Bajocco, S. and Ricotta, C. 2007. Evidence of selective burning in Sardinia (Italy): which
land-cover classes do wildfres prefer? Landscape Ecology. Pp. 557–567.
Bessie, W.C. and Johnson, E.A. 1995. The relative importance of fuels and weather on fre
behaviour in Subalpine forests. Ecology 76:747–762.
Blasi, C., Carranza, M.L., Frondoni, R. and Rosati, L. 2000. Ecosystem classifcation and
mapping: a proposal for Italian landscapes. Applied Vegetation Science 3:233–242.
Bond, W.J. and Keeley, J.E. 2005. Fire as a global ‘herbivore’: the ecology and evolution of
fammable ecosystems. Trends in Ecology & Evolution 20:387–394.
Bowman, D., Balch, J.K., Artaxo, P., Bond, W.J., Carlson, J.M., Cochrane, M.A., D'Antonio,
C.M., DeFries, R.S., Doyle, J.C., Harrison, S.P. and others 2009. Fire in the Earth System.
Science 324:481–484.
Büttner, G., Feranec, J., Jaffrain, G., Mari, L., Maucha, G. and Soukup, T. 2004. The Corine
Land Cover 2000 project. EARSeL eProc 3:331–346.
58 Towards Integrated Fire Management
Carrion, J.S. 2002. Patterns and processes of Late Quaternary environmental change in a
montane region of southwestern Europe. Quaternary Science Reviews 21:2047–2066.
Catry, F.X., Rego, F.C., Bação, F.L. and Moreira, F. 2009. Modeling and mapping wildfre
ignition risk in Portugal. International Journal of Wildland Fire 18:921–931.
Cumming, S.G. 2001. Forest type and wildfre in the Alberta boreal mixedwood: what do
fres burn? Ecological Applications 11:97–110.
Daniau, A.-L., Harrison, S.P. and Bartlein, P.J. Fire regimes during the last glacial.
Quaternary Science Reviews. In press.
De Ronde, C., Goldammer, J.G., Wade, D.D. and Soares, R.V. 1990. Prescribed fre in
industrial pine plantations. Ecological Studies Analysis and Synthesis 84:216–272.
Fernandes, P.M. and Rigolot, E. 2007. The fre ecology and management of maritime pine
(Pinus pinaster Ait.). Forest Ecology and Management 241:1–13.
Gill, A.M. 1975. Fire and the Australian fora: a review. Australian Forestry 38:4–25.
Gillett, N.P., Weaver, A.J., Zwiers, F.W. and Flannigan, M.D. 2004. Detecting the effect of
climate change on Canadian forest fres. Geophysical Research Letters 31:1–4.
Keeley, J.E. and Fotheringham, C.J. 2001. History and management of crown-fre
ecosystems: a summary and response. Conservation Biology 15:1561–1567.
Manly, B., McDonald, L.L. and Thomas, D.L. 1993. Resource selection by animals:
statistical design and analysis for feld studies. Chapman and Hall, London.
Marlon, J.R., Bartlein, P.J., Carcaillet, C., Gavin, D.G., Harrison, S.P., Higuera, P.E., Joos,
F., Power, M.J. and Prentice, I.C. 2008. Climate and human infuences on global biomass
burning over the past two millennia. Nature Geoscience 1:697–702.
Minnich, R.A., Vizcaino, E.F., Sosaramirez, J. and Chou, Y.H. 1993. Lightning detection
rates and wildland fre in the mountains of Northern Baja-California, Mexico. Atmosfera
6:235–253.
Moreira, F., Rego, F.C. and Ferreira, P.G. 2001. Temporal (1958–1995) pattern of change in
a cultural landscape of northwestern Portugal: implications for fre occurrence. Landscape
Ecology 16:557–567.
Moreira, F., Vaz, P., Catry, F. and Silva, J.S. 2009. Regional variations in wildfre preference
for land cover types in Portugal: implications for landscape management to minimize fre
hazard. International Journal of Wildland Fire 18:563–574.
Mouillot, F. and Field, C.B. 2005. Fire history and the global carbon budget: a 1º ×1º fre
history reconstruction for the 20th century. Global Change Biology 11:398–420.
Nunes, M.C.S., Vasconcelos, M.J., Pereira, J.M.C., Dasgupta, N., Alldredge, R.J. and Rego,
F.C. 2005. Land cover type and fre in Portugal: do fres burn land cover selectively?
Landscape Ecology 20:661–673.
Pausas, J. and Keeley, J. 2009. A burning Sstory: the role of fre in the history of life.
BioScience 59:593–601.
Power, M.J., Ortiz, N., Marlon, J., Bartlein, P.J., Harrison, S.P., Mayle, F., Ballouche, A.,
Bradshaw, R., Carcaillet, C., Cordova, C. and others 2008. Changes in fre regimes since
the Last Glacial Maximum: an assessment based on a global synthesis and analysis of
charcoal data. Climate Dynamics 30:887–907.
San-Miguel, J. and Camia, A. 2009. Forest Fires at a Glance: Facts, Figures and Trends in
the EU. In: Birot, Y. (ed.). Living with Wildfres: what science can tell us. EFI Discussion
Paper 15. European Forest Institute. Pp. 11–18.
Silva, J.S., Moreira, F., Vaz, P., Catry, F. and Godinho-Ferreira, P. 2009. Assessing the
relative fre proneness of different fores types in Portugal. Plant Biosystems 143:597–608.
Szczygieł, R., Ubysz, B., Piwnicki, J. and Kwiatkowski, M. 2009. Fire fammability of
forest plant formations in Poland. Deliverable D4.2-3 of the Integrated Project “Fire
Paradox”, Project no. FP6-018505. European Commission. 91 p.
Tinner, W., Conedera, M., Ammann, B. and Lotter, A.F. 2005. Fire ecology north and south
of the Alps since the last ice age. Holocene 15:1214–1226.
Humans, Climate and Land Cover as Controls on European Fire Regimes 59
Trigo, R.M., Pereira, J.M.C., Pereira, M.G., Mota, B., Calado, T.J., Dacamara, C.C. and
Santo, F.E. 2006. Atmospheric conditions associated with the exceptional fre season of
2003 in Portugal. International Journal of Climatology 26:1741–1757.
Vélez, R. 2009. The causing factors: a focus on economic and social driving forces. In:
Birot, Y. (ed.). Living with wildfres: what science can tell us. EFI Discussion Paper 15.
European Forest Institute. Pp. 21–25.
Whelan, R.J. 1995. The ecology of fre. Cambridge University Press, Cambridge.
3.2 Fire Modelling and Simulation Tools
Jean-Luc Dupuy
1
and Daniel Alexandrian
2
1
National Institute for Agricultural Research, Avignon, France
2
Agence MTDA, Aix-en-Provence, France
3.2.1 Introduction
The scientifc observation of fres in natural conditions is a diffcult task. As for
many environmental phenomena, fire modelling has been a way to reduce the
amount of observations necessary for understanding and predicting fre behaviour.
Modelling approaches can be reduced to two: empirical and physically-based.
Empirical fre behaviour models are established on the basis of a reasonable
number of fre observations and predict the rate of spread or fame size of a fre
(Sullivan 2009b). Usually, simple equations give the fre rate of fre spread as a
function of a small number of parameters (wind speed, fuel characteristics, fuel
moisture, and slope). There are well-known empirical models developed in the
Australia, Canada, and the United States, and scientists have recently developed
similar models in Spain and Portugal.
Physically-based models are based on the principles of combustion and they
attempt to quantify the basic fre mechanisms (Sullivan 2009a). It has only been
in the last decade, however, that a full representation of wildfre physics has been
able to give predictions of fre spread. The so-called full physical models enable
three-dimensional (3D) simulations of fre spread at the scale of a forest stand (<20
ha), but usually on super-computers. These models solve the transport equations of
physics over time on a spatial grid, and predict many fre characteristics.
Despite the fast development of computer resources, it is not practical – so far
and in the near future – to use this new generation of physically-based models to
simulate fre at larger scales (10 km
2
) with resolutions on the order of meters or
smaller. The other aspects of these types of model, which limit their applicability is
the accessibility of the large amounts of vegetation and atmospheric data that they
require. Two-dimensional (2D) fre simulators have been developed in parallel to
provide decision-makers with powerful tools capable of predicting and mapping fre
behaviour in real conditions. Currently, GIS-based fre simulators can automatically
compute fre growth under heterogeneous conditions of terrain, fuels and weather.
However, because the results of the simulation have to be obtained in almost or
faster than real time, fre simulators can only use simplifed models (empirical or
semi-physical). A way – and one of the objectives of the Fire Paradox project – to
fll the gap between 3D physical models and 2D fre simulators is to use, as the
“engine” of a fre simulator, parametric laws ftted to the output of 3D physical
models (see Box 1).
62 Towards Integrated Fire Management
3.2.2 Pyrolysis, heat transfer and combustion
The ignition and spread of fre are governed by the physico-chemical processes of
pyrolysis, heat transfer and combustion (Albini 1992).
Fire ignition, which is the initiation step of fuel combustion, can occur only when a
heat source, fuel and an oxidizer are all present at the same time in the same physical
space at high enough temperature to react. In the particular case of woody fuels heat
transferred to the vegetation makes the temperature of the fuel rise. Beyond some
threshold ‘ignition’ temperature (about 300°C), the vegetation releases fammable
gaseous fuel at a high rate: this is the pyrolysis process. The gaseous fuels react with
oxygen, this is the faming combustion process releasing heat. Basically, fre spread
is possible after fre ignition and when the amount of heat transferred from the pre-
existing fre to neighbouring fuel is high enough to cause its ignition.
The rate of fre spread – i.e. the speed of the fre – depends on the strength of
the heat source, on the effciency of the heat transfer processes, and on the energy
required to raise the temperature of the fuel to the ignition temperature. In wildfres,
fne fuels such as grasses, twigs, and foliage are largely responsible for fre spread
due to their large surface area per unit volume, which increases their vulnerability
Box 1. Vesta
Vesta is the Large Scale Wildfre Simulator developed in the frame of the Fire Paradox
project. The basic idea of the software is to fll the gap between time consuming 3D
physical models and existing semi-empirical 2D fre simulators.
The physical model used to establish parametric laws is Firetec, thanks to an agreement
with the LANL (Los Alamos National Laboratory).
Besides the basic usual functions of a 2D-Fire simulator, its main features are the
following:
• it is able to work on various vector and raster GIS fle formats (import and export);
• as a ‘platform’, it is able to use various propagation models;
• it may be run with or without spotting models (namely the probabilistic model
developed in the frame of the Saltus project);
• it includes a wind simulator to assess the local variations of wind direction and
speed on the terrain;
• it allows fuel types description with the best accuracy available;
• it allows the user to simulate in an interactive way some human intervention as fuel
breaks, aerial dropping, etc.
• it enables launching a series of simulations on a given area to calculate hazard
maps;
• it enables the user to compare fre simulated with real wildfres to validate the
results obtained;
• maps and fres can be visualized in 2D and 3D.
Fire Modelling and Simulation Tools 63
to ignition. The dominant heat transfer processes between these fuels are radiation
and convection.
Fire spread is not only the propagation of fre on the terrain slope, but also the
transition of fre from surface fuels (litter, grasses, shrubs) to crown fuels (tree
canopy foliage). A fre that burns only the surface fuels is called a surface fre and a
fre burning also the tree canopy is called a crown fre.
Fire spread may be enhanced by the transport of frebrands some distance ahead of
the fre. These frebrands may ignite the fuel at their landing point and a secondary
fre – called a fre spot – develops. Spotting can occur up to kilometers from the main
fre front. Crown fres are the most likely to produce fre brands causing spotting.
Empirical or physically-based models have also been developed for fre spotting.
3.2.3 Physical processes and factors infuencing fre spread
Understanding how woodlands burn amounts to understanding how the basic
processes of pyrolysis, combustion and heat transfer are affected by the different
environmental factors and the nature of the complex coupling between them.
Knowledge of these effects, their relationships, and their quantifcation has been
the purpose of many studies based on observations, experiments and modelling.
Forest fuel moisture and wind are probably the most obvious factors affecting
fre spread. Because water content of the fuel must be evaporated before ignition
can occur and because the boiling of water requires a huge amount of energy, fuel
moisture content is the most important parameter infuencing fre start. Moisture
content also infuences fre spread and how much of the existing fuel can burn. This is
why so many efforts have been dedicated to measure or predict the moisture content
of vegetation as infuenced by meteorological conditions and biological cycles. In
natural fres, wind has mainly the effect to bend the fames towards the unburned
vegetation and convect heat to its surface. This enhances heat transfer to the unburned
fuel and it has been observed that fre rate of spread is more or less proportional to
wind speed at least for low to moderate wind speeds and for surface fres.
The terrain slope also affects fre spread and a steep slope is known to represent
very dangerous situations for fre fghters.
The amount of forest fuel and how it is distributed in the different vegetation
layers is also a factor affecting fre spread. In addition, fre intensity (also known as
fre power) is directly infuenced by the amount of fuel available to burn.
3.2.4 3D-Fire modelling
Fire modelling is a powerful technology for the simulation and prediction of fre
spread, even if only average fre behaviour is predictable.
Fire behaviour research has established signifcant correlations between the fre
environment – wind, fuel properties, slope – and fre rate of spread. This is why
empirical models succeed operationally in predicting fre spread within the range of
64 Towards Integrated Fire Management
environmental conditions under which they have been developed. Fire is, of course,
governed by the laws of physics and as such is not a completely random process.
These facts mean that some aspects of fre are predictable at some scales and this is
essential to decide strategies of fre prevention and mitigation.
The question of fire behaviour predictability however remains important.
Fire observations by experts of wind fows and recent advances in coupled fre-
atmosphere modelling have revealed that wind turbulence plays a dominant role in
erratic fre behaviour often observed in the feld. In addition, the turbulence level,
which measures the amount of local wind fuctuations, is increased by buoyancy
effects. Hot gases in a fre plume are less dense than ambient air and rise vertically
in the atmosphere due to the Archimedes’ principle; the driving force is called the
buoyancy force and the upward vertical motion of hot gases is accompanied by
downward vertical motion of fresh air. This physical situation is very unstable and
explains why high-intensity fres are likely to exhibit very erratic and dangerous
behaviour. A simple physical criterion based on the relative force of wind and
buoyancy has been proposed to identify such unstable situations, but the prediction
of the criterion itself remains delicate.
It is well known in the feld of fuid mechanics that turbulence effects are only
predictable on average. That means that one must expect great variations in
fre behaviour from one moment to another and from one point to another. Fire
practitioners or fre fghters operating on the ground must be aware of and account
for this variability, which is signifcant mainly at local scales (<1 km).
The natural heterogeneity of vegetation and mechanism of spotting are additional
sources of fre behaviour variability.
Physical models, and in particular coupled atmosphere-fre models that solve the
transport equations of physics to represent the fre, are powerful tools to explore
the effects of turbulence (and more generally non-linear effects) on fre behaviour,
provided that they use adequate computation and modelling approaches. Other
models, by their nature, cannot properly render the effects of turbulence. The fact
that the fuel is spatially-distributed in these models also enables the variability
caused by the non-uniform distribution of vegetation biomass to be accounted for.
Examples of use of full physical models in the feld of research are given in Box 2.
Global change is expected to cause important changes in the environmental
factors that influence fire start and spread. In particular, changes in vegetation
characteristics are expected. The frst consequences have been observed with the
decline of some conifer forests in the south of Europe: an immediate consequence
is the sudden reduction of moisture content of the vegetation. One can also expect
that some species not yet adapted to heat and dryness will adapt to new climate
conditions by reducing their plant moisture content or more likely by producing
dead biomass, as some Mediterranean species already do. Tree species are known
to have a large capacity for adaptation. One can eventually expect that by migration
new mixtures of plant species will take the place of species no longer adapted to
their environment. In order to assess the consequences of these changes on fre
behaviour, empirical models will be of small value since they are based on fre
observations made in existing vegetation types. In contrast recent physically-based
models are powerful tools to explore new environmental scenarios.
Fire Modelling and Simulation Tools 65
3.2.5 2D-Fire simulators
A two-dimensional fre simulator predicts fre growth over a map. In a real situation,
local conditions (i.e. topography, fuel and weather in each map pixel) change over
the map. Using spatial information along with weather and wind fles, the simulator
thus requires the support of a geographic information system (GIS).
Fire simulators can potentially be used by land management agencies and fre
fghting organisations for:
• wildfre training and educational sessions by conducting all kinds of ‘what-if’
scenarios;
• prevention and pre-suppression activities by assessing the effectiveness of
fuel treatments, testing the location of potential access roads, identifying
opportunities for the use of prescribed burning;
• operational purposes by supporting fre suppression activities and predicting
wildfre behaviour;
• making decisions by producing hazard maps and risk maps, economic
assessment.
The two North-American fre simulators Farsite (Finney 1998) and Prometheus
– the most used worldwide – have the same functionalities. Using deterministic
models (i.e. they produce output directly related to the input used by the user), they:
• can use point, line or polygon ignitions;
• can automatically compute wildfre growth and behaviour for long time periods;
• produce output that are GIS compatible;
• are based on existing fre behaviour models widely used (e.g. Behave, CFFDRS);
• enable some interactivity with the user (e.g. modifcation of fuel types,
simulation of aerial and ground fre suppression actions).
3.2.6 Linkage between 3D-fre models and 2D-fre simulators
The scale of prediction of physically-based models resolving transport equations on
a three-dimensional grid is still far from the landscape scale, and even in a long-term
perspective, the operational use of such models will be limited by the computational
resources, especially when a number of simulations are requested. This is the reason
why such models cannot be a direct component of a fre simulator at a landscape scale.
So, since such 3D-fre models have a limited use for operational purposes, one
way to overcome this limitation is to incorporate their results as the ‘engine’ of
2D-fre simulators.
As a first step, the 3D physically-based model is used to provide local fire
behaviour characteristics like rate of spread or fre intensity, by making simulations
over distances up to several hundreds of meters in the direction of propagation. The
model serves to implement a database of fre characteristics for various fuel types,
topographies and weather conditions. As a second step, the library of simulated
conditions is used to ft parametric laws that yield the rate of spread as a function of
wind, slope, fuel load, moisture content, and so on. Finally, these laws are applied to
local input data to predict fre spread over the terrain map using a contagion process.
66 Towards Integrated Fire Management
Box 2. 3D-modelling of fre behaviour and effects
Three-dimensional physically-based modelling of fre has been extensively used in Fire
Paradox in the frame of research studies on the understanding of fre behaviour and
effects mechanisms. Three different models have been used: (i) FDS, the Fire Dynamics
Simulator developed by the NIST (mainly devoted to building or structural fres); (ii)
WFDS (the FDS version for wildland fre simulation); and (iii) Higrad-Firetec, a coupled
atmosphere-fre model for fre behaviour simulation developed by the LANL (Los Alamos
National Laboratory) and the INRA (Institut National de la Recherche Agronomique).
Many questions arise from forest and fre managers about the effciency of surface fuel
reduction on tree survival, about the size and location of fuel breaks in a forest landscape,
or about the effect of fuel removal or modifcation on fre intensity. Simulations of
the effects of vegetation heterogeneity on fre behaviour have been run with Higrad-
Firetec, in the context of the evaluation of fuel break effciency. A simulation of fre
spreading over a fuel-break is shown as an illustration in Figure 1. Similar simulations
have been done in a pine (Pinus halepensis) stand using varied vegetation patterns on
the fuel break. These simulations revealed that: (i) a 25% cover of trees on the fuel
break decreases signifcantly fre intensity as compared to the untreated pine stand (75%
cover fraction), but 50% is not effcient; (ii) the size of the tree clumps does not affect
signifcantly the fre behaviour, except if very dense trees (unrealistic for pines) are used
to run the simulations.
The impact of a surface fre on tree crowns is usually measured through the percentage
of crown scorched in length or volume. A description of crown injury is often used as an
input to statistical models predicting tree mortality after fre. Empirical models for the
prediction of crown scorch height have been developed in the past, based on a simple
formulation derived from the plume theory. The most representative of such models is
the Van Wagner model for scorch height. Higrad-Firetec, as well as a two-dimensional
CFD (Computational Fluid Dynamics) code, has been used to explore the relevance
and implications of plume theory in the context of crown scorch height prediction: the
assumptions of the plume theory were found to be inadequate. This conclusion could be
extended to some models for crown fre initiation that also rely on plume theory.
Fire also impacts tree trunks, and tree mortality models often also account for this
effect. In an attempt to examine fow and thermal conditions on a tree trunk, when a
surface fre spreads around it, and so to understand the mechanisms of fre impact,
both computer simulations (WFDS) and laboratory experiments were conducted
as complementary approaches to the same problem. The results obtained by both
approaches were, in general terms, very similar, showing an increase in fame residence
time and on fame height in the leeward (near-wake) side of the trunk.
Wildfres also have specifc impacts in the wildland urban interface, in particular
on houses. FDS was used to simulate the thermal impact of a fire on a house and
also the transport of frebrands around a house. Indeed frebrands are often the cause
of house ignition at roof level. Regarding the thermal impact it was shown that two-
dimensional simulations can give useful results in terms order of magnitude, but that
three-dimensional simulations are required for deeper investigations. These simulations
coupled with ignition models for building materials (also developed in Fire Paradox)
should give an evaluation of the thermal risk in the future. Regarding frebrands, three-
dimensional simulations give indication of the risk they represent by monitoring of
brands at different positions around the house, but the effects of fow instability and
unsteadiness on results should be further investigated before considering the operational
use of such approach to fre impact measurement.
Fire Modelling and Simulation Tools 67
Box 2. Continued
Suppression fre is one possible use of fre against wildfres and as such could help
solving the fre paradox. 3D-modelling using Higrad-Firetec was one approach followed
to assess the feasibility and effciency of suppression fre. Field experiments have also
been conducted for the same purpose in Fire Paradox. Although it is diffcult for the
model to strictly reproduce the experiments, both approaches indicate that it is very
diffcult to fnd proper conditions for a safe and effcient use of backfres, when a wildfre
spreads over a fat terrain under the infuence of even moderate winds: the backfre is
expected to be “sucked” by the wildfre, but this was observed only under weak wind
conditions. As already known by fire managers, however, topography can help the
development of such a “sucking” phenomenon.
Figure 1. Numerical simulation of a fire spreading over a fuel break in a Pinus
halepensis stand, using the coupled fre-atmosphere model FIRETEC (INRA – LANL
joint work). Cover proportion of trees on the fuel break is 25%. The fre spreads along
the x-axis and the fuel-break extends from x=240 m to x=440 m. Flame contours in
red are deduced from computed isotherms. Green and yellow colour contours are used
respectively for tree canopy and surface fuels (shrubs, grasses) representations (iso-
contours of computed biomass density). Adapted from Pimont (2008).
68 Towards Integrated Fire Management
References
Albini, F.A. 1992. Dynamics and modelling of vegetation fres: observations. In: Crutzen,
P.J. and Goldammer, J.G. (eds.).Fire in the environment: the ecological, atmospheric, and
climatic importance of vegetation fres. Report of the Dahlem Workshop held in Berlin,
15–20 March 1992. 400 p.
Finney, M. A. 1998. FARSITE: Fire Area Simulator-model development and evaluation.
Res. Pap. RMRS-RP-4, Ogden, UT: U.S. Department of Agriculture, Forest Service,
Rocky Mountain Research Station. 47 p.
Lecomte, I., de Coligny, F., Griffon, S., Pimont, F., Rigaud, E., Rigolot, E. and Vigy, O.
2009. Fire Paradox fuel manager: software and user’s manual. Product P6.1-6 of the
Integrated Project “Fire Paradox”, Project no. FP6-018505. European Commission. 85 p.
Box 3. Fuel editor
The Fire Paradox Fuel Manager (Lecomte et al 2009) is a piece of computer software
integrated in the data processing chain between the European data and knowledge base
on fuels and the 3D physically-based fre propagation models. The scientifc objective
was the representation of vegetation scenes and their transformation into fuel complexes
including all the necessary parameters to run a fre behaviour model. The technological
objectives are to implement a user friendly platform to generate fuel complexes in 3D,
to provide tools for managing the European database on fuels, to visualize fre effects on
trees and simulate post fre vegetation successions.
A survey of available simulation platform technologies has led us to join the
Capsis project, dedicated to hosting a wide range of models for forest dynamics and
stand growth. A new CAPSIS module – “FireParadox” – has been developed which
implements data structure and functionalities of the Fire Paradox Fuel Manager.
A 3D vegetation scenes’ editor has been implemented allowing interactive
manipulative functionalities on vegetation scenes (e.g. zoom, rotation, etc) as well as
on vegetation objects (selecting, adding, updating) through a graphical user interface.
Several renderers are available to display 3D vegetation objects. Fire damage on
vegetation objects have been mainly focused on fre-induced tree mortality. Several fre
impacts on tree crown and trunk have been defned and can be visualized at the scene
scale. Moreover, several tools are available to display information (descriptive statistics,
indicators) on the vegetation scene content or on the current selection.
Several creation modes of vegetation scenes are available including loading of a pre-
existing inventory fle or the automatic generation of a new scene respecting a set of
constraints on species distribution.
The application is connected through the Internet to the European database on fuels
and manages the user’s rights.
An export module has been developed to prepare the set of fles necessary to run the
fre propagation model. Export fles describe the composition and the structure of the
fuel complexes taking into account the physical properties of various components of
the different vegetation layers (trees, shrubs, herbs and litter) composing the vegetation
scene. Further details can be found at http://freintuition.ef.int/fuel-manager.fre.
Fire Modelling and Simulation Tools 69
Pimont, F. 2008. Modélisation physique de la propagation des feux de forêts: effets des
caractéristiques physiques du combustible et de son hétérogénéité. Université de la
Méditerranée, Aix Marseille II. Doctoral Thesis. 323 p.
Sullivan, A.L. 2009a. Wildland surface fre spread modelling, 1990–2007. 1: Physical and
quasi-physical models. International Journal of Wildland Fire 18: 349–368.
Sullivan, A.L. 2009b. Wildland surface fre spread modelling, 1990–2007. 2: Empirical and
quasi-empirical models. International Journal of Wildland Fire 18: 369–386.
3.3 Wildland Urban Interfaces, Fire Behaviour and
Vulnerability: Characterization, Mapping and
Assessment
Corinne Lampin-Maillet
1
, Antonis Mantzavelas
2
, Luis Galiana
3
, Marielle Jappiot
1
,
Marlène Long
1
, Gema Herrero
4
, Oskar Karlsson
4
, Apostolopoulou Iossifna
2
,
Lazaridou Thalia
2
, Partozis Thanassis
2
1
Cemagref, Aix-en-Provence, France
2
OMIKRON Ltd, Greece
3
Department of Geography, Autonomous University of Madrid, Spain
4
Department of Geography, University Complutense of Madrid, Spain
3.3.1 Introduction
Large areas initially consisted of contiguous forests, particularly in Europe, were
infuenced by human activities to a large extent. This infuence contributed to the
fragmentation of the rural landscape and forested areas were found surrounded by
or intermixed with urban development. Urban and economic development in or
near wildland vegetation poses a major threat to the environment (Johnson 2001,
Radeloff et al. 2005a). They are areas of human-environment conficts, such as the
destruction of homes by wildfres, natural habitat fragmentation, introduction of
exotic species, and biodiversity decline (Radeloff et al. 2005a). These areas that are
characterized by increased human activities and land use conversion make up the
wildland urban interfaces (WUIs).
The signifcance of WUIs has grown in recent years mainly because WUIs, as
landscape units, have grown worldwide (Steward et al. 2003). Essentially in USA,
Canada and Australia interest in study of WUIs increased after the huge fres of
1985 (Davis 1990). Since 2000 in Europe the WUI becomes involved in the forest
fire environment. Residential areas are increasingly affected by wildfires with
damage on goods and people. Extreme fres affected Portugal, France and Spain in
2003, Spain in 2005, Portugal and Spain in 2006, and Greece in 2007. More crown
fres involve WUI zones, mainly during heat waves, destroying houses (Lampin-
Maillet 2008).
So large efforts, aiming at the identifcation and mapping of existing or potential
WUI areas have been recorded in North America and to a lesser extent, in Europe.
The aim was to assess fre risk in WUIs and particularly their vulnerability in order
to improve the effciency of prevention actions for the protection of WUIs from
wildfre propagation.
Wildfre risk can be defned as the expected loss due to wildfres on a certain
area and period of time. Many definitions of wildfire risk exist (Hardy 2005,
Camia et al. 2004). The most common corresponds to the combination of hazard
with vulnerability: risk = hazard*vulnerability, establishing relationships between
72 Towards Integrated Fire Management
theses three concepts (Chen et al. 2003). So fre risk encompasses two different
components: (i) the probability that a wildfre affects the area during that period
of time – fre hazard; and (ii) the potential damage that the fre will cause once it
occurs – vulnerability (Blanchi et al. 2002, Jappiot et al. 2009). Blanchi et al. (2002)
proposed to defne each component of fre risk as composed of different overlapping
elements (Table 1).
But we also fnd other fre risk defnitions where intensity is considered as part of
vulnerability (Jappiot et al. 2009, Wilson et al. 2005).
The assessment of forest fre risk is generally recognized as an indispensable
component of fre prevention and suppression systems. As fre fghting resources
and fre prevention infrastructures are not infnite, the need for predicting a wildfre
incident and its consequences becomes apparent. Resources need to be allocated
wisely, in space and time, for the sake of operation and cost effectiveness. Usually
fire risk assessment consists of a combination of fire hazard assessment and
vulnerability assessment.
• Various authors in risk-related issues have used the term fre hazard, in
different and contradicting ways (Camia et al. 2004). In the present work,
we consider fre hazard as the combined outcome of the probability of fre
occurrence and the potential intensity of the resulting fre, similarly to the
defnition used by Blanchi et al. (2002). We consider the probability of fre
occurrence as the likelihood of a fre to happen in an area. Finally, fre intensity
is considered as the potential energy release per unit length of fre front, in
the case of a fre incident (Byram 1959). Fire hazard is generally computed
as an index (San Miguel-Ayanz et al. 2003). Assessment methodologies are
categorized by their temporal scales (Marzano et al. 2004). If the assessment
is based upon factors, that change very slowly over time (e.g. vegetation,
topography), then the outcome would be a structural (static or long-term) index
(Jappiot et al. 2009). On the other hand, if the assessment is based upon factors
that are likely to change frequently (daily or even hourly, e.g. fuel moisture,
weather) then the outcome would be a dynamic (short-term) index.
• Concerning the other component vulnerability, there are many different concepts
of vulnerability across disciplines and topics (Gallopin 2006). According to
natural risks, vulnerability can be considered in three ways: as a consequence, as
a state or characteristic, or as a cause (Mantzavelas et al. 2008).
Table 1. Fire risk defnition with detailed elements.
FIRE RISK
Hazard Vulnerability
Occurrence Intensity
Ignition Wildfre Threatened Wildfre Stake Reply
Probability Probability area Intensity Response

Wildland Urban Interfaces, Fire Behaviour and Vulnerability: Characterization, Mapping... 73
In the frst case vulnerability is the value that could be lost when hazard occurs.
According to Coburn et al. (1994), vulnerability is defned as the degree of loss to a
given element (or set of elements), resulting from a given hazard at a given severity
level. Caballero et al. (2004) stated that vulnerability can be expressed through the
calculation of the potential damage when a single unit is exposed to a certain level
of danger.
In the second case vulnerability can be the propensity of an element or a set of
elements to suffer damage, when a hazard occurs (D’Ercole 1998). Commonly,
vulnerability may encompass the degree of fragility of men, organized societies,
economic structures, built environments and ecosystems to the negative
consequences resulting from the exposure to hazardous events. According to Blaikie
et al. (1994), vulnerability expresses the capacity of a person or group to anticipate,
cope with, resist to and recover from the impact of natural hazards.
Finally, in the third case vulnerability corresponds to a “system which considers
a lot of variables (natural and human). Their spatial and temporal dynamic can
produce situations which can be more or less dangerous for exposed society”
(D’Ercole 1996). In this case the aim is to identify the factors (variables) that are the
source of vulnerability.
In short, it is obvious that the defnition of the vulnerability refers mainly to the
impacts of a catastrophic event. The vulnerability of an element is usually expressed
as a percentage loss (or as a value between 0 and 1) for a given hazard severity level
(Blanchi et al. 2002).
The great challenge is to express vulnerability in measurable units or indices in
order to be used for further estimation of the total fre risk. Furthermore, most of
disaster mitigation work is focused on reducing vulnerability, and in order to do it,
there is a need to understand which elements or units are the most exposed at risk,
from the principal hazards which have been identifed (Coburn et al. 1994).
The concern of this chapter is the wildland urban interface and fire risk
assessment. It supposes to defne precisely this term. In geography, interface is
defned as the contact plan or contact line between two different systems (Brunet
et al. 1993). It constitutes a privileged zone to exchange, to interact between
two systems, specifcally human and wildland systems (Carroue et al. 2002). In
the literature, defnitions present WUI as the line, area, or zone where structures
and other human developments meet or intermingle with undeveloped wildland
or vegetative fuels. The term of WUI community exists with the following
definition “the urban-wildland interface community exists where humans and
their development meet or intermix with wildland fuel”, where houses meet or
intermingle with undeveloped wildland vegetation (USDA-USDI 2001). Nowadays
and more generally, the WUI commonly described as the area where urban areas
meet and interact with rural lands (Vince et al. 2005), includes the edges of large
cities and small communities, areas where homes and other structures are intermixed
with forests and other land uses, and islands of undeveloped lands within urban
areas (Alavalapati et al. 2005, McGee 2005, Caballero et al. 2004). In these WUI
increased human infuence and land use conversion are chancing natural resource
goods, services and management (Macie and Hermansen 2002).
The issue of the chapter is to map the location of WUIs on the territory and
to determine which WUIs are at the greatest risk. For that several objectives are
74 Towards Integrated Fire Management
developed. The frst objective is to present a method to characterize and to map
WUIs at a local scale with a view to improve fre prevention, and at a landscape
scale with a view to analyse their territorial development in the landscape. The
second objective is to propose a fre hazard calculation and mapping process taking
into account structural and daily factors. The third objective is to propose a method
to assess vulnerability levels. The fourth and last objective is to present a specifc
approach developed with a view to assess and to map fre risk in WUIs through a
total fre risk index. Tools have been developed as the result of scientifc research.
In order to develop the previous objectives, three study cases have been defned
in three European countries: France, Spain and Greece. The frst one is located
in southeastern France in the Metropolitan area between Aix-en-Provence and
Marseille (site 1 in Figure 1). The second one is located in southeastern Spain in
Sierra Calderona (site 2 in Figure 1). The third one is located in northeastern Greece
in the western part of the prefecture of Thessaloniki (site 3 in Figure 1).
3.3.2 Identifcation, characterization and mapping of WUIs in
European Mediterranean countries
In order to map WUIs on the territory, a precise defnition of the WUI has been
proposed as follows (Lampin-Maillet et al. 2010):
• WUIs are composed of residential houses, which are inhabited permanently,
temporarily or seasonally (agricultural, industrial, commercial and public
buildings were not taken into consideration);
• Houses are located at 200 m from forests or shrubland to consider area where
brush-clearing is partially required or spot fres occur (Colin et al. 2002);
Figure 1. Location of the three European study cases.
Wildland Urban Interfaces, Fire Behaviour and Vulnerability: Characterization, Mapping... 75
• WUI are delineated by a radius of 100 m around the houses. This distance
takes into account the perimeter wherein fuel reduction operations can be
imposed on home owners.
These different distances are particularly adapted to the European context but they
also could be changed according the specifc local context of the country (vegetation
clearance regulation or urban organization).
3.3.2.1 At local level (Lampin-Maillet et al. 2010)
Considering the above definition, the WUI area is located up to 300 m from
forests, garrigues (200 m + 100 m) so it is also signifcantly exposed to frebrands
from vegetation in the case of fre. Therefore, in France, WUIs are the subject of
the French Forest Orientation Law of July 9, 2002 which makes brush clearing
obligatory within a 50 m perimeter around each house located at a distance of less
than 200 m from forests or shrublands. In other European countries, the effective
fuel treatments are required only within a 10-30 m radius.
We considered WUI as two intermixed elements: the frst concerned the spatial
organization of residential houses, and the second concerned the structure of fuel
vegetation. Spatial criteria had to be developed to specify the structure of dwellings
in contact with the different vegetation structures. Concerning the structure of
dwellings, after first approaches developed with housing density calculation
(Lampin et al. 2007a,b,c), we proposed a real and quantitative defnition of terms
corresponding to isolated, scattered, dense (or very dense) clustered dwelling
types, usually used by land managers and geographers. Their distinction is based
on quantitative criteria described in Lampin-Maillet et al. (2009) such as housing
density. Concerning the structure of vegetation, only the horizontal structure of
vegetation that can be spatially recognized has been characterized (no vegetation,
sparse vegetation and continuous vegetation). Then the combination of different
types of dwellings and different classes of horizontal structures of vegetation
produced a WUI typology. The method used to characterize and to map WUIs is
based on three steps.
The frst step is to characterize and map the housing confguration. The houses
considered as located in WUIs are selected. Then according to the defnition of
the dwelling types established in Lampin-Maillet et al. (2009) and through the
process of buffering and house counting described in the same paper, each house
was classifed as belonging to one of the four confgurations of houses: isolated,
scattered, dense clustered, and very dense clustered housings. These classes take
into account the distance between houses and the density of houses located within a
100 m radius around houses.
The second step is to characterize and map the structure of vegetation. The
structure of vegetation reveals its horizontal continuity which is designed for the
measurement of aggregation levels of spatial patterns within the vegetation class
in a land-cover map. Among the different existing metrics in landscape ecology
(McGarigal 2002), the most appropriate index to measure aggregation of spatial
patterns is the aggregation index (AI) (Lampin-Maillet et al. 2010). This aggregation
index has a spatial representation. Calculated on vegetation class, the aggregation
76 Towards Integrated Fire Management
index enhances spatial organization of forests and scrublands. Vegetation is
defned as wildland forests (coniferous, deciduous, and mixed forest), scrublands,
transitional lands (mostly clear-cuts). Excluded from vegetation are low- and
high-intensity residential, commercial/industrial buildings, orchards/vineyards,
pasture/hay, arable land (e.g., row crops) and pasture, small grains, fallow, urban/
recreational grasses, bare rock/sand/clay, quarries, open water, and perennial ice/
snow (Steward et al. 2003, Radeloff et al. 2005b, Lampin et al. 2006). Aggregation
metrics calculations were made within a moving window with a radius of 20 m and
a map of aggregation index values was also drawn up including three classes of AI
values. The frst class concerned values equal to zero, and the two other classes were
determined by sharing the numbers of value equally into two groups or by setting a
threshold value equal to 95%: the frst distribution of numbers were considered as
low values of aggregation, the second one as high values.
The third step is combining the two previous criteria through a geographical
information system (GIS). The calculation allowed mapping of the WUIs according
to 12 types (Figure 2) by crossing four classes of dwelling types and three classes of
vegetation aggregation indices in a raster format. The WUI method has been applied
in different areas in France and Spain. Figure 3 is an illustration of the WUI map
carried out on French site, site 1 of Figure 1.
Figure 2. Typology of wildland urban interfaces (Lampin-Maillet et al. 2010).
Wildland Urban Interfaces, Fire Behaviour and Vulnerability: Characterization, Mapping... 77
3.3.2.2 At landscape level (Galiana et al. 2007)
The local scale of analysis must be complemented by a more general approach. On
one hand, WUI expansion processes are associated with counter-urbanization and
development of second home dynamics which respond to spatial organizational
models on an urban–regional scale (Antrop 2004). On the other hand, the existing
landscape pattern exerts an infuence on the land cover and land use trajectory, and
ultimately on the pattern of new residential development. Also, fre behaviour is
largely dependent on the landscape pattern. These are two major reasons to apply
the landscape character assessment method developed as an intermediate scale
between local and regional levels (Bürgi et al. 2004).
The WUI characterization developed results from the analysis at different scales
(regional, landscape, local), and the interactions between them, i.e. multiscale
analysis.
At the regional level, the objective of analysis is the establishment of a descriptive
model of the main territorial dynamics influencing the area (suburbanization,
abandonment and transformation of the rural areas, wildfres) and defnition of its
main spatial patterns. The results from this method improve the landscape character
assessment. The model is based on the analysis of structural (distribution of land
uses) and functional elements.
Figure 3. Map of wildland urban interfaces in study site 1 (Lampin-Maillet et al. 2010).
78 Towards Integrated Fire Management
At the landscape level, the multiscale analysis consists in the Landscape
Character Assessment (LCA), based on the natural and cultural features present
in the landscape and on evaluation of functional dynamics and uses. Landscape
description established a hierarchical typology composed of two levels: landscape
units and types (Countryside Agency and Scottish Natural Heritage 2002). The LCA
provides the territorial context of the WUI studied in order to correlate urbanization
processes with the type of landscape in which they occur and with the foreseeable
risk evolution in these areas.
At the local scale, the WUI characterization process consisted of identifying the
different morphologies defned by urbanization processes, examining them in the
context of the landscape type of the area in which they occur and characterizing
them according to fre behaviour.
Figure 4 is an illustration of the WUI map at landscape level located in site 2 of
Figure 1.
3.3.3 Fire hazard assessment in WUIs (Lampin et al. 2009)
3.3.3.1 Steps for fre hazard mapping
The work presented hereafter is aiming at the defnition of a standard workfow
for obtaining a fre hazard map that will take into account structural and dynamic
factors within a GIS. The map can be updated on a daily basis.
Table 2. WUI characterization. Relation matrix from the combination between the morphology
of the settlements and the type of landscape (Specifc case: Sierra Calderona)
Type of landscape/ Towns Urbanizations Scattered rural
Morphology of settlements
settlements
I, Western fat topped I, Compact III, Spontaneous IX, Scattered
peaks towns urbanizations (not planned) settlements on
on wildland terrain terraced slopes
III, Wildland mountain II, Compact IV, Urbanizations planned X, Scattered
with lower bunter towns with on wildland terrain settlements on
sandstones and extensions wildland terrain
cultivated gullies
IV, Agroforestal slopes V, Spontaneous XI, Scattered
- urbanizations (not planned) settlements on
V, Small agricultural VI, Urbanizations planned cultivated slopes
valleys on wildland terrain
II, Agricultural foothill VII, Spontaneous XII, Scattered rural
plains - urbanizations (not planned) settlements
VIII, Urbanizations planned

Wildland Urban Interfaces, Fire Behaviour and Vulnerability: Characterization, Mapping... 79
Calculation of a structural index (Probability of occurrence)
There are several approaches attempting to assess the combined impact of structural
factors on fre hazard. Factors that are usually taken into account usually are:
- Human presence (population, distance from settlements, distance from roads
etc.);
- Vegetation (type, biomass, structure);
- Topography (slope, elevation, aspect).
Factors are integrated into equations, each one weighted according to the author’s
opinion on its relative importance, to produce hazard indices. To overcome the
subjectivity in weighting of the contributing factors, new techniques have been
elaborated like principal component analysis (Xu et al. 2006), logistic regression
(Chou and Minnich1993, Martinez et al. 2009) and neural networks (Vasilakos et al.
2007). These techniques seek to establish a relationship between those factors and
fre history. Among those new techniques, logistic regression has been extensively
used as it provides a probability of occurrence as an output. The accuracy of the
assessment can be easily estimated by the observed fre incidents.
In a frst step, a group of potentially signifcant variables for the prediction of fre
occurrence must be defned (for example: fuel types, elevation, aspect, distance
from nearest road, average annual rainfall, etc.). These variables should be regressed
against a data set containing the perimeters of historic fre incidents.
In a second step, an exploratory analysis (e.g. Chi
2
signifcance test) should be
performed, in order to identify if any of the variables in consideration are not related
to fre occurrence. Further, weights are assigned to each of the remaining variables
Figure 4. WUI map at landscape level in Spanish study site, site 2.
80 Towards Integrated Fire Management
according to its signifcance, which is calculated by means of logistic regression.
The logistic model of fre occurrence probability can be expressed as:
PFO
i
= (1)
Where PFO
i
is the probability of a fre to occur in the ith geographic unit (pixel) and
Z
i
can be calculated as:
Z
i
= b
0
+ b
1
X
1i
+ b
2
X
2i
+ ... + b
j
X
ji
+ e (2)
Where X
ji
is the value of the jth variable in consideration in the ith geographic unit
(pixel or polygon), b
j
is the weight of the jth variable and e is an error term.
As implied by equation 1, when Zi approaches positive infnity, the value of the
index approaches 1, indicating that the ith geographic unit is defnitely going to burn
in the next period. Similarly, when Zi approaches negative infnity, the value of the
index approaches 0, indicating that the ith geographic unit is not going to burn.
Value range for the PFO: PFO, being a probability, ranging between 0 and 1.
Calculation of a daily (dynamic) index
Considering the short-term component, different strategies have been applied.
Most of them consist of trying to assess the effect of fuel water content on fuel
fammability. The Canadian Fire Weather Index (FWI) is generally considered as
one of the most effcient in defning the daily risk, as it has been globally applied
and evaluated in very diverse ecosystems besides Canada (Marzano et al. 2004,
Viegas et al. 1994). For these reasons the FWI is the dynamic index proposed also
here. A schematic presentation of the FWI structure is shown in Figure 5.
The values of the required variables for the calculation of the FWI (temperature,
relative humidity, wind and rainfall) are usually derived from meteorological
stations or sensors. In order to calculate an FWI surface (map), there is a need to
obtain a surface of values for each one of the above-mentioned variables. Thus,
the point source values of those variables have to be interpolated, before the fnal
calculation of the FWI (Mantzavelas et al. 2007).
Value ranges for the FWI: FWI is open-ended with ranges from 0 to 100+. Values
from 0 to 8 are considered as low values, values from 8 to 17 are considered as
moderate values, values from 17 to 32 are considered as high values, and values
greater than 32 are considered as extreme values (Raínha and Fernandes 2002,
Alexander 2008).
Combination of structural and daily indices; the Composite Index (CI)
As discussed above, the structural index PFO is calculated by considering all factors
(human, topography, vegetation, weather, etc.) that contribute to the occurrence of
the wildfre phenomenon. But, as implied by the calculation method, the structural
index is a long-term predictor of fre occurrence and can only be considered as
an average probability, because it does not take into account the current status of
fuel, which is thought to be a very signifcant factor in the occurrence of wildfres.
For instance, a heavy rain incident can reduce the chance of fre to zero, even in a
place where the (long-term) occurrence probability is thought to be high. On the
other hand, a meteorological index like the FWI, represents the ignition risk that
Wildland Urban Interfaces, Fire Behaviour and Vulnerability: Characterization, Mapping... 81
is contributed upon the daily fuctuations of the weather and their impact on fuel
moisture status. Although the FWI is considered as quite effcient in defning the
daily ignition risk, it is reasonable to think that: (a) all ignitions cannot be explained
by a low fuel moisture level; and (b) not all ignitions develop into fres (e.g. in the
case of a lack of fuel).
Thus, it becomes apparent, that an integrated approach is needed, in order to better
understand and defne the wildfre hazard issue. We propose, for that reason, the
calculation of a Composite Index (CI), which is the product of the Structural Index
(PFO) and the Fire Weather Index (FWI), such as:
CI
i
= PFO
i
* FWI
i
(3)
where: CI
i
is the value of the Composite Index for the ith geographic unit,
PFO
i
is the value of the Structural Index for the ith geographic unit, and
FWI
i
is the value of the FWI for the ith geographic unit.
The CI can be considered as the ‘valid’ portion of the FWI calculation. That is,
provided that the calculation of the probability of fre occurrence in PFO is correct,
then the CI is the part of the FWI value that remains meaningful under the combined
infuence of structural factors and fre history. In other words, we may say that CI is
the part of the FWI value that is validated by structural factors and fre history (if the
PFO value is 0 then the CI value becomes 0, whereas if the PFO value is 1 then the
CI value is 100% of the FWI value).
Value ranges for the CI: CI, being the product of the PFO and the FWI has the
same value range as the FWI (0 to 100+).
The Potential Fireline Intensity (PFI)
Until now, we have only dealt with the issue of fire occurrence. In order to
have a broader view on the issue of fre hazard, we have to explore the possible
Figure 5. Simplifed structure diagram for the Fire Weather Index (after Van Wagner 1987).
Fire Weather
Observations
Temperature
Relative Humidity
Wind
Rain
Temperature
Relative Humidity
Rain
Temperature
Rain
Fire Behaviour
Indexes
Fuel Moisture
Codes
Initial Spread
Index
Fire Weather
Index
Build Up
Index
Wind
Drought
Code
Duff
Moisture Code
Fine Fuel
Moisture Code
82 Towards Integrated Fire Management
consequences from a certain fre incident (Alexander 2008). A way to do so is to
calculate the potential freline intensity, which is the potential energy release per
unit length of the fre front in the case of a fre incident. The calculation of freline
intensity is signifcant in fre suppression and the study of the ecological effects of
fre. The freline intensity is calculated as a function of fuel mass, the rate of spread
and a ‘constant’ number, usually taken to be equal to 18 000 kJ/kg (Byram 1959):
PFI
i
= 18 000 * W
i
* ROS
i
(4)
Where: PFL
i
is the freline intensity for the ith geographic unit (kW/m),
W
i
is the available fuel mass in the ith geographic unit (kg/m
2
), and
ROS
i
is rate of fre spread in the ith geographic unit (m/sec).
Provided that fuel types are mapped according to the classification scheme
introduced by Anderson (1982), or another classifcation scheme resulting in the
description of fuel, mean values for Wi and ROSi are normally incorporated in the
description of fuel types.
Value ranges for the PFI: PFI values range between 0 and 100 000 kW/m.
Intensity values from 0 to 350 kW/m are considered as the low values, from 350
to 1700 kW/m are considered as moderate values, from 1700 to 3500 kW/m are
considered as high values, from 3500 to 7000 kW/m are considered as very high
values, and values greater than 7000 kW/m are considered as extreme values
(Lampin et al. 2002).
3.3.3.2 The Hazard Index (HI) building and mapping
Finally the Hazard Index (HI) can be seen as a combination between the Composite
Index and the Potential Fire Intensity. We propose the relationship defned in Table
2 for the calculation of HI.
The idea that stands behind the HI calculation is to integrate into one single index,
the issues of fre ignition, fre spread, and the potential to cause damage. The HI
embodies the different aspects of the indices that this index is calculated from. On
the one hand, the CI indicates the effects of current weather conditions and fre
history upon fre occurrence, while the PFI indicates the severity of the fre should
an ignition occur.
For example, if current weather conditions do not favor ignition or spread and
this is also validated by the fre history (‘Low’ CI value), and the potential freline
intensity is low (‘Low’ PFI value), then we should expect a low intensity fre, or
Table 3. Calculation of the Hazard Index.
CI\PFI Low Moderate High Very high Extreme
Low Low Low Moderate Moderate High
Moderate Low Moderate Moderate High Very high
High Moderate Moderate High Very high Extreme
Extreme Moderate High Very high Extreme Extreme

Wildland Urban Interfaces, Fire Behaviour and Vulnerability: Characterization, Mapping... 83
no fre at all (‘Low’ HI value). If current weather and fre history indicate that fre
occurrence is very likely (‘Extreme’ CI value), and the potential freline intensity
is ‘Extreme’, then we should expect a very intense fre that may cause extensive
damage (‘Extreme’ HI value). Intermediate classes of the HI (‘Moderate’, ‘High’
and ‘Very High’) can be explained, similarly, as the kind of fre behaviour and
intensity that is expected at a given place and time. Figure 6 illustrates the map of
this HI.
3.3.4 Vulnerability assessment in WUIs
The main objective is to develop a set of processes for mapping a synthetic index
of the WUIs vulnerability to forest fres. Vulnerability assessment comes from its
threefold consideration (a consequence, a state or characteristic, and a cause) as it
is considered in the different defnitions used above. In consequence vulnerability
is formed by an internal component, related to the effects of the fre caused on
the value of the affected goods and its capacity for recovery, and by an external
component, related to the fre characteristics and by the ability developed by society
to face the danger of wildfre.
Based on this approach, a wide variety of factors infuencing vulnerability of
the territory to forest fres have been identifed and methods to obtain parameters
Figure 6. Example of calculation of the HI in the Greek study site 3 for 14/04/2007.
84 Towards Integrated Fire Management
and its cartography at local level for the calculation of these factors have been
proposed (Figure 7). Each factor will be further subcategorized into parameters.
These parameters are the basic units that have to be mapped and standardized into a
common, simple quantitative scale in order to allow them to be aggregated through
a process based on multicriteria evaluation, where each variable is given a particular
weight. The aggregation process can be repeated until all the hierarchical structure
is aggregated into one single indicator, the vulnerability index and it will show
whether a particular area is susceptible to suffer damage from wildfre. When all the
components are calculated, the vulnerability index can be determined and will show
whether a particular area is susceptible to suffer damage from wildfre.
In the task of obtaining these parameters, the modelling of potential high risk
situations and the historical analysis of wildfres have played a very important role;
also the conclusions obtained from the tasks of WUI characterization have been
Figure 7. Flow chart illustrating the elements involved in assessing vulnerability index.
Wildland Urban Interfaces, Fire Behaviour and Vulnerability: Characterization, Mapping... 85
incorporated. Finally, the factors have been grouped into four synthetic components
(diffculty of extinction, demand for forest defense, demand of civil protection and
territorial value) which are integrated into a global index. Consultations with forest
management and fre fghting experts of the experimental area (Sierra Calderona,
Valencia, España) by applying a DELPHI methodology have been fundamental for
the determining the weights given to the factors that make up each of the indices.
The fnal index which can be mapped (Figure 8) and the intermediate components,
factors and parameters, are useful for the managers of territorial planning,
emergencies and forest fre prevention and fghting.
3.3.5 A specifc approach of fre risk assessment in WUIs through a
total risk index (Lampin-Maillet 2009)
The work presented hereafter is aiming at the defnition of a process for obtaining a
fre risk map through the calculation of a total fre risk index based on a WUI map.
The map can be updated on a basis depending on the territory WUI extension.
Considering the WUI map, a new perception of the territory is possible: WUI
area, and outside WUI area. Because of their high vulnerability, ignition probability
and combustibility, it is important and effcient to focus risk assessment in the
WUIs. The method developed allows assessment and mapping of fre risk levels.
Figure 8. Map of vulnerability levels on Spanish study site (site 2 in Figure 1).
86 Towards Integrated Fire Management
A spatial analysis on the studied territory was performed in order to establish
relationships between the distribution of fre ignition points and burned areas and
different land cover data, WUI types, environmental data. For that, a digitalized
database of fire ignition points created by the French National Forest Institute
(ONF) was used. It comprised fre ignition points during the 1997–2007 period
for which the fre area was more than one hectare. Around 565 fre ignition points
were located in the study area. A digitalized database of burned area produced by
the Administration of Agriculture and Forest of Bouches-du-Rhône was used. It
comprised 109 wildfres recorded study area during the 1990–2007 period. At last,
a thematic land cover layer obtained from the Spot Thema database elaborated in
2004 by CNES, Provence-Alpes-Côte-d’Azur Region and a Spot Image from Spot
5 satellite imagery on the study area were used. The detailed level describes the
territory (urban, agricultural and natural components) at the 1:10 000 scale.
As a result of the relationships established between WUIs and fre indicators
calculated with past fire data (fire ignition density and burned area ratio), it is
possible to identify specifc WUIs which present a high level of fre risk. Figure 9
shows that WUIs corresponding to isolated dwellings present a high level of fre risk
due to high levels of ignition density and burned area ratio. WUIs corresponding to
very dense clustered dwellings present also a high level of ignition density linked
with human activities but a low burned area ratio (high urban component and low
vegetation component) (Lampin-Maillet et al. 2010).
The spatial analysis also allowed identification of a set of conditions that
correspond with high fre risk in WUIs: housing density, road density, vegetation
more or less continuous. Results of this analysis are expressed through three main
functions (or fre risk indicators) based on statistical multiple regressions with R
2

more or less high:
- Fire Ignition Density FID = Exponential function (territory type, land cover
type, housing density) with R
2
= 51%;
- Wildfre Density WD = Exponential function (territory type, land cover type,
housing density, coniferous forests, exposure to very warm temperatures) with
R
2
= 57%;
- Burned Area Ratio BAR = Polynomial function (territory type, land cover
type, housing density, road density, country road density, garrigues, altitude,
low aggregation of vegetation) with R
2
= 36%.
A total index of wildfre risk was developed combining the three previous indicators
(Lampin-Maillet 2009). In reference to fire risk definition, each of the three
indicators includes information about hazard and/or vulnerability: Fire Ignition
Density FID and Wildfre Density WD are particularly concerned by fre occurrence
(ignition probability and wildfre probability) and Burned Area Ratio BAR is related
to hazard and vulnerability through the intensity element. Their combination can
contribute to a pertinent and effcient assessment of fre risk. So a Fire risk total
index RI has been built corresponding to a linear combination of the three indicators
having the same weight but corrected by their explanation level (R
2
value). In the
case of the study area the equation is:
Fire risk total index RI = 0.89 FID + WD + 0.63 BAR
Wildland Urban Interfaces, Fire Behaviour and Vulnerability: Characterization, Mapping... 87
Figure 9. Fire ignition density and burned area ratio according to WUI types in Lampin-
Maillet (2009).
Figure 10. Map of wildfre risk global index in WUIs in the French study site (site 1 in Figure 1).
88 Towards Integrated Fire Management
With respectively a correction of 0.89 corresponding to the ratio 51% / 57% for FID,
a value of 1 (the best R
2
value) corresponding to the ratio 57% / 57% for WD, and a
correction of 0.63 corresponding to the ratio 36% / 57% for BAR.
A map of this fre risk total index can be produced as illustrated at Figure 10 in the
South of France.
Certain types of WUIs represent a high level of fre risk in terms of fre ignition
density, wildfire density and burned area ratio. Regarding fire ignition density
and burned area ratio, isolated WUIs with low and high aggregation indices of
vegetation presented the highest values. Scattered WUIs with both low and high
aggregation indices of vegetation also represented a high level of fire ignition
density and burned area ratio even if these values were lower than those for isolated
WUIs (Lampin-Maillet et al. 2010). Results also highlighted the fact that the burned
area ratio generally decreased from isolated WUIs to dense and very dense clustered
WUIs, and also decreased from a high aggregation index to a zero aggregation index
(Lampin-Maillet et al. 2010).
3.3.6 Conclusions
In this chapter we proposed reproducible methods for characterizing and mapping
WUIs at large scales and over large areas (Lampin-Maillet et al. 2010) and at
landscape scale (Galiana et al. 2007). We also proposed on the one hand a method
for fre hazard assessment and mapping (Lampin et al. 2009), and on the other
hand a method for vulnerability assessment and mapping (Galiana et al. 2009).
These two processes contribute to assessment of fre risk, combining fre hazard and
vulnerability assessments. A specifc approach has also been developed to calculate
and to assess a total index of wildfre risk in WUIs (Lampin-Maillet 2009).
The WUI is often a location which is particularly appreciated as an area in which
to live. However, in our European Mediterranean context this lifestyle carries a
certain risk: people should always be aware of the existing fre risk in such WUI
and should respect and apply effcient recommendations to insure against risky
situations.
These results could have interesting implications for fre prevention and land
management. Introducing the risk of wildfre with such maps, and particularly the
vulnerability of the territory, is a way to make the inhabitants becoming aware of
fre risk in WUIs. The WUI map is key information to identify locations where
vegetation has to be reduced in order to protect the houses and their inhabitants in
case of wildfre and where careful behaviour are essential to avoid fre ignition. This
will globally decrease the risk of fre either by reducing fre propagation through
biomass removal and/or by reducing fre ignition probability together with less
carelessness. Accomplishing this goal is strictly related to the designation of suitable
prevention messages and preventive actions which can be different according to
WUI types.
WUIs have increased considerably all over the world in recent decades and this
trend will certainly continue in the coming years due to the continuing trend of land
abandonment combined with urbanization. The method we developed for mapping
Wildland Urban Interfaces, Fire Behaviour and Vulnerability: Characterization, Mapping... 89
WUIs is an appropriate tool for the assessment of WUI dynamics and associated
fre risk dynamics in the context of ongoing changes in climate, urbanization and
vegetation.
References
Alexander, M.E. 2008. Proposed revision of fre danger class criteria for forest and rural
areas in New Zealand. 2
nd
Ed. National Rural Fire Authority, Wellington, in association
with Scion, Rural Fire Research Group, Christchurch. 62 p.
Anderson, H.E. 1982. Aids to determining fuel models for estimating fre behaviour. USDA
Forest Service General Technical Report INT-122, Intermountain Research Station,
Ogden, UT. 22 p.
Antrop, M. 2004. Landscape change and the urbanization process in Europe. Landscape and
urban planning 67: 9–26.
Alavalapati, J.R.R., Carter, D.R. and Newman, D.H. 2005. Wildland-urban interface:
Challenges and Opportunities. Forest policy and Economics 7: 705–708.
Byram, G.M. 1959. Combustion of forest fuels. In: Davis K.P. (ed). Forest Fire: Control and
Use. McGraw-Hill, New York.
Blaikie, P., Cannon, T., Davis, I. and Wisner, B. 1994. At risk: natural hazards, people
vulnerability and disaster. Routledge, London, New York. 280 p.
Blanchi, R., Jappiot, M. and Alexandrian, D. 2002. Forest fre risk assessment and
cartography. A methodological approach. In: Viegas, D. (ed.). Proceedings of the IV
International Conference on Forest Fire Research. Luso, Portugal.
Brunet, R., Ferras, R. and Théry, H. 1993 (3
rd
Ed.). Les mots de la géographie. Paris,
Montpellier, La documentation française. 518 p.
Bürgi, M., Hersperger, A.M. and Schneeberger, N. 2004. Driving forces of landscape change
– current and new directions. Landscape Ecology 19: 857–868.
Caballero, D., Giroud, F., Picard, C. and Xanthopoulos, G. 2004. EUFIRELAB/ Euro-
Mediterranean Wildland Fire Laboratory, a "wall-less" laboratory for wildland fre
Sciences and technologies in the Euro-Mediterranean Region - Deliverable D-10-01-
Wildland-Urban Interface Management: A state of the art. 28 p.
Camia, A., Guglielmet, E., Bovio, G. Deshayes, M., Lampin, C., Salas, J., Martinez, J.,
Molina, D., Carrega, P., Fox, D., Berolo, W., Sabate, S., Vayreda, J., Martin, P., Martinez,
J., Conese, C., Bonova, L., Karteris, M., Mallinis, G., Sergiopoulos, I. and Gitas, I. 2004.
Eufrelab: Euro-Mediterranean Wildland Fire Laboratory, a "wall-less" laboratory for
wildland fre Sciences and technologies in the Euro-Mediterranean Region Deliverable D
08-02: Wildland fre danger and hazards: a state of the art. 57 p.
Carroue, L., Claval, P., Di Meo, G., Miossec, A., Renard, J.P., Simon, L. and Veyret, Y.
2002. Limites et discontinuité en géographie. Dossiers des images économiques du
monde, Edition SEDES. 159 p.
Chen, K., Blond, R. and Jacobson, C. 2003. Towards an integrated Approach to natural
Hazards risk assessment using GIS: with reference to bushfres. Environmental
Management 31 (4): 546–560.
Chou, Y.H. and Minnich, R.A. 1993. Mapping Probability of Fire Occurrence in San Jacinto
Mountains, California, USA. Environmental Management 17(1): 129–140.
Coburn, A.W., Spence, R.J.S. and Pmomonis, A. 1994. Vulnerability and risk assessment.
Disaster management training Program, UNDP/DHA, 70 p.
90 Towards Integrated Fire Management
Colin, P.Y., Lampin, C., Delboube, E., Coste, N., Marcillat, J., Pereira, J.C., Binggeli, F.,
Gaulier, A. ,Botelho, H., Loureiro, C., Loddo, G., Ditana, E., Guijarro, M., Hernando,
C., Diez, C., Martinez, E., Madrigal, J., Vega, J.A., Gorostiaga, P., Alexandrian, D. and
Dimitrakopoulos, A. 2002. SALTUS - Spot fres : mechanisms analysis and modelling -
Final report 1998, ENV98-CT98-0701.
Countryside Agency and Scottish Natural Heritage, 2002. Landscape Character Assessment:
guidance for England and Scotland (Wetherby, Countryside Agency and Scottish Natural
Heritage).
Davis, J. B. 1990. The wildland-urban interface: paradise or battleground? Journal of
Forestry 88(1): 26–31.
D’Ercole, R. 1998. Approches de la vulnérabilité et perspectives pour une meilleure logique
de réduction du risque. Pangea 29(39): 20-28.
D’Ercole, R. 1996. Représentation cartographiques des facteurs de vulnérabilité des
populations exposées à une menace volcanique. Application à la région du volcan
Cotopaxi (Equateur). Bulletin de l’IFEA 25(3): 479–507.
Galiana, L., Herrero, G. and Solana, J. 2007. Caracterización y clasifcación de Interfaces
Urbano-Forestales mediante análisis paisajístico. El ejemplo de Sierra Calderona
(Comunidad Valenciana, España). Book of Abstracts of the IV International Wildland Fire
Conference, Seville, Spain. Organismo Autónomo de Parques Nacionales, Ministerio de
Medio Ambiente. 285 p.
Gallopin, G.C. 2006. Linkages between vulnerabilty, resilience and adaptative capacity.
Global Environmental Change 16: 293–303.
Hardy, C.C. 2005. Wildland fre hazard and risk: Problems, defnitions, and context. Forest
Ecology and Management 211: 73–82.
Jappiot, M., Gonzalez-Olabarria, J.R., Lampin-Maillet, C. and Borgniet, L. 2009. Assessing
Wildfre Risk in Time and Space. Partie de Living with Wildfres: What science can tell
us. Yves Birot (ed.) EFI Discussion Paper 15. European Forest Institute. Joensuu. Pp.
41–47.
Johnson, M. P. 2001. Environmental impacts of urban sprawl: a survey of the literature and
proposed research agenda. Environment and Planning A 33: 717–735.
Lampin-Maillet, C. 2009. Caractérisation de la relation entre organisation spatiale d’un
territoire et risque d’incendie : le cas des interfaces habitat-forêt du sud de la France.
Thèse de doctorat de l’Université de Provence, Aix-Marseille I. 321 p.
Lampin-Maillet, C., Jappiot, M., Long, M., Morge, D. and Ferrier, J.P. 2010. Mapping
wildland-urban interfaces at large scales integrating housing density and vegetation
aggregation for fre prevention in the South of France. Journal of Environmental
Management 91 (2010):732–741.
Lampin-Maillet, C., Jappiot, M., Long, M., Morge, D. and Ferrier, J.P. 2009.
Characterization and mapping of dwelling types for forest fre prevention. Computers,
Environment and urban systems 33(2009): 224–232.
Lampin-Maillet, C. 2008. 2007, Summer Fires in the European Mediterranean – The Cases of
Greece, Italy and Spain. Mediterranean yearbook. European Institute of the Mediterranean.
Med. 2008, Economy and Territory- Sustainable Development. Pp 243–247.
Lampin, C., Jappiot, M., Long M., Morge, D., Bouillon, C., Galiana, L., Herrero, G., Solana,
J., Mantzavilas, A., Loddo, G. and Delogu G. 2007a. Characterization and mapping of
wildland-urban interfaces: a methodology applied in the study case in Sardinia. Book of
Abstracts of the IV International Wildland Fire Conference, Seville, Spain. Organismo
Autónomo de Parques Nacionales, Ministerio de Medio Ambiente. 289 p.
Lampin, C., Long, M., Jappiot, M., Morge, D., Bouillon, C., Mantzavellas, A.,
Apostolopoulou, I., Lazaridou, T., Partozis, T., Topaloudis, T., Galiana, L., Herrero,
G., Solana, J., Loddo, G., Brigaglia, S., Dettori, G. and Delogu, G. 2007b. Method to
characterize and to map wildland urban interfaces. Deliverable D5.2-1 of the Integrated
Project “Fire Paradox”, Project no. FP6-018505. European Commission. 104 p.
Wildland Urban Interfaces, Fire Behaviour and Vulnerability: Characterization, Mapping... 91
Lampin, C., Long, M., Jappiot, M., Morge, D., Bouillon, C., Mantzavelas, A.,
Apostolopoulou, I., Lazaridou, T., Partozis, T., Topaloudis, T., Galiana, L., Herrero, G.
and Solana, J.L. 2007c. Guide for end-users to characterize and to map wildland-urban
interfaces. Deliverable D5.2-2 of the Integrated Project “Fire Paradox”, Project no. FP6-
018505, European Commission. 18 p.
Lampin, C., Jappiot, M., Long, M. and Borgniet, L. 2006. Cartographie des interfaces
habitat-forêt. Une approche spatiale pour estimer le risque d’incendie de forêt, Revue
internationale de Géomatique. European journal of GIS and Spatial analysis, Information
géographique et gestion des risques l16(3–4): 320–340.
Lampin, C., Jappiot, M., Alibert, N., Manlay, R. and Guillande, R. 2002. Prototype of an
Intensity Scale for the Natural Hazards: Forest Fire. In Viegas, D. (ed.). Proceedings of
the IV International Conference on Forest Fire Research. Luso, Portugal.
Macie, E.A. and Hermansen, L.A (eds). 2002. Human infuences on Forest Ecosystems: The
Southern Wildland-Urban lnterface Assessment.
McGarigal, K. 2002. Landscape Pattern Metrics. Chapitre du livre Encyclopedia of
Environmentrics, Volume 2. John Wiley & Sons, Sussex, England.
Lampin, C., Mantzavelas, A., Apostolopoulou, I., Lazaridou, T., Partozis, T. and Topaloudis,
T. 2009. Adapted method to assess and map fre hazard in wildland urban interfaces.
Internal Report IR5.1-4 of the Integrated Project “Fire Paradox”, Project no. FP6-018505.
European Commission. 19 p.
Mantzavelas, A., Apostolopoulou, I., Lazaridou, T., Partozis, T., Lampin, C., Long, M.
and Jappiot, M. 2008. Literature review and synopsis of the methodologies to estimate
vulnerability. Deliverable D5.2-5 of the Integrated Project “Fire Paradox”, Project no.
FP6-018505. European Commission. 13p.
Mantzavelas, A., Apostolopoulou, I., Lazaridou, T., Partozis, T., Topaloudis, T., Lampin,
C., Borgniet, L., Bouillon, C., Brewer, S., Curt, T., Ganteaume, A., Jappiot, M., Defossé,
G., Gómez Fernán, M. and Lencinas Daniel, J. 2009. Method to assess with good spatial
accuracy the meteorological and fuel moisture components of the fre risk. Deliverable
D5.1-1 of the Integrated Project “Fire Paradox”, Project no. FP6-018505. European
Commission. 89 p.
Martinez, J., Vega-Garcia, C. and Chuvieco, E. 2009. Human-caused wildfre risk rating for
prevention planning in Spain. Journal of Environmental Management 90(2): 1241–1252.
Marzano, R., Bovio, G., Guglielmet, E., Jappiot, M., Lampin, C., Dauriac, F., Deshayes, M.,
Salas, J., Aguado, I., Martinez, J., Molina, D., Martinez, E., Carrega, P., Fox, D., Berolo,
W., Martin, P., Gomez, I., Conese, C., Bonora, L., Karteris, M., Gitas, I., Mallinis, G.
and Giannakopoulos, V. 2004. Common methods for mapping the wildland fre danger,
EUFIRELAB, Deliverable D-08-05.
McGee, T. K. 2005. Completion of recommended WUI fre mitigation measures within
urban households in Edmonton, Canada. Environmental Hazards 6: 147–157.
Radeloff, V.C., Hammer R.B., Stewart, S.I., Fried, J.F., Holcomb, S.S. and McKeefry, J.F.
2005a. The wildland urban interface in the United States. Ecological applications 15(3):
799–805.
Radeloff, V.C., Hammer R.B. and Stewart, S.I. 2005b. Rural and suburban sprawl in the
U.S. Midwest from 1940 to 2000 and its relations to forest fragmentation. Conserv. Biol.
19(3): 793–805.
Rainha, M. and Fernandes, P.M. 2002. Using the Canadian Fire Weather Index (FWI) in the
Natural Park of Montesinho, NE Portugal: calibration and application to fre management.
In Viegas, D. (ed.). Proceedings of the IV International Conference on Forest Fire
Research. Luso, Portugal.
San Miguel-Ayanz, J., Carlson, J.D., Alexander, M., Tolhurst, K., Morgan, G., Sneeuwjagt,
R. and Dudley, M. 2003. Current Methods to Assess Fire Hazard Potential. In: Chuvieco,
E. (ed.). Wildland Fire Hazard Estimation and Mapping. The Role of Remote Sensing
Data. World Scientifc Publishing, Singapore. Pp. 21–61.
92 Towards Integrated Fire Management
Stewart, S.I., Radeloff, V.C. and Hammer, R.B. 2003. Characteristics and location of the
Wildland-Urban-Interface in the United States. p. 6. Proceedings of the 2
nd
International
Wildland Fire Ecology and Fire Management Workshop, November 16–20, 2003,
Orlando, Florida.
USDA-USDI. 2001. Urban wildland interface communities within vicinity of Federal lands
that are at high risk from wildfre. U.S. Department of Agriculture and U.S. Department of
Interior, Fed.Reg. 66: 751–777.
Van Wagner, C. E. 1987. Development and Structure of the Canadian Forest Fire Weather
Index System. Can. For. Serv., Ottawa. For. Tech. Rep. 35. 37 p.
Vasilakos, C., Kalabokidis, K., Hatzopoulos, J., Kallos, G. and Matsinos, Y. 2007.
Integrating new methods and tools in fre danger rating. International Journal of Wildland
Fire 16: 306–316.
Viegas, D.X., Sol, B., Bovio, G., Nosenzo, A. and Ferreira, A. 1994. Comparative study
of various methods of fre danger evaluation in Southern Europe. In: Viegas, D. (ed.).
Proceedings of II International Conference on Forest fre Research. Coimbra, Portugal.
Pp. 571–590.
Vince, S.W., Duryea, M.L., Macie, E.A. and Hermansen, L.A. (eds.) 2005. Forests at the
wildland-urban interface: conservation and management. CRC Press, Boca Raton.
Wilson, K., Pressey, R.L., Newton, A., Burgman, M., Possingham, H. and Weston, C. 2005.
Measuring and incorporating vulnerability into conservation planning. Environmental
Management 35(5): 527–543.
Xu, D., Guofan, S., Dai, L., Hao, Z., Tang, L. and Wang, H. 2006. Mapping forest fres risk
zones with spatial data and principal components analysis. Science in China: Series E
Technological Sciences 49(I): 140–149.
3.4 The Importance of Economics in Fire
Management Programmes Analysis
Robert Mavsar
1
, Armando González Cabán
2
and Verónica Farreras
1

1
Forest Technology Centre of Catalonia, Barcelona, Spain
2
USDA Forest Service, Pacifc Southwest Research Station, Riverside (CA), USA
3.4.1 Introduction
Wildfres are a societal problem that threatens many ecosystems, affects millions
of people worldwide, and causes major ecosystem and economic impacts at local
regional, national and global scales. In Europe, and especially in the Mediterranean
countries (France, Greece, Italy, Portugal and Spain), wildfres continue to be a
major environmental threat (Requardt et al. 2007) where, an average 500 000 ha
of forests are burned annually (San-Miguel and Camia 2009). Wildfres affect the
forests and other wooded land, and neighbouring systems such as urban areas,
infrastructure networks (i.e. power-lines and transportation corridors), agriculture
lands, and the civil society. These impacts can be refected in many ways – for
example, loss of human life or health, decreased well-being of the population (local
and wider), and temporary or permanent loss of employment possibilities and
economic activities. For a worldwide perspective on the effects of fre on the earth
systems see Bowman et al. 2009.
The last decades have seen increased social awareness and growing concerns for
wildfres’ negative environmental and economic consequences, and, particularly, the
loss of human lives (McCaffrey 2008). The growing importance of wildfre issues
at EU level is also refected in the increasing number of research projects funded to
better understand and address this problem (e.g. Eufrelab, Fire Paradox).
In Europe, this change was galvanized by the large-scale wildfires and their
consequences in the Mediterranean region during the 2000s. For example,
during 2003 in Portugal about 400 000 ha of forests lands were burned; during
2005 in Spain some 190 000 ha of forests lands were damaged; and during 2007
in Greece around 270 000 ha of forests and other wooded lands were destroyed
(JRC 2007). Events in other parts of the world also helped to increase the concern
on this topic in Europe. These included, Indonesia 1997/1998, where 8 million ha
burned; Australia 2007, more than 1 million ha; and California, USA 2007, 120
000 ha affected. These major wildfre events clearly showed that they are not only
an environmental problem, but have also a signifcant social dimension, affecting
millions of people, having major economic impacts, and causing signifcant human
casualties (González-Cabán 2007). For example, the wildfres affecting vast forest
areas of Portugal in 2005 caused economic damage worth almost €800 million and
caused 13 fatalities. Even worse, the large fres affecting Greece during the summer
of 2007 caused 64 casualties, and according to the Greek authorities the economic
94 Towards Integrated Fire Management
damage was estimated at €2–5 billion (Papachristou 2007; Petsini Arlapanou and
Petsini Arlapanou 2007).
These events increase the public interest and concern about the wildfires
and prompted the development and implementation of improved policies and
management measures at different levels. The main objective of these improved
policies and management measures is to minimize the negative environmental,
economic and social impacts of wildfres (EU 2005). However, the implementation
of such measures requires substantial investment of financial, human and
organizational resources, which must be justifable and effcient.
Decision support systems based on economic models can help to decide what
would be the optimal use of the resources. For example, economic models can
help to estimate whether investments in wildfre related measures (e.g. prevention,
suppression, fuel management) are financially justified, or to choose the most
effcient amongst several alternatives (i.e. the combination of investments in fre
prevention, fre fghting and amount of wildfre acceptable). For the implementation
of these decision support systems reliable knowledge and data are needed on
the physical and economic impacts (negative and positive) of wildfires and
the economic efficiency of fire management measures. The understanding and
assessment of socio-economic impacts of wildfres should be considered an essential
part of the fre risk assessment, development of wildfre related policies, as well as
planning and implementation of management practices (Morton et al. 2003).
This chapter will frst introduce the basic model used for the estimation of the
economic efficiency of fire management programmes and revise some of the
problems related with its application. In the subsequent section we discuss why
the social preferences should be considered when planning fire management
programmes and present a study dealing with this topic that was conducted in Spain.
In section three of this chapter, we look in detail at the problems and a possible
solution for the adequate estimation of costs related to wildfres. The last part of the
chapter presents the main conclusions.
3.4.2 Estimating the economic effciency of and social preferences for
fre management programmes
3.4.2.1 Economic effciency of fre management programmes
To combat the wildfres problem forest managers can apply different management
measures. These include prevention (e.g. education, publicity campaigns), fuel
treatment (e.g. prescribed burning, thinning, mechanical fuel removal), pre-
suppression and suppression, and restoration measures. One of the thorniest
questions in fre management is to determine how the limited fnancial, equipment,
and human resources should be most efficiently spent and distributed among
alternative fre management options.
To answer this question, the economic analysis of the efficiency of fire
management is generally evaluated by the cost plus net value change concept
(C+NVC) (Althaus and Mills 1982; Donovan and Rideout 2003; González-Cabán
The Importance of Economics in Fire Management Programmes Analysis 95
2007). This approach estimates all costs and benefts related to fre management
measures. The model is derived by adding pre-suppression costs to the costs of fre
suppression and the net value of the change in resources outputs that result from the
fres (difference between benefts and damage to the resource). The pre-suppression
costs represent the expenditure on fre management prior to the wildfre season (e.g.
purchase of fre fghting resources such as fre engines, equipment for fre fghting
crews). Suppression costs are fre fghting expenditures during a fre season (e.g.
wages for fre fghting crews, fuel for aerial resources).
In the C+NVC model pre-suppression and suppression expenditures are considered
as independent inputs, related through the Net Value Change (NVC) function
(Donovan and Rideout 2003). The independence of pre-suppression and suppression
expenditures means that one input does not determine the level of the other. For
example, the purchase of some fre fghting equipment before the fre season, does
not determine how frequently this equipment will be used during the fre season.
Nevertheless, the pre-suppression may affect the optimal level of suppression
through the NVC function (Figure 1). In Figure 1 the slope of the level curve is
equal to the negative of the ratio of the marginal contributions of suppression and
pre-suppression in reducing damage. It should be noted, that pre-suppression can be
substituted by suppression, but only in the case that NVC is held constant (Donovan
and Rideout 2003). Thus, both inputs are allowed to vary independently, but remain
related through the NVC function (Donovan and Rideout 2003).
As mentioned at the beginning of this chapter, the objective of the C+NVC
model is to fnd the most effcient management levels (González-Cabán et al. 1986;
Lankoande 2005). Selecting the most effcient management level is an optimization
problem, where we intend either to minimize the costs and the net value change, or
to maximize the social benefts (Donovan and Rideout 2003; Mercer et al. 2007).
Figure 1. NVC level curve.
Suppression
P
r
e
-
s
u
p
p
r
e
s
s
i
o
n

Net Value Change
96 Towards Integrated Fire Management
For example, given that the pre-suppression expenses are at an optimal level (Figure
2), the most effcient level of fre programme is the point where the summation of
costs and the net value change is minimized (i.e. P* in Figure 2). In the model, the
cost part (C) is obtained by the estimation of the programme costs (pre-suppression
and suppression activities); while the net value change (NVC) sums all the changes
in the quantity/quality of resources outputs (goods and services) that result from the
fre multiplied by the unit value of the output.
The objective of the evaluation of economic efficiency of fire management
measures is not only to help identify the most effcient programme level, but also to
decide how to distribute the resources (spatial component). Namely, once agencies
have decided how many resources to use (e.g. fre detection facilities, suppression
resources, fuel treatments), their most effcient and cost-effective allocation must be
determined (Wei et al. 2008).
To be able to apply the C+NVC model and to include it into a decision
support system (DSS) for planning and evaluation of the economic effciency of
management programmes it requires adequate information. For example, for each
evaluated fre management program alternative it would at least need information
on: monetary estimates of economic effciency, quantitative estimates of effects on
resource outputs, and assessments of the risk associated with these values. However,
there are still considerable shortcomings in the knowledge and data availability.
One of the diffculties related to the application of the C+NVC model is the
estimation of the NVC component. It requires information about the direct and
indirect effects of fre on spatial and temporal provision of goods and services, and
information about how fre-induced marginal changes in the quality and quantity
of goods and services will affect social welfare. It is apparent that information is
Figure 2. Illustration of the C+NVC model (Donovan and Rideout 2003).
Suppression costs
Suppression cost
Net value change
Pre-suppression cost
C¹NVC
C
o
s
t
s

P*
The Importance of Economics in Fire Management Programmes Analysis 97
lacking in one or both of these areas for many of the types of goods and services
that can be affected by wildfre (Venn and Calkin 2007). More on this issue will be
presented in section 3.4.3 of this chapter.
The estimation of the costs and the NVC is further complicated by the possible
temporal interrelation of fre events or management measures; meaning that present
actions can infuence decisions on applications of fre management measures in
future fre seasons. For example, Mercer et al. (2007) showed that fuel management
treatments applied in the current season can produce benefts (e.g. reduced fre
risk) also in subsequent seasons. These, ‘extended’, benefts should also be taken
into account when estimating the costs and benefts of fre management measures.
Lagged benefts are not only a characteristic of management measures, but also of
wildfres. Namely, wildfres signifcantly reduce the amount of fuels (similar to the
effect of fuel management); thus, they notably affect the fre risk and intensity in the
following seasons. According to Mercer et al. (2007) this reduced risk of wildfres
can last up to 11 years.
Another important issue to consider when estimating costs and benefts of fre
management is that multiple objectives can be accomplished by these measures. For
example, prescribed burning can reduce fuels, dispose of logging debris, prepare
sites for seeding and planting, improve wildlife habitat, manage invasive species,
control insects and diseases, improve accessibility, enhance aesthetics and recreation
activities, improve forage and grazing, and manage endangered species (Wade and
Lunsford 1988).
At the European level, another important problem, and maybe the most
fundamental, is the general lack of reliable data not only on the effects of fres on
the quantity/quality of goods and services produced by the natural resources (in our
case, forest lands), but the social value of different goods and services, as well as the
costs and expenses of fre management and suppression activities. In many cases,
there is no common methodology to collect and report such data, and therefore the
available data is most often incomplete, unreliable, and available only for some
countries or regions within countries. (Mavsar et al. 2007).
3.4.2.2 Social preferences for fre management programmes
Economic effciency is an important criterion in deciding which fre management
programme would be the most effcient. However, it should not be the only attribute
considered when deciding about fre policies or management measures. In addition,
it is important to consider other factors, for example social values and preferences
for a certain programme or management measure (Daniel et al. 2007; Martin et
al. 2008). Knowing the social preferences for different management actions and
identifying when they might differ from the manager’s viewpoint can help agencies
to understand and predict how different audiences may react to these decisions. For
instance, this knowledge may (i) help the administration recognize when policies
or actions may be supported or opposed by the public
1
, and (ii) help to develop
1 For an interesting discussion on how public perceptions affect the implmentation of fre management programmes
see Laband et al. (2006) and Laband et al. (2008).
98 Towards Integrated Fire Management
information or education campaigns, which might help to get public support for a
certain policy or action. Thus, learning about the social preferences could be useful
in different ways. It can serve as a support tool in the design of policies and/or
specifc fre management programmes; or, if faced with a fxed budget for wildfre
mitigation, land managers may want to design a fre prevention measure that mirrors
society’s preferences – for example, for fre behaviour with a lower impact on forest
resources. Therefore, social preferences may allow policy makers to better identify
priority-attention areas in fre management. This is seldom the case in Europe and
there are only a couple of case studies where social preferences for fre management
measures have been elicited (see Box 1).
3.4.3 Estimating the costs of wildfres
A precondition for the estimation of the socio-economic damages caused by
wildfres is to understand what the important factors are and how they infuence
the quantity and quality of goods and services provided by a natural resource (e.g.
forest). This knowledge is vital for the identifcation of goods and services that
can be or are damaged by a fre. Only a complete picture of the damages enables a
reliable estimation of the socio-economic impacts of wildfres. Further, it should be
acknowledged that wildfres can have positive effects (González-Cabán 2007). For
example, improving the wildlife habitat, improving understory forage and grazing,
managing endangered species and fire dependent species (Wade and Lunsford
1988). These positive effects (benefts), when existing, must also be included into
the estimation methodology.
A wide range of costs can be related to a wildfre, and there are different ways to
categorize them (Ashe et al. 2008; Dale 2009; Zybach et al. 2009). In general, we can
divide costs into direct and indirect. Direct costs are those which are directly related
to a wildfre event, and incurred as a result of the fre and/or exposure to the fre,
such as losses or damages of environmental (forest) goods and services, property,
and direct suppression costs. In contrast, indirect costs are related to the risk of
wildfre occurrence or as a response to the occurrence of a wildfre, like prevention,
monitoring and pre-suppression costs, restoration costs, and other costs associated to
the individuals who suffer a loss of benefts as a consequence of a wildfre.
An important concern in evaluation of fre management effciency and estimation
of wildfre costs is the proper consideration of environmental (forest) goods and
services that might have been damaged (or enhanced) as a consequence of a
wildfre. In estimating the economic impacts of wildfres, only a few of them have
traditionally been considered, mainly the decreased quantity/quality of timber.
However, in the last decades it has become obvious that forests are also important
as providers of environmental and social goods and services (Farrell et al. 2000).
Thus, nowadays estimating the costs of wildfres must also include these goods and
services (Butry et al. 2001; Dunn et al. 2003). However, it is important to understand
that in contrast to forest goods and services that have a market price (e.g. timber,
fruits, and mushrooms) that refects their value, there are goods and services (e.g.
biodiversity, recreation activities, erosion prevention, water purifcation) that are not
The Importance of Economics in Fire Management Programmes Analysis 99
traded in traditional markets (hereafter non-market goods and services) and thus, we
have no information about their value (González-Cabán 1998; Mavsar et al. 2008).
Non-market valuation methods can be applied to value these goods and services.
Even though these methods have improved considerably in the last decades, they still
have limitations, which should be considered when used. One of the most important
limitations is that the estimated values cannot be easily extrapolated. For example,
recreation values in different forests are probably different and this difference is
not necessarily only the result of different site characteristics, but might also be
infuenced by other factors (e.g. population size, income in the region, accessibility).
This limitation is especially important when developing a system for the assessment
Box 1: Case study on social preferences regarding fre management
measures
2

Burned forest area or dead trees? A choice for Catalan citizens
The objective of the study, conducted in Catalonia (NE Spain), was to elicit the social
preferences regarding fre prevention measures in terms of their impact on fre behaviour
– fre propagation and intensity – and to estimate the value of these measures for the
society. We applied the choice experiment methodology. The label ‘choice experiment’
refers to a survey-based valuation method that simulates actual market behaviour
(Hanemann and Kanninen 1999; Bennett and Blamey 2001). This technique is based
on the idea that any alternative, or good, can be described in terms of its attributes, or
characteristics. In a choice experiment, respondents are presented with a series of choice
sets comprising at least two alternatives and are asked to choose which alternative they
prefer (Hanley et al. 2001; Bateman 2002).
In our empirical application, 207 interviews were conducted in three Catalan provinces
– Barcelona, Gerona and Lérida – in June 2007. The frst part of the interview presented
the attributes to be valued, as well as the payment mechanism and the consequences of
each choice. The second part contained the choice experiment exercise and a number of
debriefng questions. The fnal part was designed to collect some socio-economic data
about the respondents.
The study results suggest that additional fre prevention measures increase the welfare
of the Catalan population (similar results were reported in Riera and Mogas (2004).
Further, results show that the attribute of highest concern is fire propagation (area
burned). A plausible explanation for this may be the publics’ familiarity with information
on the quantity of area burned, as this is often used by the media in Spain to quantify
the consequences and the severity of a wildfre. The result implies that from a social
perspective reducing the forest area burned is one of the most relevant elements when
designing fire prevention and protection programmes. However, the design of fire
prevention and protection programmes also depends on other factors such as vegetation
type and characteristics, complexity of the fire problem, available funds, available
experience and expertise. Thus, social preferences should be considered in the decision
making process, but are not the only factor infuencing the design of fre management
programmes.
2 For detailed information see Mavsar and Farreras (2008).
100 Towards Integrated Fire Management
of socio-economic impacts of wildfres, because the fre may spread over areas where
no values for non-market goods and services have been estimated. In this case,
simply applying values from another site may generate signifcant estimation errors.
Nevertheless, with the use of the right data, adequate techniques (e.g. beneft transfer
method) or employing proxies (e.g. restoration costs) this problem could potentially
be overcome (for example see Rideout et al. 2008).
Nevertheless, as mentioned earlier, in most cases only fre suppression costs and
loss of timber production are considered, while damage to forest non-market goods
and services are often omitted or very limited (Pettenella et al. 2009). The absence
of non-market values is somehow expected, given the complexity and high cost in
measuring them; in addition, the obtained values are site specifc and transferable
to other sites only under some specific conditions (e.g. similar forest site and
population characteristics).
3.4.4 Conclusions
The extreme wildfre events of the past decade have shown that fres are not only an
ecological, but even more a socio-economic problem. Economics has an important
role in helping to quantify the magnitude of the problem (i.e. assessing the costs of
wildfres) and to fnd adequate solutions (i.e. evaluate effciency of fre management
measures and provide information on public preferences). This also implies that the
role of economics in the fre management process should change. Therefore, it is
important for decision makers to start considering economic analysis as an integral
part of a proactive fre management; a tool that provides information that can help
managers make better informed decisions that can lead to selection of the most
effcient alternatives in a given situation.
Although there has been considerable development of economic methods (e.g.
non-market valuation methods) and models (e.g. C+NVC model), important
problems still remain to be solved in the future. In this respect, the main issues are
the inadequate understanding of:
• the impacts of wildfres on the spatial and temporal provision of goods and
services (e.g. how the quality and quantity of a good or service is affected and
for how long);
• the potential effect of the changes caused by wildfres on society’s wellbeing
(e.g. what is the value of the losses);
• the impact of fre management measures on risk, extent and severity of
wildfres (e.g. quantify the effects of different management measures).
Thus, development of a decision support system (DSS) requires, frst, answers to
these three concerns. Once this is accomplished, the next step is to develop models
that would adequately simulate the behaviour and impact of wildfres, the effects
of fre management measures, and the potential economic impacts (positive and
negative). Furthermore, development and implementation of a DSS necessitates
that a significant amount of information be collected and processed (e.g. fuel
models, historical fre data, weather parameters, geographic data, availability of fre
The Importance of Economics in Fire Management Programmes Analysis 101
management practices and resources, values of market and non-market goods and
services). In addition, standardized procedures must be established on how to collect
and process the needed data.
Finally, it is important to recognize that no decision support system or computer
simulation model is a substitute for a decision maker. A decision maker must still
weigh and assimilate a large number of relevant factors, many of which cannot be
measured in common units or even be measured quantitatively at all, and must place
fre programs within the context of other management programs and institutional
constraints (Mills and Bratten 1982). A DSS simply quantifes some of the factors,
which are relevant to fire management program planning and helps trace the
interaction of relationships that are too complex and numerous for a person to easily
follow. In the case of a fre economic effciency DSS, this would provide the decision
maker only with information on how to possibly design the most economically
effcient fre management programme. However, the results depend on the model
assumptions and restrictions (e.g. fre management policies). A DSS could show
some of the costs of imposing institutional constraints on the programme; and while
many of those constraints are valid, their costs must always be considered (Mills
and Bratten 1982). After careful consideration of all the information provided by the
DSS, and other sources, the fnal decision on what kind, and to what extent the fre
programme will be implemented rest squarely on the decision maker’s shoulders.
There are still many obstacles to overcome on the path to the development and
application of a model for the evaluation of economic effciency of fre management
and protection programs at the European level. One of the main problems is the lack
of reliable and comparable data on fre effects, the social value of affected goods and
services, and the costs of fre management activities. Considering these limitations,
there is a need not only to develop a relevant DSS, with the potential to deal with
territorial differences, but furthermore to ensure that adequate data is collected in a
standardized form. Hopefully, this will change for the best in the near future.
References
Althaus, I.A. and Mills, T.J. 1982. Resource values in analyzing fre management programs
for economic effciency. USDA Forest Service. 9 p.
Ashe, B.S.W., McAneney, K.J. and Pitman, A.J. 2008. Total Cost of Fire in Australia.
Proceedings of the III International Symposium on Fire Economics, Planning and Policy:
Common Problems and Approaches. Carolina, PR. Pp. 20.
Bateman, I. 2002. Economic valuation with stated preference techniques : a manual. Edward
Elgar, Cheltenham.
Bennett, J. and Blamey, R. 2001. The choice modelling approach to environmental
evaluation. E. Elgar Pub, Cheltenham, UK.
Bowman, D.M.J.S., Balch, J.K., Artaxo, P., Bond, W.J., Carlson, J.M., Cochrane, M.A.,
D'Antonio, C.M., DeFries, R.S., Doyle, J.C., Harrison, S.P. and others 2009. Fire in the
Earth System. Science 324: 481–484.
Butry, D.T., Mercer, D.E., Prestemon, J.P., Pye, J.M. and Holmes, T.P. 2001. What is the
price of catastrophic wildfre? Journal of Forestry 99: 9–17.
Dale, L. 2009. The True Cost of Wildfre in the Western U.S. Western Forestry Leadership
Coalition, Lakewood, CO. 16 p.
102 Towards Integrated Fire Management
Daniel, T.C., Carroll, M.S., Moseley, C. and Raish, C. 2007. People, fre and forests : a
synthesis of wildfre social science. Oregon State University Press, Corvallis.
Donovan, G.H. and Rideout, D.B. 2003. An Integer Programming Model to Optimize
Resource Allocation for Wildfre Containment. Forest Science 49: 331–335.
Dunn, A.E., González Cabán, A. and Solari, K. 2003. The Old, Grand Prix, and Padua
Wildfres: How much did these fres really cost? USDA Forest Service, Pacifc Southwest
Research Station. 18 p.
EU 2005. Council Regulation (EC) No 1698/2005 of 20 September 2005 on support for
rural development by the European Agricultural Fund for Rural Development (EAFRD).
Offcal Journal L: 1–40.
Farrell, E.P., Fuhrer, E., Ryan, D., Andersson, F., Huttl, R. and P., P. 2000. European
forest ecosystems: building the future on the legacy of the past. Forest Ecology and
Management 132: 5–20.
González-Cabán, A. 1998. Aspectos Económicos de la Evaluación del Danõ de Incendios.
USDA Forest Service, Pacifc Southwest Research Station.
González-Cabán, A. 2007. Wildland Fire Management Policy and Fire Management
Economic Effciency in the USDA Forest Service. Proceedings of the Wildfre 2007-4th
International Wildland Fire Conference. Seville, Spain.
González-Cabán, A., Shinkle, P.B. and Mills, T.J. 1986. Developing fre management
mixes for fre program planning. U.S. Department of Agriculture, Forest Service, Pacifc
Southwest Forest and Range Exp. Stn. 8 p.
Hanemann, W.M. and Kanninen, B.J. 1999. The statistical analysis of discrete-response CV
data. In: Bateman, I. and Willis, K.G. (eds.). Valuing environmental preferences: theory
and practice of the contingent valuation method in the US, EU, and developing countries.
Oxford University Press, Oxford [England]; New York. Pp. 302–442.
Hanley, N., Mourato, S. and Wright, R., E. 2001. Choice Modelling Approaches: A Superior
Alternative for Environmental Valuatioin? Journal of Economic Surveys 15: 435–462.
JRC. 2007. Forest Fires in Europe 2006. EC Joint Research Centre, Institute for
Environment and Sustainability.
Laband, D.N., González-Cabán, A. and Hussain, A. 2006. Factors that Infuence
Administrative Appeals of Proposed USDA Forest Service Fuels Reduction Actions.
Forest Science 52: 477–488.
Laband, D.N., Hussain, A. and González-Cabán, A. 2008. The impact of Forest Service
litigation success on administrative appeals of proposed fuels reduction actions. Forest
Policy and Economics 10: 444–449.
Lankoande, M.D. 2005. Three essays on wildfre economics and policy [Thesis (Ph D )].
Washington State University, August 2005.
Martin, W.E., Raish, C. and Kent, B.M. 2008. Wildfre risk: human perceptions and
management implications. Resources for the Future, Washington, DC.
Mavsar, R. and Farreras, V. 2008. Database on perception indicators. Deliverable D4.4-3 of
the Integrated Project "Fire Paradox", Project no. FP6-018505. European Commission. 48 p.
Mavsar, R., Farreras, V., Kovac, M., Japelj, A., Kraus, D. and Piwnicki, J. 2007. Database of
market values and on non-market values. Deliverable D4.4-1-2 of the Integrated Project
“Fire Paradox”, Project no. FP6-018505. European Commission. 56 p.
Mavsar, R., Weiss, G., Ramčilović, S., Palahí, M., Rametsteiner, E., Tykkä, S., van
Apeldoorn, R., Vreke, J., van Wijk, M., Prokofeva, I. and others. 2008. Study on the
Development and Marketing of Non-Market Forest Products and Services, Study Contract
No: 30-CE-0162979/00-21. EC DG AGRI. 137 p.
McCaffrey, S. 2008. Understanding Public Perspectives of Wildfre Risk. In: Martin, W.E.,
Raish, C. and Kent, B.M. (eds.). Wildfre Risk: human perceptions and management
implications Resources for the Future, Washington, DC. Pp. 11-22.
The Importance of Economics in Fire Management Programmes Analysis 103
Mercer, D.E., Jeffrey, P.P., David, T.B. and John, M.P. 2007. Evaluating Alternative
Prescribed Burning Policies to Reduce Net Economic Damages from Wildfre. American
Journal of Agricultural Economics 89: 63–77.
Mills, T.J. and Bratten, F.W. 1982. FEES: designg of a Fire Economics Evaluation System.
Pacifc Southwest Forest and Range Experiment Station, Forest Service, U.S. Department
of Agriculture. 26 p.
Morton, D.C., Ressing, M.E., Camp, A.E. and Tyrrel, M.L. 2003. Assessing the
Environmental, Social, and Economic Impacts of Wildfre. Research Paper. Yale University,
School of Forestry & Environmental Studies - Global Institute of Sustainable Forestry.
Papachristou, H. 2007. Greek August Wildfre Damage Seen at €2 Billion, Minister Says.
Bloomberg. http: //www.bloomberg.com/apps/news?pid=20601085&sid=ax1IYpsKfGmQ
&refer=europe
Petsini Arlapanou, A. and Petsini Arlapanou, L. 2007. 5 billion Euro damages. http: //www.
stockwatch.com.cy/nqcontent.cfm?a_name=news_view&ann_id=74961
Pettenella, D., Florian, D., Mavsar, R., Goio, I. and González-Cabán, A. 2009. Review of
the methodologies for the assessment of economic impacts of forest fres and economic
effciency of fre management. EFIMED, Barcelona. 108 p.
Requardt, A., Köhl, M., Schuck, A., Poker, J., Janse, G., Mavsar, R. and Päivinen, R. 2007.
Feasibility Study on means of combating forest dieback in the European Union. EC DG
Environment. 79 p.
Rideout, D.B., Ziesler, P.S., Kling, R., Loomis, J.B. and Botti, S.J. 2008. Estimating rates of
substitution for protecting values at risk for initial attack planning and budgeting. Forest
Policy and Economics 10: 205–219.
Riera, P. and Mogas, J. 2004. Evaluation of a risk reduction in forest fres in a Mediterranean
region. Forest Policy and Economics 6: 521–528.
San-Miguel, J. and Camia, A. 2009. Forest Fires at a Glance: Facts, Figures and Trends
in the EU. In: Birot, Y. (ed.). Living with Wildfres: What Science Can Tell Us? EFI
Discussion Paper 15. European Forest Institute. Pp. 11–20.
Venn, T.J. and Calkin, D.E. 2007. Challenges of accommodating non-market values in
evaluation of wildfre suppression in the United States. Proceedings of the American
Agricultural Economics Association Annual Meeting. Portland, OR.
Wade, D.D. and Lunsford, J.D. 1988. A Guide for Prescribed Fire in Southern Forests.
USDA Forest Service, Southern Region. 56 p.
Wei, Y., Rideout, D. and Kirsch, A. 2008. An optimization model for locating fuel treatments
across a landscape to reduce expected fre losses. Canadia Journal of Forest Research 38:
868–877.
Zybach, B., Dubrasich, M., Brenner, G. and Marker, J. 2009. U.S. Wildfre Cost-Plus-Loss
Economics Project: The “One-Pager” Checklist. John Wildland Fire Lessons Learned
Center. 20 p.
3.5 Improving Fire Management Success through
Fire Behaviour Specialists
Domingo Molina
1
, Marc Castellnou
2
, Daniel García-Marco
3
, and António
Salgueiro
4
1
Unit of Forest Fires, University of Lleida, Spain
2
GRAF, Department of Internal Affairs, Barcelona, Spain
3
UNAP- INFOCAM, Guadalajara, Spain
4
GAUF, National Forest Authority, Lisbon, Portugal
3.5.1 Introduction
In numerous areas of Europe, and principally in the Mediterranean countries, the last
decades have been characterized by drastic changes in land use, especially grazing
pattern and land abandonment, combined with forest extension. These changes have
caused a more dramatic wildfre spread than in previous decades. Therefore, the
task of managing a wildfre is more complex than before and requires well trained
managers who have a good understanding of fre propagation.
Specifically, in Europe dramatic fire events are characterized by being often
beyond the capacity of suppression in large areas, including the wildland urban
interface (WUI), with fast changes in fre behaviour, offering very few and localized
opportunities of extinction. In these large fres, two critical tools are required to
improve the effciency in the complex task of managing a wildfre.
1. Fire analysis to plan fre fghting tactics in advance (pro-active way) instead
of simply reacting to fre (re-acting way), thus taking advantage of the
existing opportunities.
2. The wise use of fre as a tool to fght high intensity fres. Understanding and
anticipating fre behaviour is also a critical part in managing fre as a tool.
Here, there is the need of three new Fire Specialists, requiring a higher level of
knowledge in wildfre propagation:
1. Wildland Fire Analyst (WFA), or in more detail, a Wildland Fire Propagation
Expert;
2. Firing Operations Leader (FOL);
3. Firing Specialist (FS).
Three aspects of the WFA position should be addressed.
1. How relevant is this job task in each region for this job?
2. What should be the required training (and education)?
3. Which kind of experiences can be transferred from some countries (or
regions) to others?
106 Towards Integrated Fire Management
In relation to the second aspect above, signifcant new training materials (i.e. Grillo
et al. 2008; Molina et al. 2009), specifcally aimed for such training, were recently
produced in Europe, and sharing of existing ones has been promoted. These training
documents are an innovative contribution to introducing changes into regional and
national forest and fre management services across Europe. It should be noted
that the approach toward the WFA position is different in Europe than in the USA.
A major reason is that large wildfres in the USA last longer than in Europe and
maybe far from towns. In Europe, most often the main fre runs occur in a matter
of a few hours to a few days and often affect housing areas. Therefore, in Europe,
suppression agencies not only need to forecast wildfre behaviour changes, but also
to identify opportunities in the anticipated fre spread, and to employ tactics that
take advantage of these opportunities.
This forecast has to be supported by tools like the Campbell Prediction System
Language (CPSL; Campbell 1995) and the large wildfres propagation patterns
(Castellnou et al. 2009). CPSL provides for a qualitative analysis that includes
a language to communicate estimated fire behaviour changes. CPSL has been
developed in southern California, where there are fast spreading large wildfres
that affect housing developments in a very short time; it could also be appropriately
applied for the densely populated Mediterranean Coast in Europe. On the other
hand, the propagation patterns of large wildfres, developed after detailed study
of past wildfres, helps to identity best management action in pre-attack or fuel
management. These large wildfire propagation patterns have worked well in
Catalonia (Spain) and have been used later with success in other regions of Spain
(i.e. Canary Islands, Castilla-La-Mancha, Aragon) and beyond (i.e. Portugal, Italy).
This chapter is about the need for a comprehensive knowledge on fre propagation
in support of fre management actions and the possibility to introduce some new job
positions of fre behaviour specialists.
3.5.2 The Fire Specialist Group
3.5.2.1 Tasks
As experts on fre propagation, the tasks of the Fire Specialist group (WFA, FOL,
FS) include:
1. To help plan the suppression action to be carried out. This is implemented
by means of identifying which fre perimeter sectors are above the fre
suppression capabilities (by which tool, at any given position in time and
space). Therefore, they spot safe and effcient opportunities in time and space.
Additionally, they set up a hierarchical classifcation of opportunities to
contain the fre perimeter. Then, they assign priorities on which sectors of the
fre perimeter should be contained frst and which should be addressed later
to diminish the potential fre damage.
2. To forecast fre behaviour changes in both time and space and to determine
critical points or lines where a greater alignment of factors (i.e. slope, wind,
fuel model, aspect) might put the fre behaviour out of suppression capabilities.
Improving Fire Management Success through Fire Behaviour Specialists 107
3. To train workers in safety and effcient work. Safety of personnel must be
always a priority. Second in line, fre control effciency has to be re-assessed
often. Therefore, the fre behaviour specialists must train workers in basic fre
analysis and in job hazard abatement actions all year round. In this training,
an important role has to be played by a careful study of past fre propagation
patterns and suppression actions (Molina et al. 2009) to allow for a secure
and effcient fre control.
4. To use fre as a tool. To assess or be in charge of prescribed burning actions.
To plan and perform backfres and burn out operations.
5. To lead operations combining different tools (i.e. aerial means, technical fre,
hand tools, hoses).
Each of the Fire Specialist group tasks is performed by different specialists at
different scales (i.e. time, fre sectors).
The FS is the one who decides how a fre operation should be performed in order
to fulfll objectives, in prescribed burning in rangelands or understory burning, and
in wildfres. Also, this fring specialist should forecast fre behaviour changes in time
and space and determine critical points or lines where a greater alignment of factors
(i.e. slope, wind, fuel model, aspect) might put a given fre front out of suppression
capabilities. Later, the FS has to identify and select opportunities to contain the
fre perimeter and propose tactics to implement the control. Personnel safety must
always be a priority. Finally, the FS participates and often leads operations merging
diverse suppression tools (i.e. hand tools and fre engines). A fring specialist may
train other fre fghters in basic safety procedures, basic fre analysis and merging
fre suppression tools for optimal results.
The FOL can be in charge of any operation involving the wise use of fre. Also,
the FOL helps in selecting opportunities associated to changes of fre behaviour,
recommending tactics and strategies. The FOL may also lead operations merging
different suppression tools working in a given fre sector. A FOL may also train
other fre fghters in advanced fre analysis and control and supervision of operations
to ensure that both safety and effciency would be accomplished.
The WFA is a key specialist who facilitates the wise use of fre as a tool (both in
fuel management when there is no wildfre and in actual fre suppression actions).
3.5.2.2 What to expect from a Wildland Fire Analyst?
A WFA is an expert in both fre spread patterns and fre behaviour forecast that is
employed (ideally all year round) by the agency in charge of wildfre control (i.e.
Forest Service, Emergency & Fire Fighting Service). Among the duties assigned are:
1. To elaborate an advanced forecast of fre behaviour potential in active fres,
and therefore to make an assessment on the best strategy that will take
advantage of every opportunity to weaken the fre head, and to confne or
extinguish the fre.
2. To elaborate full reviews of major past and present fres, to incorporate
lessons learned specially in fre behaviour, but also in tactics and in strategy,
to the training and response of the whole organization.
108 Towards Integrated Fire Management
3. To characterize the different wildfre propagation patterns from past fres
in the region, and therefore, to allow that new fres could be managed
as incidents with known patterns of spread (and not as random in their
behaviour), both in suppression and pre-suppression operations.
4. To assist the Incident Commander to set up the best fre control strategy for
a given fre (based on specifc wildfre propagation patterns or fre typologies
(Castellnou et al. 2009). Based on this assumption, suppression action has
a ‘winning label’ from the very beginning. In this way, both resources and
efforts are aimed to victory and there is not room for an ‘unpredictable fre
blow-up’. However, there is a need to plan for any eventual misinterpretation
or failed forecast (or failure of adequate accomplishment of the tactic by fre
workers). For this, we have the LACES protocol (Lookout, Anchor lines,
Communications, Escape routes, and Safety zones). Elsewhere are specifc
details on LACES safety protocol (Molina et al. 2009) or LCES protocol
(USDA Forest Service 2004).
5. Identify areas where forecast fre behaviour allows opportunities and track
opportunities with a light vehicle. When a potential opportunity is found, it is
assessed, and if selected, an appropriated tactic is proposed.
The WFA position flls a gap that it is too often present when setting up a Local
Emergency Command Group. It is very common to see a stressed out Incident
Commander (fre boss) not receiving the best help possible to address how to match
resources and effort with actual and future wildfre behaviour.
Specifc duties of the WFA when not involved in active fre fghting operations are:
• To document current and past wildfres and to report them in detail.
• To work on developing, applying, reviewing and improving a high-quality
standard fre safety education and training program. In this training, an
important role has to be played by detailed case studies of past fre propagation
patterns and suppression actions (Molina et al. 2009) to make future fre
control actions more secure and effcient.
• To set up the fuel management plan, the prescribed burning plan and the
infrastructure enhancement plan.
• To assess or be in charge of the implementation of any given fuel management
action or any given infrastructure enhancement action.
• It should be highlighted that the WFA is the expert that assesses actual and
forecast fre behaviour to report to the Regional Fire Resources Boss about:
• The distribution of resources covering risk and about the potential outcome of
different on-going fres to ensure that proportional action is taken and that not
all available resources go to the frst fre to break out.
• Where and how to implement slow down operations, and also how to arrange a
hierarchical classifcation of opportunities to contain the fre perimeter.
At the same time, the WFA participates with the ‘fre management specialist’ in
the tracking, the assessment and supervision of any proposed suppression action
(in terms of effciency and personnel safety). This is especially true for backfring
or burn out operations and operations combining multiple tools (Figure 1). In this
respect the contribution of the WFA is like an additional check on those actions,
Improving Fire Management Success through Fire Behaviour Specialists 109
taken from standard protocols. Therefore, it can be concluded that the role of a WFA
in Europe should go beyond just planning and should include proposing suppression
actions. This is different from what is the standard role of the equivalent specialist
in the USA.
3.5.3 Wildfre Behaviour Forecast
The main task of the WFA and FOL is to have an educated, detailed look at the
wildfre to assess potential changes in spatial and temporal fre behaviour. The WFA
forecasts behaviour of one or several fres and, as a result of this, develops a fre
control plan for the fre boss (Incident Commander) to approve. The FOL does the
same within a fre sector, and the FS does this within a fre line. This plan would
have an objective (and thus a strategy), a proposed methodology (and thus a tactic
and explicit limitations), and an execution window as a result of the objective and
methodology. In short, this plan would have a strategy, a tactic and an execution
window. The tactic and execution window would be addressed most likely by a FOL
and a FS, and maybe not by a WFA. The WFA does this not only during fre events
to support suppression, but also in advance to support both planning and training.
Figure 1. Backfring operations supervision in Gran Canaria 2007 (Canary Islands, Spain).
(Photo by Daniel García-Marco).
110 Towards Integrated Fire Management
3.5.4 Suppression planning (objective and strategy)
This is the process of establishing the best plan to accomplish secure and effcient
fre control when suppression forces arrive to the fre scene. To do so, it is necessary
to forecast fre behaviour changes in time and space and determine critical point or
lines where a greater alignment of factors (Campbell 1995) might put the fre out of
our suppression capabilities. The suppression plan is a set of sequential actions to be
taken within a time frame. It may be a written document or an oral communication
(most agencies record all radio telecommunications today). It is not required that
this plan is something too intricate. On the contrary, it has to be simple; easy to
understand by those in charge of executing it (i.e. this requires simple language and
clear maps). The WFA helps the Incident Commander to set up this plan in one or
several wildfres. The FOL may help the Division Supervisor, and the FS may help
the Crew Boss.
This suppression plan has to be communicated and we must check that it has been
properly understood (mandatory checking action). Security issues must always be a
major part of the plan. Additionally, a drawing in a map may be provided to allow a
better understanding of tasks and duties, as well as the current fre scenario and its
potential runs (dead man zone concept, see below).
In case of failure to control the fre through this suppression plan, maps and
drawings might help to set up an alternative plan. Failure could be due to any
miscommunication among fre workers.
3.5.4.1 Strategy to control fre propagation
The strategy for fire propagation control is about how to establish the fire
control objectives. The WFA helps the Incident Commander to plan for a secure
and effcient fre control with the best strategy possible. It is up to the Incident
Commander to decide the strategy after listening to the WFA and the leaders of the
different agencies involved in the Incident Command System Group.
The objectives have to be communicated to all personnel involved in this
emergency. Objectives have to be easy to communicate, easy to recognize, and with
clear magnitudes. A bad objective could be “We want to control the fre as soon as
possible”. A better objective could be “we want to control the fre, slowing down
the head in these felds in the bottom of the valley. The left fank should be stopped
before this creek while the right fank should be anchored before this other creek on
next morning. So in this moment, the priorities are the head and then the left fank
with larger potential over the right one”. But the priorities should be re-assessed
according to changes in fre behaviour. The second objective has clear magnitudes
(referenced to a map) and could be tested later to see if it was accomplished or not.
This is not the case with the frst objective.
It is the duty of the WFA to re-asses (or validate) that the proposed fre control
lines are both secure and effcient to establish.
Lastly, sometimes the chosen strategy may be not to address the head fre because
it is out of suppression capability even with indirect attack. It is important to
recognize when there is not any possible safe and effcient tactic to use. In that case
the head fre should be addressed later in time and space (see text box).
Improving Fire Management Success through Fire Behaviour Specialists 111
Box 1. Torre de Fontaubella 2007 (Catalonia, Spain)
Initial strategy: Set up by Incident Commander (the commandment of the frst arriving
crews) helped by the FOL. The objective is to contain fre in area 1. Tactic, set by this same
Incident Commander, is anchoring the back and fanking towards the head. Therefore,
1. avoiding backing fre spreading to area 2;
2. avoiding right fank spreading to area 4;
3. containing downhill fre approaching area 3.
The fre goes from area 1 to area 4 by spotting driven by strong northwest wind (not
surface topographic wind). The key factor is the lack of pre-attack infrastructure to help
suppression efforts.
Area 2 is a key sector in the spread of the fre, is contained out of the sector with few
hoses and fghting multiple spotting.
The fre goes slowly through area 3 and the execution window is very wide.
Extended attack strategy: Set up by Incident Commander helped by the Operation
Section Chief and the WFA, present in the feld one hour after the fre began.
The fre is in areas 1 and 4. The objective is to hold the fank where fre runs will begin
driven by wind in the night when the northeast wind goes down to surface. Therefore,
1. hold the back of the fre out of area 2;
2. slow the head spread to reduce the fre front that arrives at area 6 and, with it,
reduce the amount of suppression resources in a front with no potential of spread
but with populated areas;
3. reduce the wide of right fank that goes to area 7, where the northeast expected
winds can spread the fre to area 8 and then to area 5;
4. stop the fre that goes downslope in area 6 and the left fank.
The head and the back of the fre are under control (the fre does not spread to area 2, and
nor to area 6). Part of left fank is under control. For safety reasons the hose lines on the
right fank cannot hold all the line but is narrow. The front that jumps to area 7 is narrow
and this slows the rate of spread.
Night extended attack strategy: Set up by Incident Commander helped by the Operation
Section Chief and the WFA.
Fire on area 7. The objective is to avoid the fre reaching area 8. Therefore,
• avoiding that fre spreading to area 2 and to the populated areas of the village Torre
de la Fontaubella, where the fre is making head runs;
• avoiding fre runs following the crests in area 7 and spreading to area 8 where there
are no anchor points to begin the fre fghting work;
• stopping the left descending fank.
112 Towards Integrated Fire Management
3.5.4.2 Tactics
The term tactics refers to the planning and execution of actions for accomplishing
the strategy. Different tactics are available. The best tactics (to allow for a secure
and effcient fre control) should be taken considering the resources available, the
training, and the forecasted fre behaviour changes in time and space and trying to
determine critical points or lines where a greater alignment of factors might put the
fre out of suppression capabilities. Safety of personnel must always be a priority.
For a given strategy, several tactics could be possible in order to implement
the control goals. Choosing one tactic among others is based on specifc training,
resources available, suppression protocols, and forest regulations. It is the way to
pursue the goals set by the strategy, always looking for opportunities that (for the
given resources) will maximize fre fghting effectiveness without compromising
safety. Strategy is up to the Incident Commander, while tactics are up to the Division
Supervisors and Crew Bosses. The Incident Commander has the leadership, the
planning responsibility, and the management and co-ordination of resources. As
large fres may change so quickly, decisions about tactics in a particular fre sector
can be taken by the Division Supervisor and in any fre perimeter site by the crew
boss. Let us look at an example: the chosen strategy in a fre sector is to control that
fre line at the road. Then, the Division Supervisor may choose:
• to control it with engines and hoses;
• to rely on both engines and air tankers;
• to backfre from the road.
All of them are different possible tactics. In Figure 1, fre fghters are igniting fre
to consume fuels in advance of the spreading wildfre; this approach can be very
effective in limiting the spread of the main fre.
3.5.4.3 Execution window for the tactic
It is the third part of the wildfre control plan. It refers to the limits in time and space
of the tactic that was decided upon. As we mentioned above, any possible tactic
has to be validated by checking whether there is an appropriate “time execution
window” and “space execution window”, but also the “weather window” and the
“fre behaviour window”.
Sometimes, there is a starting restriction in the “execution window”; i.e. a
backfring operation may require steady wind suction in the fre line to conduct an
effcient suppression fre. Therefore, we cannot light the fre line until suction (from
the main fre) is noticeable. Another example could be when a wind change (i.e.
timed with day/night topographic changes, or land/sea breeze) is required to allow a
safe and effcient tactic to be performed.
The execution window is not only useful to ensure effciency, but it is also part
of the LACES safety protocol (Molina et al. 2009) or LCES protocol (USDA
Forest Service 2004). Setting an appropriate execution window helps the look-outs
understand what they should be looking for, and what they should be communicating.
Somehow adding new dimensions the execution window in a given fre sector
can be described like this: “our execution tactic would be effective (and therefore,
Improving Fire Management Success through Fire Behaviour Specialists 113
it would be still in shape) before the fre reaches the counter-wind area, before the
daily topographic wind arrives, before wind changes direction due to storms, and
while the fre perimeter sends frebrands less than 20 m away”.
3.5.5 Wildfre Analysis: a needed tool
Fire spread analysis is the needed tool to accomplish a safe and effcient wildfre
propagation control. This is especially true in fast large wildfres involving wildland
urban interface areas. There are three keys to success:
• To have a fast forceful response: increasing distributed vigilance and
suppression resources, directing prevention efforts to increase suppression
effciency and widening and combining the techniques used to contain fre
spread.
• To have a suppression agency fexible enough to respond to rapid changes of
fre behaviour: It will be necessary to plan in advance through fre behaviour
anticipation (pro-active approach) and not to react to fre (re-acting approach).
It will require lowering the decision levels to ensure a faster response to
changes. Lastly, it will require a common technical language to describe fre
behaviour changes,
Box 2. Tactics in an extended attack at Torre de Fontaubella 2007
(Catalonia, Spain)
• Backfre at head of the fre front. Set up by the WFA and the Incident Commander.
Executed by the WFA.
• Hose line in the back of the fre. Upslope hose lines in the fanks set up by the
Division Supervisor.
• Downslope hose line in both fanks helped by aircrafts as lookout, burned area as
safety zone. Burn out to make wider the safety zone, made successfully in the left
fank and stopped for safety reasons in right fank. Set up by WFA and Incident
Commander, applied by Division Supervisor and FOL, and executed by Strike
Team Leader (the boss of a three fre engine team).
• Aircrafts help the backfre operation and later on right fank. The suppression fre is
set up by the WFA under the Incident Commander supervision.
Execution window for backfre on the head fre
• Meteorology. Starting the backfre when the suction of the fre holds the ignition
inside the fre line.
• Fire Behaviour. Starting the ignition before the fre begins spotting.
• Safety. Starting the ignition after successful evacuation of the people remaining on
the road.
• Safety. Firing from left fank to right fank for a better way out for fring specialists.
114 Towards Integrated Fire Management
• To be able to use fre as a tool to contain high intensity fre spread situations. It
will require slowing down the fre head (high intensity fre behaviour), while
allowing the standard fre-line construction to contain the fre perimeter.
As we have seen, for all these key aspects, applying fre analysis in every job position
of a fre fghting organization is of paramount importance. The existence of Fire
Specialists fosters these three key aspects, promoting the use of fre analysis in the
three spatial scales of the fre (fronts, sectors, wildfres), and so improving knowledge
and understanding all through the different levels in the chain of command.
In the words of some Incident Commanders we have interviewed, “to have a fre
analyst as an assistant to the chief of the suppression forces (Incident Commander)
is like having the peaceful feeling that our efforts will end up in success or that we
are not just playing around the fre spread to justify our salary”. In summary, safe
and effcient wildfre suppression is harder to be accomplished without a proper
WFA on duty.
Going into the details, the vision of an expert in fre analysis (i.e. WFA, FOL, FS)
does differ from the perspective of an Incident Commander who is a manager of
resources. Specifcally, the WFA has specialized knowledge on fre behaviour, fre
effects, and fre safety, and has the time and focus to apply this knowledge to fre
analysis. Therefore, the WFA can properly identify fre suppression opportunities
but can also assess in time when some efforts are either meaningless (ineffcient to
control the fre perimeter) or unsafe for the workers.
3.5.5.1 Agencies with Fire Specialist positions on duty in 2009
The current situation is rapidly improving over recent years in many regions in
Europe. Clearly this progress needs to continue and to multiply. In Table 1, some
specifc data is provided regarding Fire Specialists and their training. It is perceived
that the investment in training and positions results in improved fre management
effectiveness. These frst steps in Europe are promising, but the efforts need to
expand and accelerate – positions need to be recognized and become a part of the
regular fre operations during fres and year-round to support planning and training.
3.5.5.2 Required competences training and experience
The required level of education of a WFA should include a university degree in
Forestry (or related degree), and an additional specifc postgraduate or master degree
in Forest Fire Management. Several years in fre fghting operations are certainly
important to qualify as a WFA. However, it should be highlighted that it should be
possible to transfer experience through high quality case studies. Lifelong learning
(not necessary at universities) might permit the acquisition of formal qualifcations
and competencies. Therefore, some people may qualify for these positions with
no formal academic degrees. However, the usual way to reach the qualifcations is
through attendance at a course run by a university.
This is a needed job position and requires a specific education and training.
This specifc degree in Forest Fire Management must include credits in prescribed
Improving Fire Management Success through Fire Behaviour Specialists 115
T
a
b
l
e

1
.

R
e
g
i
o
n
s

(
a
n
d

c
o
u
n
t
r
i
e
s
)

a
n
d

t
h
e
i
r

F
i
r
e

S
p
e
c
i
a
l
i
s
t

p
o
s
i
t
i
o
n
s

o
n

d
u
t
y

i
n

2
0
0
9
.

R
e
g
i
o
n

W
i
l
d
l
a
n
d

F
i
r
e

A
n
a
l
y
s
t

F
S

/

F
O
L


F
i
r
e

A
n
a
l
y
s
i
s

t
r
a
i
n
i
n
g
C
a
t
a
l
o
n
i
a

(
n
o
r
t
h
e
a
s
t

S
p
a
i
n
)


8

f
u
l
l

t
i
m
e
,

4

a
d
d
i
t
i
o
n
a
l

S
e
v
e
r
a
l

F
S
,

a
n
d

F
O
L

F
i
r
e

a
n
a
l
y
s
i
s

f
o
r

a
l
l

l
e
v
e
l
s
1
.

C
o
n
t
i
n
u
o
u
s

t
r
a
i
n
i
n
g
:

P
a
s
t

w
i
l
d
f
r
e
G
R
A
F
.

S
i
n
c
e

1
9
9
9

a
s
s
i
s
t
a
n
t
s
.


a
n
a
l
y
s
i
s

a
n
d

a
c
t
u
a
l

w
i
l
d
f
r
e

r
e
p
o
r
t
s
.

G
R
A
F

u
n
i
t
s

t
r
a
i
n
i
n
g



i
n
c
l
u
d
e
s

2
0
0

h

o
f

t
r
a
i
n
i
n
g

(
F
S
)
,

1
5
0

h

f
o
r

F
O
L
,

i
n
c
l
u
d
i
n
g



p
r
e
s
c
r
i
b
e
d

b
u
r
n
i
n
g

a
n
d

a
n
a
l
y
s
i
s

e
x
e
r
c
i
s
e
s

f
o
r

G
R
A
F
.
P
o
r
t
u
g
a
l

(
a
l
l

c
o
u
n
t
r
y
)
.

4



T
h
e
r
e

i
s

n
o
t

y
e
t

S
e
v
e
r
a
l

F
S
,

a
n
d

F
O
L

G
A
U
F

u
n
i
t
s

t
r
a
i
n
i
n
g

i
n
c
l
u
d
e
s

5

d
a
y
s

t
r
a
i
n
i
n
g

s
e
s
s
i
o
n
s

o
f

f
r
e
G
A
U
F
2
.

S
i
n
c
e

2
0
0
6
.

a
n

o
f
f
c
i
a
l

f
u
n
c
t
i
o
n


a
n
a
l
y
s
i
s

a
n
d

p
r
e
s
c
r
i
b
e
d

b
u
r
n
i
n
g
s
.

L
´
A
u
d
e

(
F
r
a
n
c
e
)
,

G
R
A
F
F

-

F
S

T
r
a
i
n
i
n
g

f
r
e

I
n
c
i
d
e
n
t

C
o
m
m
a
n
d
e
r
s

i
n

f
r
e

a
n
a
l
y
s
i
s
.
G
r
a
n

C
a
n
a
r
i
a

(
S
p
a
i
n
)
.


2

f
u
l
l

t
i
m
e
,

1

a
d
d
i
t
i
o
n
a
l

S
e
v
e
r
a
l

F
S
,

a
n
d

F
O
L

F
i
r
e

a
n
a
l
y
s
i
s

f
o
r

m
o
s
t

l
e
v
e
l
s
.

C
o
n
t
i
n
u
o
u
s

t
r
a
i
n
i
n
g
:

P
a
s
t

w
i
l
d
f
r
e
S
i
n
c
e

2
0
0
2
.

a
s
s
i
s
t
a
n
t

i
n

s
u
m
m
e
r


a
n
a
l
y
s
i
s

a
n
d

w
i
l
d
f
r
e

r
e
p
o
r
t
s

(
i
n

p
r
o
g
r
e
s
s
)
.



C
a
s
t
i
l
l
a
-
L
a

M
a
n
c
h
a

S
i
n
c
e

2
0
0
9
,

1
4

f
u
l
l

t
i
m
e


I
n

p
r
o
g
r
e
s
s

F
i
r
e

a
n
a
l
y
s
i
s

f
o
r

m
o
s
t

l
e
v
e
l
s

(
i
n

p
r
o
g
r
e
s
s
)
.

C
o
n
t
i
n
u
o
u
s

t
r
a
i
n
i
n
g
:

(
S
p
a
i
n
)
.

U
N
A
P

s
i
n
c
e

2
0
0
9
.

(
s
o
m
e

i
n

t
r
a
i
n
i
n
g

p
h
a
s
e
)
.



P
a
s
t

w
i
l
d
f
r
e

a
n
a
l
y
s
i
s

a
n
d

w
i
l
d
f
r
e

r
e
p
o
r
t
s

(
i
n

p
r
o
g
r
e
s
s
)
.
A
r
a
g
ó
n

(
S
p
a
i
n
)

5

f
o
r
e
s
t

f
r
e

m
a
n
a
g
e
r
s

i
n

t
r
a
i
n
i
n
g

N
o

A
n
d
a
l
u
c
í
a

(
S
p
a
i
n
)
.


E
x
p
e
c
t
e
d

2
0
1
0
.

M
e
t
h
o
d
s

a
n
d



F
i
r
e

c
o
m
m
a
n
d
e
r
s

a
r
e

b
e
i
n
g

t
r
a
i
n
e
d

i
n

f
r
e

a
n
a
l
y
s
i
s
.

p
r
o
t
o
c
o
l
s

a
r
e

t
e
s
t
e
d

i
n

2
0
0
9
3


T
e
n
e
r
i
f
e

(
S
p
a
i
n
)

T
h
e
r
e

a
r
e

m
o
v
e
s

t
o
w
a
r
d
s


4

F
O
L

(
i
n

t
r
a
i
n
i
n
g
)

4

f
o
r
e
s
t

m
a
n
a
g
e
r
s

a
n
a
l
y
z
e

p
a
s
t

f
r
e
s
,

f
o
l
l
o
w

a
c
t
i
v
e

f
r
e
s

e
m
p
l
o
y
m
e
n
t

o
f

W
F
A
.
4

S
a
r
d
i
n
i
a

(
I
t
a
l
y
)
.

M
a
s
t
r
o
s


N
o

F
S

i
n

t
r
a
i
n
i
n
g

B
a
s
i
c

t
r
a
i
n
i
n
g

i
n

w
i
l
d
f
r
e

a
n
a
l
y
s
i
s
d
e

F
o
g
u
.

2
0
0
8



C
o
n
t
i
n
u
o
u
s

t
r
a
i
n
i
n
g

t
h
r
o
u
g
h

p
a
s
t

w
i
l
d
f
r
e

a
n
a
l
y
s
i
s
.
N
o
r
t
h
u
m
b
e
r
l
a
n
d

(
U
K
)

-

N
o

T
r
a
i
n
i
n
g

h
i
g
h

a
n
d

i
n
t
e
r
m
e
d
i
a
t
e

f
r
e

c
o
m
m
a
n
d
e
r
s

i
n

F
r
o
n
t

a
n
d



S
e
c
t
o
r

f
r
e

a
n
a
l
y
s
i
s
.

U
n
d
e
r
s
t
o
r
y

p
r
e
s
c
r
i
b
e
d

b
u
r
n
i
n
g
.

1

F
i
r
e

s
u
p
p
r
e
s
s
i
o
n

s
u
p
p
o
r
t
,

B
a
s
i
c

F
i
r
e

s
u
p
p
r
e
s
s
i
o
n
,

F
i
r
e

B
r
i
g
a
d
e

u
n
i
t

c
o
m
m
a
n
d
e
r

f
o
r

F
i
r
e

s
u
p
p
r
e
s
s
i
o
n

o
p
e
r
a
t
i
o
n
s
,

D
i
v
i
s
i
o
n

c
o
m
m
a
n
d
e
r

f
o
r

f
r
e

s
u
p
p
r
e
s
s
i
o
n

o
p
e
r
a
t
i
o
n
s

i
n

M
e
d
i
t
e
r
r
a
n
e
a
n

E
c
o
s
y
s
t
e
m
s
,

I
n
c
i
d
e
n
t

C
o
m
m
a
n
d
m
e
n
t
,

L
o
o
k
-
o
u
t
,

C
o
n
t
r
o
l

o
p
e
r
a
t
o
r
s
,

A
e
r
i
a
l

m
e
a
n
s

o
p
e
r
a
t
o
r
s
.
2

A
l
l

G
A
U
F

m
e
m
b
e
r
s

a
r
e

f
o
r
e
s
t
r
y

e
n
g
i
n
e
e
r
s

a
s

w
e
l
l

a
s

c
e
r
t
i
f
e
d

a
s

p
r
e
s
c
r
i
b
e
d

f
r
e

s
p
e
c
i
a
l
i
s
t
.

E
v
e
r
y

y
e
a
r

s
i
n
c
e

2
0
0
6

t
h
e
y

o
r
g
a
n
i
z
e

a

5

d
a
y
s

t
r
a
i
n
i
n
g

s
e
s
s
i
o
n
s

c
o
n
c
e
r
n
i
n
g

t
h
e

w
i
l
d
f
r
e

a
n
a
l
y
s
i
s
.

I
n

t
h
e

c
i
v
i
l

p
r
o
t
e
c
t
i
o
n

s
y
s
t
e
m
,

i
n

P
o
r
t
u
g
a
l
,

t
h
e
r
e

i
s

n
o
t

y
e
t

a
n

o
f
f
c
i
a
l

f
u
n
c
t
i
o
n

o
f

f
r
e

a
n
a
l
y
s
t
,

b
u
t

G
A
U
F

p
e
r
s
o
n
n
e
l

a
r
e

o
f
t
e
n

r
e
q
u
e
s
t
e
d

t
o

h
e
l
p

i
n

t
h
i
s

t
a
s
k
.
3

A
n

a
n
a
l
y
s
i
s

u
n
i
t
,

f
o
r
m
e
d

b
y

f
u
l
l

t
i
m
e

W
F
A

a
s

w
e
l
l

a
s

p
a
r
t

t
i
m
e

W
i
l
d
l
a
n
d

F
i
r
e

O
p
e
r
a
t
i
v
e

T
e
c
h
n
i
c
i
a
n
s
,

i
s

e
x
p
e
c
t
e
d

t
o

b
e

o
p
e
r
a
t
i
v
e

b
y

s
u
m
m
e
r

2
0
1
0
.
4

T
h
e
r
e

i
s

a

m
o
v
e

t
o
w
a
r
d
s

t
h
i
s

n
e
w

j
o
b

p
o
s
i
t
i
o
n

b
u
t

t
h
e
r
e

i
s

a

c
o
m
p
l
e
x

p
r
o
t
o
c
o
l

(
i
n

l
a
r
g
e

w
i
l
d
f
r
e
s
)

t
h
a
t

a
c
c
o
m
m
o
d
a
t
e
s

s
e
v
e
r
a
l

d
i
f
f
e
r
e
n
t

t
a
s
k
s

f
o
r

u
p

t
o

f
o
u
r

f
o
r
e
s
t

f
r
e

m
a
n
a
g
e
r
s

o
n

d
u
t
y
.

T
h
i
s

c
o
u
l
d

e
v
e
n
t
u
a
l
l
y

a
l
l
o
w

f
o
r

a

m
o
r
e

s
p
e
c
i
f
c

W
F
A

p
o
s
i
t
i
o
n
.
116 Towards Integrated Fire Management
burning techniques, fre behaviour, fre effects, fre analysis, strategies and tactics,
and the use of prescribed fre in ecosystem management. Fire analysts may enhance
their performance by becoming knowledgeable about the body of works by
Campbell (1995), Castellnou et al. (2009), and Cheney et al. (2001).
By addressing European-wide training needs, the new training systems and
practices will strengthen European co-operation, will develop common principles
for informal learning, and will amplify support at the local service level for the
development of qualifcations and competencies as WFA. These training materials
will include fre science, fre ecology, fre weather, the social and cultural role of fre
in Europe, fre prevention and suppression methods and technologies and the use of
prescribed fre in ecosystem management. Additionally, they will cover the state-of-
the-art of scientifc knowledge in fre ecology in the European biota, the impacts of
fre on atmospheric chemistry, climate, human health and security. Country ‘Annexes’
providing specific information on particularities of wildland fuels, management
options and country specifc legislature and protocols may be included.
In terms of suitable education, training, experience and certification of a
WFA, these should be similar to what it is required for an Incident Commander.
Additionally, the WFA should be:
• Familiar with Wildfre Typologies in the region (Castellnou et al. 2009)
and capable to forecast and assess fre behaviour changes under changing
topography and weather (Campbell, 1995; Castellnou et al. 2009). The WFA
should also be able to use fre simulators (Finney 1998, Finney et al. 1997,
Molina et al. 2006).
• Familiar with prescribed burning techniques. This helps to understand minor
changes in fre behaviour and how to address them. Expertise (both education
and extensive training) in the wise use of fre (prescribed burning as well as
backfring operations in suppression actions).
• Knowledgeable in fre ecology and fre effects (this may be obtained partially
in prescribed burning actions).
• Familiar with job hazards abatement actions and aptitudes (Beaver 2001,
Braun 1995, and Geller 1996), and with conditions leading to accident
identifcation (Putnam 1995, Hart 1995, USDA Forest Service 2001, Mangun
1995, Mangun 1999) in order to carry-out fre analyses that help achieve
effcient and safe fre suppression.
The required extensive training may include:
• A manager level position in Forest Fire Management during at least two fre
seasons.
• A close-up participation in recreation and analysis of fre spread in numerous
past fres as shown by Molina et al. (2006) and Grillo et al. (2008). In those
works, several case studies are shown, in which the fre control actions either
did match or did not match the fre spread pattern, as a result of this pattern
either being properly perceived or not being properly perceived.
• A deep knowledge of fre ecology, in regard to how fre effects are linked to
fre behaviour. This is something not required for the Incident Commander but
it is a must for a WFA.
Improving Fire Management Success through Fire Behaviour Specialists 117
In summary, it is obvious that the WFA position is very important and deserves
recognition. It is also clear that a proper certifcation system is needed to ensure
that the most qualifed candidates are hired for the job. There are already different
attempts towards this accreditation throughout Europe (see Miralles et al. in the
present volume).
As to the required competences needed for the Firing Specialist and the Firing
Operations Leader positions, see Miralles et al. in the present volume.
3.5.6 Discussion
Molina et al. (2009) have shown several case studies in which the fire control
actions did not match the fre spread pattern because this last one was not perceived
properly. It should be highlighted that having an expert wildfre analyst on duty in
those cases could have made a signifcant difference in the fnal fre perimeter and in
human safety (including job hazard abatement actions).
The certifed WFA may perform several tasks in both prevention and suppression.
WFAs are able to write down fuel management plans or integrated fre management
plans. Additionally, prescribed burning plans are their duty.
They can also carry out a host of other tasks, as listed above, in relation to fre
suppression. These assignments may be carried out by any fre manager, but they
can be accomplished much better by a certifed WFA.
It is rewarding that there are new positions as WFA in different agencies; it has
also contributed in enhancing the credit of those already doing this highly needed
job. Having paved the way, it is certain that there will be even more WFAs in more
agencies in the coming years. For example, some recent developments (i.e. Castilla-
La-Mancha, Spain) have set up excellent working documents to ensure a smooth
transition to a system that relies heavily on the WFA tasks.
This chapter has clearly been based on Fire Paradox activities including
monitoring of active wildfres, demonstration sites and training actions in prescribed
burning and detailed documentation of fre analysis of recent medium and large
wildfres. Therefore, it integrates diverse aspects of wildfre propagation using
the knowledge obtained. These diverse aspects (that indeed provide background
knowledge to Fire Specialists) merge wildfre propagation with the wise use of
fre. In fact, learning and training in prescribed burning provides an outstanding
understanding of wildfre propagation because they allow for many hours to observe
changes in fire behaviour and spread. If these hours are spent under adequate
supervision, the trainees should be encouraged to anticipate fre behaviour changes
(a pro-active look at fre spread).
3.5.7 Conclusions and recommendations
Field experience, large wildfire analysis, and arguments stated in this chapter
strongly suggest that it is clearly benefcial to have a WFA on duty.
118 Towards Integrated Fire Management
Fire specialists (i.e. WFA, FOL, and FS) would effciently identify critical points
(or lines) where fre intensity and spread could increase seriously (i.e. critical points)
and opportunity sites (i.e. infection points) where fre suppression actions would
have an increased probability of success.
Fire specialists would have the knowledge to propose adequate strategies and
organize tactical operations. However, knowledge on fre propagation should be
common to all people involved in fighting fires. Therefore, as the top experts,
the Fire Specialists should train teams in job hazard abatement actions and in fre
fghting exercises all year round.
Additionally, it is necessary to address how to certify the required qualifcations
to ensure that those job positions will be flled with the best workers available.
It is envisioned that the Fire Specialists should be standard job positions in those
wildland suppression actions that exceed the limits of initial dispatch.
It is strongly recommended that Fire Specialists should be viewed as being
equally indispensable as helicopters, airplanes, fre engines or logistic experts.
Therefore the message to policy makers is to encourage agencies to have WFAs on
duty for mid- and large-size wildfres.
In short, fre behaviour specialists would help to solve the fre paradox by means
of merging a comprehensive knowledge on fre propagation with mastery of the
wise use of fre.
References
Beaver, A.K. 2001. Evaluating risk and reward relationships in wildland fre-fghter safety.
Proceedings of the 5
th
International Fire Safety Summit. International Association of
Wildland Fire. Missoula, MT.
Braun, C.C. 1995. Addressing the common behavioural element in accidents and incidents.
In: Putnam, T. (ed.). Proceedings of the fndings from the wildland fre-fghters human
factors workshop. U.S. Department of Agriculture, Forest Service. 9951-2855-MTDC.
Missoula, MT. Pp. 28–30.
Castellnou, M., Miralles, M., Molina, D.M. and Martinez, E.R. 2009. Patrones de
Propagación de Incendios Forestales: In: Vélez, R. (ed.). Incendios Forestales:
Fundamentos y Aplicaciones. McGraw-Hill. Pp. 274–282.
Campbell, D. 1995. The Campbell Prediction System: A Wild Land Fire Prediction System
& Language. D. Campbell ed. 129 p.
Cheney, N.P., Gould, J.S. and McCaw, L. 2001. Dead Man Zone – a neglected area of fre-
fghters safety, Australian Forestry 64(1): 45–50.
Finney, M.A. 1998. FARSITE: Fire Area Simulator-model development and evaluation.
USDA Forest Service, Research Paper RMRS-RP-4, Rocky Mountain Research Station,
Ft. Collins, CO. 47 p.
Finney, M.A., Sapsis, D.B. and Bahro, B. 1997. Use of FARSITE for Simulating Fire
Suppression and Analyzing Fuel Treatment Economics. Symposium on Fire in California
Ecosystems: Integrating Ecology, Prevention, and Management. San Diego, CA.
Geller, E.S. 1996. The psychology of safety. Chilton Book Company, Radnor, PA.
Hart, D.O. 1995. Cultural attitudes and change in high-stress, high-speed teams. In:
Proceedings of the fndings from the wildland fre-fghters human factors workshop.
Putnam, T. (ed.). U.S. Department of Agriculture, Forest Service. 9951-2855-MTDC.
Missoula, MT. Pp. 34–39.
Improving Fire Management Success through Fire Behaviour Specialists 119
Grillo, F.F., Castellnou, M., Molina, D.M, Martínez, E.R., and Fababú, D.D. 2008. Análisis
del Incendio Forestal: planifcación de la extinción, Editorial AIFEMA, Granada, Spain.
144 p.
Mangun, R.J. 1995. Investigating wildland fre entrapments. U.S. Department of
Agriculture, Forest Service. 9551-2845-MTDC. Missoula, MT.
Mangun, R.J. 1999. Wildland fre fatalities in the United States 1990 to 1998. U.S.
Department of Agriculture, Forest Service. 9951-2808-MTDC. Missoula, MT.
Molina, D.M. 2009. Planes de quemas. Prescripciones. In: Vélez, R. (ed.). Incendios
Forestales: Fundamentos y Aplicaciones. McGraw-Hill. Pp. 274–282.
Molina, D.M., Blanco, J., Galan, M., Pous, E., García, J.B. and García, D. 2009. Incendios
Forestales: Fundamentos, Lecciones Aprendidas y Retos de Futuro. Editorial AIFEMA,
Granada, Spain. 256 p.
Molina, D.M., Grillo-Delgado, F. and García-Marco, D. 2006. Uso del fuego prescrito para
la creación de rodales cortafuegos: estudio del caso “Las Mesas de Ana López”, Vega de
San Mateo, Gran Canaria, España. Invest Agrar: Sist Recur For 15(3): 271–276.
Putnam, T. 1995. (ed.). Findings from the wildland fre-fghters human factors workshop.
U.S. Department of Agriculture, Forest Service. 9951-2855-MTDC. Missoula, MT.
USDA Forest Service 2001. Thirtymile fre investigation: Accident investigation factual
report and management evaluation report. USDA Forest Service, Winthrop. 106 p.
USDA Forest Service 2004. Safety Protocol Review. Southern California Geographic Area.
Fire Siege of 2003. USDA Forest Service, Pacifc Southwest Region. 34 p.
3.6 Wildfre Scenarios: Learning from Experience
Marc Castellnou
1
, Asier Larrañaga
1
, Marta Miralles
1
, Oriol Vilalta
1
and
Domingo Molina
2
1
GRAF, Department of Internal Affairs, Barcelona, Spain
2
Unit of Forest Fires, University of Lleida, Spain
3.6.1 Introduction
The occurrence of episodes with large wildfres has affected different regions of
Europe; Portugal in 2003 and 2005, southeast France in 2003 and 2009, northwest
Spain in 2003 and 2009, and Greece in 2000, 2007 and 2009. This demonstrated that
strong suppression resources capabilities are still inadequate when faced by extreme
fre behaviour. During all these large fres suppression crews faced extreme freline
intensities and fast fre spread which made it impossible to quickly suppress them by
the known attack techniques (direct attack).
The applied strategy for most wildfre suppression cases was based on a fast and
aggressive attack of all ignition points in the landscape, at every place and under
all meteorological conditions. The effciency of this strategy is outstanding and it
succeeds in limiting most wildfres to very small burned areas. But under certain
meteorological, fuel and topographical conditions, fre behaviour becomes extreme.
When fre behaviour is clearly superior to the power of the suppression resources, the
system becomes overwhelmed and turns out not to be effcient for various reasons.
• The distribution of the resources along the perimeter is slower than the growth
of the wildfre perimeter. In other words, fre spread is faster than the fre
suppression rate. For example, massive spotting at 200 m is very usual in large
fres, so in an instant fre can advance by 200 m.
• With such extreme fre intensities, the suppression system cannot carry its
operations effciently as in the case under normal conditions. Extreme fre
intensity thus overruns suppression limits. For example, crown fres with large
walls of fame greater than 25 m length are not unusual in large fres, and no
aerial or terrestrial resource can stop it.
• Fire perimeter growth, which is faster than the suppression capacity, increases
the probability of damage to houses and people and the necessity of actions
and resources to protect them. In the wildland urban interface situation,
a negative feedback process has developed, where the relative amount of
resources used to fght the wildfre is lower in comparison than the resources
which are used to protect the houses and people. Every house can be regarded
as an emergency (Rifà and Castellnou 2007).
Suppression of small fres can be effective under virtually all conditions however
it is reasonable to say that once a fire develops under certain extreme weather
conditions it is impossible to limit fre spread until the weather moderates. But
this strategy that only focuses on total suppression has shown not to be successful
122 Towards Integrated Fire Management
under large wildfres. So, to mitigate wildfre effects it is necessary to focus on
changing the landscape, towards a more tolerant and resistant land use. However,
during the next 30 years, we will not have the desired landscape yet, and we will
have large fres that no suppression effort can stop. So we need to gain suppression
effectiveness to address these fres. Fire fghters cannot stop them. A change in
strategy is needed, from strategies based on direct attack to strategies based on
slowing and reducing fire behaviour and defending urban assets. This clearly
requires an important investment in knowing and anticipating fre behaviour.
Anticipating wildfire spread pattern provides the opportunity to increase the
success-rate of the suppression attacks by allowing fre fghters to pre-plan decisions
and timely application of suppression methods. The anticipation of the attack
opportunities is far from being a standard in suppression systems. A great opportunity
to begin this task is before the fre occurs, included in prevention programs: locating
the places where adequate and proactive forest management could reduce fire
intensity, creating opportunities for the likelihood of successful suppression.
Different research initiatives have attempted to make progress in this direction:
• Wildfre spread patterns. The dominant spreading factor can be determined, and
in this way it is possible to pre-determine which type of changes can be expected
and will generate opportunities (developed in the frst half of the 20
th
century).
• Surface-fre prediction by static simulations (Behave) (Andrews 1986) or
spatial simulators of fre spread (Finney 1998) can be used with advantage to
anticipate sites where lower fre intensities can be expected.
• Simulations based on minimum travel time which permit identifcation of
fre opportunities (Finney et al. 2002; Finney 2007), based on main paths and
predicted fre spread schedules.
• CPSL (Campbell Prediction System Language) analysis, helps to predict where
the fre behaviour will change (Campbell 1995) based on factor alignments
(slope, aspect – fuel temperature, wind,).
• Fire scenarios: large wildfre spread patterns that occur in a given forest massif,
and ‘provinces of fre’ (with homogeneous fre behaviour) (Castellnou 1996).
In the same way, Bahro et al. (2007) proposed a methodology for planning
based on a “problem-fre”, in a “fre shed” and homogeneous fre behaviour.
There are several forecasting tools for fire behaviour (i.e. Anderson 1983;
Andrews 1986; Rothermel 1991; Finney 1998), and the investments in models
and simulations are impressive. However wildfires are very complex events,
especially those with extreme freline intensities. Processes like the interaction
between different fre fronts or the changes in oxygen supply following changes
in the convection column (by means of instability of the atmosphere) are very
diffcult to simulate but are essential for predicting fre behaviour (Dupuy 2009;
Mazzoleni et al. 2008). Furthermore, the available data is scarce, mainly because of
the complexity of the wind-relief interactions.
For these reasons, it is important not only to create and to improve simulation
models (Vega et al. 2009; Dupuy 2009; Mazzoleni et al. 2008), but to put emphasis
on the investigation of historical large wildfres. Based on this, we recommend a
methodology to anticipate the local spread paths followed by large fres moving in
a given fre scenario (Castellnou 2000; Bahro et al. 2007), to anticipate as much as
Wildfre Scenarios: Learning from Experience 123
possible the suppression opportunities, and to anticipate extreme weather conditions
associated with extreme fre behaviour in each scenario.
By understanding the schedules of fire spread we can now determine which
areas are inaccessible for suppression and which ‘zones’ present opportunities for
successful fre fghting. These strategic zones are:
• Critical areas to be managed: where the suppression resources can have an
opportunity to reduce fre intensity during the day of a large wildfre. These
areas are in the main axe of the local spread path, areas associated with large
fre potential. To be safe and useful for suppression attack it should be prepared
before, and fre prevention should locate the needed infrastructures (Rigolot et
al. 2009).
• Areas to reduce fuel combustibility: not directly related to the main local
spread path, but forest management should be directed to reduce the potential
freline intensity. This cannot be seen as a genuine opportunity.
The methodology to determine strategic zones is based on the development of the
following:
• Research and study of historical large wildfres in an area. The fre reports
should describe fre behaviour, based on the analyses of past fre spread
(propagation) and CPSL (Campbell Prediction System Language, Campbell
1995) to determine the main paths followed by the fre and its trigger points
1
.
Those are classifed for given fre types, which are based on repeated,
observations of paths of spread following the topography (Castellnou et al.
2009). The objective of these reports should include capturing the lessons
learned from these fres.
• Determine the fre scenarios in zones with homogeneous fre regime:
- Identify the schedules of fre spread which are repeated in every zone with a
homogeneous fre regime.
- Study the probability that the determined scenario of opportunities would be
useful.
3.6.2 Research of former wildfres and related synoptic situations
Predicting fre behaviour and fre spread patterns requires going back to the records
of wildfires that have occurred in the past. This process requires a structured
standard (fre types) which allows the relevant information to be identifed and
highlighted, and which characterizes the wildfres; this information usually becomes
part of the fire fighters’ knowledge. Every wildfire can be related to a certain
meteorological condition and a synoptic situation of the parameters which defne the
fre (wind intensity, type of wind, and relative humidity during day and night); these
determine fre spread pattern and fre behaviour. Vegetation structure and drought
1 Trigger point: site in which fre spread and freline intensity will be beyond the holding capacity of the suppression
forces in direct attack. The implication is that the fre spread potential increases notably by not identifying such sites.
124 Towards Integrated Fire Management
affect the quantity and distribution of fuel, and therefore also fre behaviour. It is
important to link the former wildfres to the drought level at that time, to make it
feasible to determine possible return-period/recurrence-interval.
For every fre scenario, the most relevant historical fres are analyzed, focusing on
the following aspects:
• Main conclusions. The transfer of useful information and the lessons which
are learned from the fre are the main objectives of these reports. Lessons
with respect to the organization of the suppression operations applied, the
implementation of selected tactics, or lessons related to safety are some likely
products.
• General information about the fre: Date, hour, size, general behaviour and
existing resources. This includes patterns of spread and fre type. (Castellnou et
al. 2009).
• Meteorological conditions at a synoptic scale (Castellnou 1996; Millan et al.
1998; Montserrat 1998) and, if available, at local scale.
• Location of the fre with respect to the topography. For analyzing the local
winds, the interaction between wind and topography, and the infuence of
topography by itself on fre behaviour.
• Used strategy and discussion of the performed tactics and operations.
3.6.3 Fire scenarios
In Europe, the problem of huge wildfres is changing. The developments of different
land use, in combination with the socio-economic changes during the last century,
have infuenced the landscape and the load and distribution of fuels. The development
of the territorial scenario has caused changes in the scientifc world, in operational
actions and in politics. Hence, we do not only transfer data from fire perimeter
simulators like Farsite (Finney 1998) to trajectory simulators which draw spread
path schedules and identify points where fre fronts multiply (Finney et al. 2002).
Nowadays there is also a change at the operational site, from temporary workers
during the 1960s to highly specialized suppression crews like GRAF (Catalonia) or
GAUF (Portugal). These crews have a solid training background (not only in using
suppression tools, but also in anticipating fre management opportunities), operate in
extended attack and most of the personal is involved full time. In addition, more and
more regions and countries integrate new tools of fre prevention in their systems, like
prescribed burning and new regulations with respect to wildfres (Herrero et al. 2009).
The contrast between wildfires with large perimeters (large areas) and very
complex wildfres with crown fres, wildland urban interface problem areas and
simultaneous fires, requires a shift from few resources exploiting the existing
opportunities, into numerous suppression troops which, unfortunately, also only
have a small number of real opportunities. This has caused a change:
- from prevention based on infrastructures and water points to facilitate
the creation of anchoring points, and roads and observation posts for fast
suppression;
Wildfre Scenarios: Learning from Experience 125
Box 1. Examples of large wildfre reports and their use in simulation.
When working with reports of large wildfres, it is necessary to identify those in which
the most important lessons were learned, and those which faced major problems in fre
suppression
An example of the extracted information from former wildfre reports can be seen here:
• In Casarabonela, 2009 (Andalucía, Spain) the main fre potential was limited by
the presence of agricultural areas, which act as a fuel break making suppression
actions easier (Ruíz and Senra 2009).
Another way to explain the lessons learned from past fires is by the use of video
and documentaries. These are very educational for civilian people to help them
understanding how a wildfre behaves (e.g. Binggeli 2009). The videos can also be used
as for a professional public as an education tool about the use of fre.
Figure 1. Picture from the Epuyén-report (Argentina, 1979). When the fre crossed the
area of the main river, the direction and speed of fre spread changed drastically due to
the slope, and the wind increase on the 3
rd
day (Sagarzazu Ansaldo and Defossé 2009).
Figure 2. Fire behaviour in Sant
Llorenç Savall 2003 (Catalonia, Spain).
This was the first significant wildfire
in terms of size, duration and social
importance since the formation of
the specialized crews for wildfires in
Catalonia (GRAF). These crews have a
large capacity to analyze fre behaviour
and to identify local operations and
possibilities. Furthermore they assisted
in coordinating the various different
volunteers following the strategy of the
fre fghters (GRAF 2009).
126 Towards Integrated Fire Management
- to prevention based on the strategic opportunities of large wildfres; adapting
the land use and forest management to wildfres by reducing the vulnerability
of forest stands.
There is a signifcant change: from the objective of eliminating fre from ecosystems,
an adaptation to fre, using the opportunities provided by fre. With this purpose it
is necessary to understand very well how fre behaves, and especially how large
wildfires develop. Large wildfires are not very selective concerning fuels, and
propagate in almost every type of fuel, including agricultural areas or decorative
vegetation (Bajocco and Ricotta 2007; Rigolot et al. 2009). We have also seen
examples of this type of destructive behaviour in Spain, Portugal, the south of France
and Greece in 1994, 2003, 2007 and 2009. These fres do not respond to small details,
but to larger landscape-scale features. The macro-topography (with synoptic situations
associated to large wildfres) is the main factor which controls fre propagation in
complex terrains (Taylor and Skinner 2003; Expósito and Cordero 2004; Iniguez et
al. 2008). This is based on the repetitive observations of fre spread under similar
topography and meteorological conditions, where subsequently the underlying pattern
can be seen (Castellnou 1996; Expósito and Cordero 2004), only differing in intensity
and therefore the distance which the fre reaches, depending on the available fuel.
The existence of repeated fire spread patterns linked to a specific territory
provides an opportunity to identify a ‘problem fre’ and to set up the proper planning
(Castellnou 2000; Bahro et al. 2007). From the former wildfres, a series of fre
scenarios in different synoptic and topographical situations are determined. The
history of wildfres is essential, connecting the world of suppression to the world of
prevention.
Wildfre type characterisation for homogeneous areas of risk
From the three basic patterns of fre spread (topographical fres, wind-driven or
plume-dominated fres) several specifc and repeated aspects are determined for
every pattern. Common ‘fire types’ are proposed by Castellnou et al. (2009),
based on the study of the wildfre history of Catalonia, from the characterization
of coincidences of synoptic and topographic conditions with repeated fre path
schedules. This fre types are a useful way to extract information from the main
observers of fre behaviour, the fre fghters.
For a given scenario the frst step is to identify areas with a common fre type.
Nevertheless on the same site, there may be more than one type of wildfre.
Identifying fre-design and strategic zones
Identifying where the fre behaviour will change, to the advantage or disadvantage
for the suppression crews (Campbell 1995) and predicting how fre front will sub-
divide into several directions (Finney et al. 2002) will assist in understanding fre
spread of the large wildfres (Castellnou et al. 2009). Based on these fre spread
patterns and the associated fre potential, the pre-suppression actions and prevention
needs are able to be effectively planned.
The planning of the use of terrain features is not done based on some general
and abstract fre types, but each fre type and its associated opportunities should
be applied to a specifc territory. Therefore well documented and relevant wildfre
history for the region should be studied. Fire behaviour and spread pattern of
Wildfre Scenarios: Learning from Experience 127
these fres are analyzed, as well as the operations conducted. In these past fres,
different theoretical opportunities considering relevant fre types for the zone are
studied, as well as real opportunities for suppression. We can then consider if these
opportunities could be used or not in this terrain. This will lead to a development
of common schedules, similar scenarios and repeated windows of opportunity for
each wildfre type. With this information, particular properties of the fre spread are
defned for the specifc study area, related to the topography, typical development of
the synoptic episodes, and properties of vegetation types.
By specifying the fre types adjusted to the specifc properties of the site, we end-
up with a concept of ‘wildfre-design’. The fre-design represents a ‘best-reference
fre’ or a fre with the ability to become a large wildfre in a specifc environment
(Castellnou et al. 2009). The fre-design is a tool to understand how large wildfres
perform in a certain territory, and to predict how it spreads, also taking notes on its
weak points. It is a way to provide information and provides criteria to locate the
infrastructures which will support fre suppression actions, and will result in safer and
more effcient suppression actions. The use of simulations permits confrmation and
adjustment of several parameters which are used for the planning of strategic zones:
• map inaccessible areas for the suppression crews;
• confrm the location of the opportunities, associate to rates of spread and
intensities safe and effcient to work;
• determine sites with different vegetation treatments, to establish safety zones
from where suppression operations can be applied;
• select the best sites where fuel reduction methods should be applied to reduce
fre spread potential;
• recognize when resource allocation will not result in a positive outcome.
Taking the information collected of the fire-design as a reference point, and
gathering the experiences of the validated opportunities of historical wildfres,
allows the identification of strategic sites for each environment. Each of these
strategic sites is associated with a fre potential area, which will be one of the criteria
to select the best sites where to invest resources.
3.6.4 Conclusions and recommendations
The most useful tool to stop wildfires – which escape from the conventional
suppression methods (mainly because of high fame lengths and fast rate of spread),
is the anticipation of fre trajectories (fre spread schedule in time and space) in the
near future, i.e. by identifying the type of each particular wildfre.
This anticipation is not only based on insight gained by people or on mathematical
models adjusted to real interaction between fre and a landscape, but on information
extracted from past fires. This type of understanding (based on the underlying
mechanisms) is obtained from detailed analyses of former wildfres. By gathering
this information, we create databases of wildfres which will describe the shape of
fre spread patterns in a particular synoptic (fre weather) and elevation (topography)
framework. This will then serve to make sound decisions in fire fighting, and
128 Towards Integrated Fire Management
Box 2. Instructions to integrate fre in forest management.
This box describes the fire types and related opportunities; with similar fire paths
schedules with analogue synoptic situations and macro topographical conditions
(Castellnou et al. 2009).
The properties of each fre type in a given scenario due to local topography, to the
development of the synoptic situations in the area or to vegetation characteristics, are all
captured in the fre-design.
Figure 3. Perimeters of fres in 1949, 1970, and 2000. The three fres show similar
spread patterns with a synoptic situation of northern wind. Location: Albiol, Catalonia,
northeast Iberian Peninsula. Fire spreads following ridges until the end, and open to new
ridges if available (for example, as the ignition in 1979 was more to the east than the
other two, a second ridge was available and burned).
Figure 4. Propagation by lee winds. The counter directive winds reach the higher part of
the ridge where the main wind occurs, spotting fre downhill. In this way, new fre-runs
develop uphill following the lee winds.
Acknowledging and understanding the fre paths in a lee wind area, provides possibilities
to identify the suppression potential at the end of a ridge or where the ridge drastically
changes direction.
Spot fres, blown
downhill by the
main wind fow
Wind direction
Uphill backwind
Wildfre Scenarios: Learning from Experience 129
Box 2. Continued.
Figure 5. Perimeters
of the fre in the main
valley of the River
Segre, Catalonia.
Figure 6. Topographical fire spread
pattern near a canyon.
will provide possibilities to plan the pre-attack needs and extinction resources
accordingly. In this way we can search for minor actions (CSM – Critical Sites to be
Managed) which will reduce large wildfre spreads.
Therefore, studying the repeated local spread path of past wildfires allows
managing a fre scenario, creating discontinuities of vegetation disrupting these
paths and managing vegetation to manage the whole scenario.
One single methodology of gathering general data for the whole European Union
should cover the complete range of different fre types, and increase the number
and accuracy of fire-designs. A platform of exchanges of information for the
whole Europe is needed. Using one common language in the characterization of
wildfres will simplify the know-how transfer of specialists/analysts between the
regions’ fre fghting projects and between the states of the European Union. With
respect to wildfre prevention, this exchange of knowledge allows the creation of a
130 Towards Integrated Fire Management
common working guide. This could help to implement suppression models and their
requirements, while establishing an assortment of general opportunities for every
fre type in all the regions.
With this methodology it is also possible to highlight the different problems
which the suppression crews (in different regions) may face. It is possible to search
for common/shared solutions within the mix of measures that every suppression
system can provide. Thus we can clearly map the problems, and investigate different
solutions.
Box 3. Wildfre Scenarios – l’Empordà, Catalonia, Spain.
The study of large wildfires (those in the country but also nearby) resulted in the
identifcation and confrmation of the fre scenarios which affected each area. From a
database of 46 perimeters representing 67 471 ha, we can conclude that the Cap de Creus
(the little peninsula in the northwest part of the picture) burns every 26 years mostly by
wind-driven fres (also with topographical constraints).
The validation of former wildfire control opportunities permits us to extrapolate
the lessons learned to each new wildfre scenario. Then, the CSM (Critical Sites to be
Managed) can be determined. Every CSM is linked to a potential, planned burning
surface that can be protected.
Figure 7. Main CSM as
black ellipses, to protect
against eventual 5000 ha
wildfres.
Wildfre Scenarios: Learning from Experience 131
In this manner it is possible to work on preventing the worst possible wildfres in
these zones. Besides, we will know beforehand what the critical situations will be,
which can occur simultaneously (or within a short period of time) in different parts
of Europe. This was the case within the synoptic pattern at the end of July 2009
when there were large wildfres occurring at the same time in different parts of Spain
(Castilla La Mancha, Aragón and Catalonia) and one day later in Sardinia (Italy).
The study of present and past large wildfres is essential for integrating forest
management and suppression actions where both have the same objective: the
persistence of structures and species as far as possible. We should focus on the larger
wildfres, because for the smaller fres the present suppression methods are adequate.
This will allow to plan several actions like pre-attack (locate the places where the
fre fghters could have suppression opportunities) and prevention (locate the places
where adequate forest management could reduce fre intensity, and by this creating
opportunities for the likelihood of successful suppression). This will also assist in
deciding where to make use of agricultural breaks, and to determine where houses
are most vulnerable. It helps also to reinforce or favor vegetation structures which
are more resistant in areas of full alignment (i.e. worst fre behaviour). In the same
way as it is possible to place structures which are less vulnerable to crown fres in
trigger points, so, the study of wildfre scenarios is also an integral tool for forest
management.
Nevertheless, it is important to remember that determining where to act
and increasing the planned activities is necessary but not sufficient to limit
the destructive effects of the large wildfres. The integrated actions will then be
completed by an increase in the most effective forest management options in which
the risk of wildfres is explicitly incorporated, with expanding forest tracks which
facilitate establishment and maintainance of forest structures which are more
tolerant to the fre spread and related damage. This will also reduce the capacity for
sustaining crown fres because of the more discontinuous nature of the vegetation,
which will assist fre suppression in general.
References
Anderson, H.E. 1983. Predicting wind-driven wild land fre size and shape. Intermountain
Forest and Range Experiment Station, Ogden, UT, USA. 26 p.
Andrews, P.L. 1986. Behave: fre behavior prediction and fuel modeling system - BURN
subsystem Part 1. USDA Forest Service, Intermountain Research Station, Ogden, UT,
USA. 130 p.
Bahro, B., Barber, K.H., Sherlock, J.W., and Yasuda, D.A. 2007. Stewardship and freshed
assessment: a process for designing a landscape fuel treatment strategy. Restoring fre-
adapted ecosystems. Proceedings of the 2005 national silviculture workshop, Gen. Tech.
Rep PSW-GTR-203. Pp. 41–54.
Bajocco, S. and Ricotta, C. 2007. Evidence of selective burning in Sardinia (Italy): which
land-cover classes do wildfres prefer? Landscape Ecology 23: 241–248.
Binggeli, F. 2009. Linking bulletin and video reports for the website. Internal Report
IR8.5-1 of the Integrated Project “Fire Paradox”, Project no. FP6-018505. European
Commission. 6 p.
132 Towards Integrated Fire Management
Campbell, D. (ed.) 1995. The Campbell Prediction System: A Wild Land Fire Prediction
System & Language. 129 p.
Castellnou, M. 1996. Reconstrucción de la progresión de los incendios históricos y su
efecto modelador en la vegetación: Ribera de Ebro. In: Gonzalez, J.M. (ed.). Seminario
sobre incendios forestales. Centre Tecnològic Forestal de Catalunya. Solsona, Spain. Pp.
207–223.
Castellnou, M. 2000. Nuevas Metodologías de Prevención de Grandes Incendios. Congreso
Forestal Ibérico. Castelo Branco, Portugal.
Castellnou, M., Miralles, M., Pages, J. and Pique, M. 2009 Tipifcación de los incendios
forestales de Cataluña. Elaboración del mapa de incendios de diseño como herramienta
para la gestión forestal. Proceedings of the 5th Congreso Forestal Español. Ávila, Spain.
Dupuy, J.L. 2009. Fire start and spread. In: Birot, Y. (ed.) Living with wildfres, what
science can tell us. EFI Discussion Paper 15. European Forest Institute. Pp. 27–31.
Expósito, R. and Cordero, T. 2004. Albiol fre. Did we learn at the third time? In: Viegas,
D.X. (ed.). Proceedings of the IV International Conference on Forest Fire Research.
Coimbra, Portugal.
Finney, M.A. 1998. FARSITE: Fire Area Simulator - Model development and evaluation.
RMRS-RP-4. USDA Forest Service Rocky Mountain Forest and Range Experiment
Station, Fort Collins, CO, USA. 47 p.
Finney, M.A., Sapsis, D.B. and Bahro, B. 2002. Use of FARSITE for simulating fre
suppression and analyzing fuel treatment economics. Sugihara, N.G., Morales, M.E.,
Morales, T.J. (eds.). Proceedings of the symposium: Fire in California ecosystems:
integrating ecology, prevention and management. San Diego, CA, USA. Pp. 121–136.
Finney, M.A. 2007. A computational method for optimizing fuel treatment locations.
International Journal of Wildland Fire. 16: 702–711.
GRAF 2009. The Sant Llorenç Savall Forest Fire – Operational Report. Generalitat de
Catalunya. 23 p.
Herrero, G., Montiel, C., Agudo, J. and Aguilar S. 2009 Assessment document on the main
strengths and weaknesses of the legislation and policy instruments concerning integrated
wildland fre management in EU, in European Member States and in North African
countries. Deliverable D7.1-1-2 of the Integrated Project “Fire Paradox”, Project no. FP6-
018505. European Commission.117 p.
Iniguez, J.M., Swetnam, T.W. and Stephen, R.Y. 2008. Topography affected landscape
fre history patterns in southern Arizona, USA. Forest Ecology and Management 256:
295–303.
Mazzoleni, S., Giannino, F., Beccarisi, L. and Ricotta C. 2008. Comparison of
contamination models. Deliverable D6.2-3 of the Integrated Project “Fire Paradox”,
Project no. FP6-018505. European Commission. 43 p.
Millan, M., Estrela, M.J. and Badenas, C. 1998. Meteorological processes relevant to forest
FIRE dynamics on the Spanish Mediterranean COAST. Journal of Applied Meteorology.
37: 83–100.
Montserrat, D. 1998. La predicció dels Grans Incendis Forestals. Catalunya Rural i Agrària.
54: 5–13.
Rifà, A. and Castellnou, M. 2007. El modelo de extinción de incendios forestales catalan.
Proceedings of the IV International Wildfre Conference. Seville. Spain.
Rigolot, E., Fernandes, P. and Rego, F. 2009. Managing Wildfre Risk, Prevention,
Suppression. In: Birot, Y. (ed.) Living with wildfres, what science can tell us. EFI
Discussion Paper 15. European Forest Institute. Pp. 49–52.
Rothermel, R.C. 1991. Predicting behaviour and size of crown fres in the Northern Rocky
Mountains. Res. Pap. INT-438. USDA Forest Service, Intermountain Research Station,
Ogden, UT, USA. 46 p.
Wildfre Scenarios: Learning from Experience 133
Ruíz C. and Senra F. 2009. The Casarabonela 2009/07/14 Fire - Operational Report. Junta de
Anadalucia. 15 p.
Sagarzazu Ansaldo, M.S. and Defossé, G.E. 2009. Study of a 2
nd
set of largest past fres.
Deliverable D8.3-1 of the Integrated Project “Fire Paradox”, Project no. FP6-018505.
European Commission. 82 p.
Taylor, A.H. and Skinner, C.N. 2003. Spatial patterns and controls on historical fre regimes
and forest structure in the Klamath Mountains. Ecological Applications. 13: 704–709.
Vega, J.A., Jiménez, E. and Pérez, J.R. 2009. Effectiveness of suppression fres: methods
and frst results. Internal Report IR2.3-2 of the Integrated Project “Fire Paradox”, Project
no. FP6-018505. European Commission. 27 p.
4. Prescribed Burning:
Preventing Wildfres and More
4.1 Overview of Prescribed Burning Policies and
Practices in Europe and Other Countries
Andrea Lázaro and Cristina Montiel
Research Group Forest Policy and Socioeconomics, University Complutense of
Madrid, Spain
4.1.1 Introduction
Prescribed burning is a technique that represents the controlled application of fre
to vegetation under specifc environmental conditions to attain planned resource
management objectives (FAO 1986). Twenty years after the first experimental
initiatives of prescribed burning in Europe, launched in southern countries in the
1980s (Botelho and Fernandes 1997), the use of fre for different management
purposes remains rather limited. The main constraints for prescribed burning
introduction and implementation in the continent are fire bans, complex land
structure, lack of professional experience and negative public perception
(Xanthopoulos et al. 2006). However, important changes have taken place for
prescribed burning in the last years, especially in those countries which were
pioneers in its introduction, and mainly for wildfre hazard reduction objectives.
Portugal and France were the frst European countries to initiate policy processes
aiming to develop prescribed burning practices, based on specifc legal frameworks
and national systems for the professional accreditation in the use of this technique.
Spain has also progressed in this direction, but mainly at the regional level. In fact,
Catalonia and the Canary Islands have created professional teams in the use of fre
for management purposes. Moreover, objectives for the use of this technique have
broadened towards more environmental oriented purposes as more experience has
been accumulated both in prescribed burning research and management felds. Thus,
prescribed burning has been progressively included in nature conservation policy
documents (Rigolot 2005).
In Northern European countries (e.g. Sweden and Finland) the use of prescribed
burning is at present included as a management practice required by forest
certifcation schemes and also for nature conservation of protected areas. Moreover,
although still maintained at the experimental level, changing paradigms in nature
and landscape policies in Atlantic and Central European countries have propitiated
the development of prescribed burning initiatives in the frame of research projects
(Goldammer et al. 2007).
Thus, the introduction and implementation of prescribed burning practices with
different management objectives is a progressive trend in fre and land management
policies, affecting even the more reluctant countries. The support of Fire Paradox
project and the infuence of the exchanges between fre professionals of Europe
138 Towards Integrated Fire Management
and other continents have allowed the barrier against the use of prescribed burning
to be overcome in countries which traditionally had maintained a negative attitude
towards the use of prescribed burning. The most prominent example in this sense
is the development of a professional fire use team in Sardinia (2008) and the
development of the frst prescribed burning initiatives for fre prevention in the
Cilento National Park, southern Italy (2009).
The research on prescribed burning practices and policies carried out within
the frame of Fire Paradox has enabled a general overview of prescribed burning
spatial patterns at the EU level to be obtained, identifcation of the different existing
policy scenarios in European countries towards prescribed burning practices,
and identifcation of the main factors that are key to promote these practices as
well as to provide recommendations to overcome the principal constraints for the
future. Besides, the research has considered neighboring countries in the southern
Mediterranean basin by acquiring a general sense of their fre uses, and where the
possibility to use prescribed burning for fre hazard reduction has similarly aroused
technical and political interest.
4.1.2 Inventory and classifcation of fre use practices
The available information about the development and distribution of fire use
practices in Europe is scattered and rather scarce, with a few exceptions focusing
on some European regions (Pyne 1997; Botelho and Fernandes 1997; Goldammer
et al. 2007). Thus, the frst aim of fre use practices assessment within Fire Paradox
project was to obtain a complete summary of spatial distribution of traditional
and prescribed burning in Europe and North African countries. At this point, we
fnd it relevant to include also traditional fre use practices since it is considered
as an important factor for the development of prescribed burning policies in the
future, that is either regulating traditional fre use and profting from this experience
to develop prescribed burning policies, or by (re)introducing prescribed burning
as a modern technique for management purposes within a professional structure
(see Box 1). However when considering both types of practices, it is important to
clearly differentiate between traditional burning practices and prescribed burning
practices. In this regard the main characteristic that clearly distinguishes prescribed
burning from traditional fre use is an adequate planning (Pyne et al. 1996) as well
as the latter evaluation which determines whether the pre-determined management
objectives have been reached and allow for future improvements (Fernandes 2002).
For this purpose, information was collected, classified and referenced at the
regional level for the EU-27 member states and the North African countries
(Morocco and Tunisia), based on a questionnaire distributed to national experts with
responsibilities in forest and fre management
1
.
1 For more information on sources and methods see Lázaro et al. (2008).
Overview of Prescribed Burning Policies and Practices in Europe and Other Countries 139
4.1.3 Antecedents for prescribed burning practices: traditional fre use
As in other continents around the world, anthropogenic fre has been recognized
as one of the most significant alterations to fire regimes in Europe (Conedera
et al. 2008; Scott et al. 2000). This has been the case especially since Neolithic
times where fire became an essential tool to expand agricultural lands. In the
Mediterranean region, the pastoral use of fre has continued throughout history until
present (Di Pasquale et al. 2004) and together with grazing and other traditional
cultivation systems have become part of integral components which have shaped
Mediterranean landscapes (Naveh 1975). In countries from temperate and
boreal Europe, fre became an essential tool for the expansion and occupation of
new lands). The most representative techniques were the use of fre in swidden
agriculture and shifting cultivation systems (Pyne 1997; Goldammer 1998;
Bradshaw and Hannon 2006). Also on lands that were too poor for agriculture the
burning of grasslands for pasture was common. In particular the heather and moor
landscapes characteristic in Atlantic countries such as Belgium, Germany, France,
Norway and United Kingdom are product of more than 5000 years of burning by
rural communities (Haaland 2003; Dodgshon and Olsson 2006).
The current situation in the maintenance of the traditional practices of burning in
Europe consists on a general abandonment in northern countries and central Europe
while it is still a widespread practice in Mediterranean countries where there can be
a wide range of management objectives (e.g. grazing improvement, burning of agro-
forestry remains, hunting, etc.) (see Figure 1).
In southern and eastern European countries, traditional burning practices have
been maintained and in some rural areas the use of this technique still constitutes
the most effective and economic tool for stubble burning and grazing improvement
(Metailié 2006; Velez 2005; Merou and Papanastasis 2002), However changing
spatial and socio-economic conditions in these areas (e.g. vegetation encroachment,
aging of rural population, etc.) have increased the risk of traditional fire use
practices which have evolved to be at present the main cause for wildfire in
Mediterranean countries (FAO 2007; JRC-IES 2008). In countries from the southern
Mediterranean basin like Morocco, traditional burning practices are maintained
although fre has not being used in their grazing management but for other purposes
(e.g. wars between tribes, wildlife management, forest harvesting).
In addition, it is important to highlight the maintenance of traditional burning
outside the Mediterranean, as in the case of the British heathlands (Davies et al.
2008). Heathlands dominated by Calluna vulgaris species are frequently burned
by shepherds to develop a mosaic habitat favorable for the red grouse (Lagopus
lagopus), sheep, deer and improvement of grasses (SEERAD 2001, DEFRA
2007). Similarly, traditional burning methods have been recently re-established in
Atlantic heathlands in Norway (Haaland 2003) and in the Black Forest in Germany
(Goldammer et al. 2007) (see Box 1).
140 Towards Integrated Fire Management
4.1.4 Development of Prescribed Burning practices in Europe
With regard to prescribed burning practices, a general pattern can be differentiated
with regard to its management objectives (see Figure 2). Since its frst introduction
in the 1980s fre hazard reduction has been the predominant goal of prescribed
burning in southern European countries. However, as more experience has been
accumulated, the objectives of its application have broadened towards other
management targets (e.g. nature conservation, forest or game management). This
is the case of France, Portugal and some Spanish regions like Catalonia, Galicia
or the Canary Islands which during the last decade have developed professional
prescribed burning teams, specifc training for prescribed burning specialists, expert
networks, and first legal instruments to regulate this practice. In the last years
Italy has progressed in the development of prescribed burning promoted through
exchanges between professionals from Portugal. Results from this progress have
been the initiation of the frst prescribed burning initiatives in the Cilento National
Park (Campania region) included in the Forest Fire Defense Plan (2008) with the
objective to manage fuel accumulation under pinewoods (Pinus halepensis and
Pinus pinaster), as well as with the creation of professional teams in the use of fre
in Sardinia in 2008 (Mastros do Fogu).
Figure 1. Distribution of traditional burning practices in Europe and North African countries.
Overview of Prescribed Burning Policies and Practices in Europe and Other Countries 141
Box 1. The re-establishment of traditional burning practices: new
perspectives
The recognition of the role that traditional management techniques have had for the
maintenance of many European landscapes and ecosystems (Pyne 1997; Goldammer
1998; Goldammer et al. 2007) as well as its contribution as a management scheme
for the reduction of forest fuel accumulation have contributed in recent years to the
development of programs aimed at the support of traditional burning practices.
In those areas where these type of practices had been lost, mostly in Northern
countries as well as in Central European and Atlantic countries, the re-establishment
of traditional burning systems are valuable not only to the cultural heritage values but
also the role that these management practices had in the conservation and maintenance
of valuable ecosystems related to them. Some examples of this type of initiative are
found in the Black Forest (Germany) and in the Koli National Park (Finland) (Loven
and Äänismaa 2004). Also in Norway, in the Heathland Centre (Lygra, Bergen) the
traditional management has been continued in cooperation with local farmers in order to
preserve the open heathlands and the culture connected to them (Haaland 2003).
On the other hand, in those places where traditional burning practices have been
maintained, pasture burnings have been subject to support schemes and in some cases
have converged in a modern management technique used for fre hazard reduction or the
management of natural areas, that of prescribed burning. In those cases the traditional
practice is re-established through institutional supports based on historical precedents
(Ribet 2009). However, in practice, the development of PB programs can be carried
out with greater or lesser integration of traditional know-how – from a complete
implementation by professionals, to implementation by professionals with the support of
shepherds, to training of and then implementation by shepherds.
Figure 4. Pasture burning in Sardinia, Italy.
Photo by N. Ribet, 2006.
Figure 3. Dialogue between professionals
and shepherds in Salamanca, Spain.Photo
by I. Juárez, 2006.
In Central European and Atlantic countries the use of prescribed burning is
focused on the management of endangered habitats and the maintenance of open
landscapes. Although the first experiments where initiated in the early 1970s
(Goldammer 1978) the map demonstrates how prescribed burning within this
region is still used at the experimental level, within the framework of demonstration
142 Towards Integrated Fire Management
projects in nature conservation (e.g. LIFE) in Germany (Goldammer et al 1998),
Netherlands (van der Zee 2004), Denmark (Jensen 2004), United Kingdom (Bruce
2002) etc. Some of the most commonly managed landscapes are the Calluna
heathlands and other open landscapes such as grasslands and fen mires. Other
examples include the management endangered species habitats such as for the
black grouse (Tetrao tetrix) and the woodland grouse (Tetrao urogallis). Exceptions
to these experimental fre uses are in Northern European countries, Sweden and
Finland, where the use of prescribed burning is promoted as a sustainable practice
for forest certifcation (FSC Suecia 2006; FSC Finlandia, 2005) and for biodiversity
management in protected areas with promising examples such as the nature reserves
of Jämtgaveln and Stormyran-Lommyran in Sweden and the Koli National Park in
eastern Finland (Goldammer et al. 2007).
New scenarios for the development of prescribed burning practices are being
initiated in other regions. For instance in eastern European countries experimental
trials are taking place for both nature conservation in open landscapes (e.g.
Poland) as well as for its potential application in fre risk reduction use within the
frame of the EU sponsored SEERANET (southeast-European Research Area), in
Figure 2. Distribution of prescribed burning practices with regard to its management
objectives in Europe and North African countries.
Overview of Prescribed Burning Policies and Practices in Europe and Other Countries 143
which Germany, Hungary and the Former Yugoslav Republic of Macedonia are
participating. In North Africa, Morocco began in 2008 to apply prescribed burning
on shrubs located along the freeways for clearing the external and central bands as
an alternative method to the mechanical clearing.
4.1.5 National and regional legal frameworks related to prescribed
burning in Europe
The assessment of national legal and policy frameworks with respect to prescribed
burning practices in Europe and North African
2
countries has identifed that there are
few legal initiatives that exist with regard to prescribed burning, with most examples
restricted to a few Mediterranean examples.
Portugal and France were the frst European countries to develop legal instruments
regulating the use of prescribed burning. France, for the frst time, introduced in the
Forest Management Law (2001), modifying the Forest Code (1987), the possibility
to use fre for prevention purposes by the responsible authorities outside of the risk
period (art. L321-12)
3
. At the regional level, the “Départements” have regulated the
use of prescribed burning through local regulations (arrêtés préfectorales) and by the
development of the corresponding prescribed burning plan (cahier des charges). A
few years later Portugal enacted a specifc regulation on prescribed burning (fogo
controlado) (Portaria 1061/2004), and in 2009 major policy changes have used the
term fogo técnico, including both components: prescribed burning and suppression
fre (fogo de supressão) (Decree-Law No. 17/2009 and Despacho nº143031/2009).
Spain and Italy as decentralized countries have progressed in this feld only at
the regional level. In Spain, the region of Catalonia which has developed a specifc
regulation on this subject, the Decree 312/2006 on Tactical Fire which regulates the
use of fre both in wildfre prevention and suppression duties by the forest and civil
protection services. Other regions like Galicia and Balearic Islands have introduced
this possibility within their basic forest fre regulations. In Italy some regions have
considered the possibility to use this technique in their regional laws and programs
(e.g. Piemonte, Basilicata, and Liguria) but except for Sardinia, little has been
translated into practice.
However the majority of European countries do not have a legal framework for
prescribed burning and this factor constitutes one of the main impediments to the
development of this practice. This is due to the general prohibition of the use of fre
in forest lands and other open landscapes (e.g. grasslands, wetlands, mires) which
on occasions come from different legal acts (e.g. Nature Conservation and Forest
Acts) and explains the need to obtain special permits to be able to implement this
technique.
2 For more information on sources and methods see Montiel et al. (2009).
3 This article applies only to the regions of Provence-Alpes-Côte d’Azur, Languedoc-Rousillon, Corse, Midi-Pyrénées,
Aquitaine, Poitou-Charentes and the departments of L’Alderche and Drôme.
144 Towards Integrated Fire Management
In Central European and Atlantic countries fre bans have arisen very often from
new air quality standards and the widespread opinion from the authorities of the
detrimental effect that fre had on the stability and biodiversity of the ecosystems
(Goldammer et al., 1998). For instance, in the late 1970s Germany imposed a total
fre ban in open spaces through the Federal Law of Nature Conservation. It provided
the legal framework legislation to ban the application of these practices by the
Landers which enacted respective laws in the subsequent years (Goldammer et al.
2007). In Belgium there is a general fre use prohibition. It is prohibited in forest
areas, within a distance of 100 m from forest edges (art. 167. Forest Code 1854-
2003) as well as in nature reserves (art. 11, Nature Protection Act 1973).
For eastern European countries prohibitions might have been either derived
from the examples of their neighboring countries, either because of the increasing
perception of fre risk derived from the spatial and socio-economic changes that
have taken place in the region during the last decades (FAO 2007; JRC-IES 2008).
For example, in Poland different legal acts have meant that there has been a total ban
in force during the last three decades: the Forestry Law (1991), prohibited the use
of fre outside the designated areas in “forests and in areas of grass, peat bogs and
heather moor, as well as in a perimeter of 100 m of the forest zone”; and the Law
on Protection of Nature (2004) prohibited the use of fre in “meadow vegetation,
pasture grounds, waste land, ditches, along roads, train trails, or in the zones of
bulrush and reed”. In Hungary, the Law of Nature Protection (1996) prohibited the
use of fre in open spaces. The Forest Act (1996) specifes that it is forbidden to
light any kind of fre in forest areas as well as within a distance of 200 m from the
edge of a forest area, except for the forest managers or any person who has a written
permission (art. 53).
Other countries just do not include provisions for fre use in their main legislation
texts. In Denmark, neither the Forest Act (2004) nor the Nature Protection Act
(1992) includes provisions for the use of fre. This is the same situation for national
legislation in Switzerland in both the Forest Law (1991) and Nature Protection and
Landscape Act (1966). In Scandinavian countries (Sweden, Norway and Finland) the
use of fre is not referenced within the main legislative instruments at national level.
However there are also exceptions to this restrictive general framework. The most
prominent case is the regulation of the practice of burning heather and grasses in the
United Kingdom which is subject to regulations and codes of practice in Scotland
(SEERAD 2001), England (DEFRA 2007) and Wales (by the Welsh Assembly
Government in 2008). With regard to legislation, the Hill Farming Act (1946)
includes provisions to regulate of the practice for the three regions. Additionally,
England (2007) and Wales (2008) have developed Heather and Grass Regulations.
Other exceptions are those acquired at the regional level in administrative
decentralized or decentralized countries. For instance in Germany, although most
states followed the general fre ban of the Federal Government two exceptions have
regulated the use of fre: Hessen State and Baden-Württemberg State. In the case
of Switzerland, exceptions for the use of fre in forest areas and its perimeter (e.g.
10 m) are also contemplated at the regional level in western Switzerland (French
cantons) and the Ticino region (Italian canton).
Overview of Prescribed Burning Policies and Practices in Europe and Other Countries 145
4.1.6 Strengths and weaknesses of prescribed burning practices in
Europe
The assessment of national and regional legal frameworks and policy instruments
shows the main strengths and weaknesses for the introduction and development of
prescribed burning practices in the short to medium term for the main European
regions (See Table 1). It is noteworthy that the use of prescribed burning is supported
on present demands and needs to solve different fre problems. Furthermore, at
present prescribed burning constitutes an opportunity to deal with new problems
that have emerged due to recently changing socio-economic conditions (e.g. new fre
scenarios at the wildland urban interface, loss of traditional landscapes).
Also experience gained in the experimental and management feld, as well as
the creation of network structures joining principal stakeholders involved in the
implementation of prescribed burning practices and policies, both constitute
strengths in those countries with longer experience. On the other hand, these
countries have a valuable potential to export and adapt key factors for success to
other countries where this technique is facing greater limitations. Nonetheless,
even in those countries where prescribed burning practices are well-established,
continued training of personnel and an increased number of qualifed technicians is
needed to consolidate the use of prescribed burning for the future.
On its side, the maintenance or loss of traditional fre use know-how (fre culture)
might constitute a strength or weakness in either way. For instance in those countries
where the fre culture has been maintained, this might constitute a strength since part
of the general public, mainly rural are still familiar with the use of fre. This culture
also constitutes a starting point for forest and civil protection services for the (re)-
introduction of fre as a management tool. However where fre culture is maintained
but not aligned with policy, conficts may arise between the rural communities and the
responsible administrations. The challenge in these areas is to achieve the same level
of recognition for competences in traditional burning practices than in prescribed
burning as well as to assure a real integration of both traditional and prescribed
burning practices in the development of future fre use policies (Ribet 2009).
Finally, other important limitations for prescribed burning practices are the over-
restrictive character of some legal frameworks for fre use in some countries, or the
still non-adapted traditional fre use regulations in other countries.
4.1.7 Conclusions and recommendations
The recent evolution of prescribed burning policies and practices in Europe and
North African countries demonstrates that new opportunities are opening for the
use of fre as a management tool for wildfre hazard reduction as well as for nature
conservation and other objectives. However, the context for fre use differs greatly
between countries, regions and even within the local level. Furthermore, the fre
problem is different in each European region and is evolving within the context of
global change.
146 Towards Integrated Fire Management
T
a
b
l
e

1
.

M
a
i
n

s
t
r
e
n
g
t
h
s

a
n
d

w
e
a
k
n
e
s
s
e
s

o
f

p
r
e
s
c
r
i
b
e
d

b
u
r
n
i
n
g

(
P
B
)

p
r
a
c
t
i
c
e
s

i
n

E
u
r
o
p
e
.


S
t
r
e
n
g
t
h
s

a
n
d

o
p
p
o
r
t
u
n
i
t
i
e
s

W
e
a
k
n
e
s
s
e
s

a
n
d

c
h
a
l
l
e
n
g
e
s
N
o
r
t
h
e
r
n

c
o
u
n
t
r
i
e
s

R
e
c
o
g
n
i
t
i
o
n

o
f

t
h
e

n
a
t
u
r
a
l

r
o
l
e

o
f

f
r
e

i
n

b
o
r
e
a
l

a
n
d

s
u
b
-
b
o
r
e
a
l

P
r
e
d
o
m
i
n
a
n
c
e

o
f

t
h
e

e
n
v
i
r
o
n
m
e
n
t
a
l

a
n
d

t
i
m
b
e
r

p
r
o
d
u
c
t
i
o
n

e
c
o
s
y
s
t
e
m
s
:

p
r
o
m
o
t
i
o
n

o
f

P
B

i
n

t
h
e

f
e
l
d

o
f

F
S
C

a
n
d

i
n

p
r
o
t
e
c
t
e
d

f
u
n
c
t
i
o
n
s

i
n

f
o
r
e
s
t

a
r
e
a
s

a
r
e
a
s



A
b
s
e
n
c
e

o
f

a

r
e
s
t
r
i
c
t
i
v
e

l
e
g
a
l

f
r
a
m
e
w
o
r
k

a
n
d

n
e
g
a
t
i
v
e

p
u
b
l
i
c

L
a
c
k

o
f

a
w
a
r
e
n
e
s
s

f
o
r

f
o
r
e
s
t

f
r
e

r
i
s
k

i
n

t
h
e

r
e
g
i
o
n

p
e
r
c
e
p
t
i
o
n
.


E
x
p
e
r
i
e
n
c
e

f
o
r

P
B

p
l
a
n
n
i
n
g

a
n
d

o
p
e
r
a
t
i
o
n
a
l

i
m
p
l
e
m
e
n
t
a
t
i
o
n

i
n

L
o
s
s

o
f

f
r
e

c
u
l
t
u
r
e

S
w
e
d
e
n
.



E
x
i
s
t
e
n
c
e

o
f

t
h
e

E
u
r
o
p
e
a
n

E
u
r
a
s
i
a
n

F
i
r
e

i
n

N
a
t
u
r
e

C
o
n
s
e
r
v
a
t
i
o
n

N
e
t
w
o
r
k

(
E
F
N
C
N
)

f
o
r

t
e
c
h
n
o
l
o
g
y

t
r
a
n
s
f
e
r

a
n
d

e
n
h
a
n
c
i
n
g

d
i
a
l
o
g
u
e

b
e
t
w
e
e
n

c
o
u
n
t
r
i
e
s

C
e
n
t
r
a
l

E
u
r
o
p
e
a
n


R
e
-
e
v
a
l
u
a
t
i
o
n

o
f

t
h
e

r
o
l
e

o
f

f
r
e

i
n

n
a
t
u
r
e

c
o
n
s
e
r
v
a
t
i
o
n

a
n
d

R
e
s
t
r
i
c
t
i
v
e

l
e
g
a
l

f
r
a
m
e
w
o
r
k

f
o
r

t
h
e

u
s
e

o
f

f
r
e

i
n

o
p
e
n
a
n
d

A
t
l
a
n
t
i
c

c
o
u
n
t
r
i
e
s

l
a
n
d
s
c
a
p
e

p
o
l
i
c
i
e
s

l
a
n
d
s
c
a
p
e
s

a
n
d

l
o
s
s

o
f

f
r
e

c
u
l
t
u
r
e

(
e
x
c
e
p
t

f
o
r

U
K
)

E
x
p
e
r
i
e
n
c
e

a
c
c
u
m
u
l
a
t
e
d

i
n

P
B

r
e
s
e
a
r
c
h

i
n

t
h
e

f
r
a
m
e

o
f

n
a
t
u
r
e

C
u
r
r
e
n
t

d
e
b
a
t
e
s

a
n
d

c
o
n
f
i
c
t
s

r
e
l
a
t
e
d

t
o

t
h
e

u
s
e

o
f

t
h
i
s

c
o
n
s
e
r
v
a
t
i
o
n

a
n
d

l
a
n
d
s
c
a
p
e

m
a
i
n
t
e
n
a
n
c
e
.

t
e
c
h
n
i
q
u
e

a
n
d

i
t
s

i
m
p
a
c
t

(
e
.
g
.

c
a
r
b
o
n

b
u
d
g
e
t
,

w
a
t
e
r

q
u
a
l
i
t
y
)

E
x
i
s
t
e
n
c
e

o
f

t
h
e

E
u
r
o
p
e
a
n

E
u
r
a
s
i
a
n

F
i
r
e

i
n

N
a
t
u
r
e

C
o
n
s
e
r
v
a
t
i
o
n

N
e
t
w
o
r
k

(
E
F
N
C
N
)

f
o
r

t
e
c
h
n
o
l
o
g
y

t
r
a
n
s
f
e
r

a
n
d

e
n
h
a
n
c
i
n
g

d
i
a
l
o
g
u
e

b
e
t
w
e
e
n

c
o
u
n
t
r
i
e
s


Overview of Prescribed Burning Policies and Practices in Europe and Other Countries 147
T
a
b
l
e

1
.

C
o
n
t
i
n
u
e
d
.


S
t
r
e
n
g
t
h
s

a
n
d

o
p
p
o
r
t
u
n
i
t
i
e
s

W
e
a
k
n
e
s
s
e
s

a
n
d

c
h
a
l
l
e
n
g
e
s
E
a
s
t
e
r
n

c
o
u
n
t
r
i
e
s

O
p
p
o
r
t
u
n
i
t
i
e
s

f
o
r

t
h
e

d
e
v
e
l
o
p
m
e
n
t

o
f

P
B

d
e
r
i
v
e
d

f
r
o
m

c
h
a
n
g
i
n
g

R
e
s
t
r
i
c
t
i
v
e

l
e
g
a
l

f
r
a
m
e
w
o
r
k

f
o
r

t
h
e

u
s
e

o
f

f
r
e

i
n

o
p
e
n

s
o
c
i
o
-
e
c
o
n
o
m
i
c

c
o
n
d
i
t
i
o
n
s

(
e
.
g
.

i
n
c
r
e
a
s
e
d

f
r
e

h
a
z
a
r
d

a
n
d

l
o
s
s

o
f

l
a
n
d
s
c
a
p
e
s

o
p
e
n

l
a
n
d
s
c
a
p
e
s

a
n
d

b
i
o
d
i
v
e
r
s
i
t
y
)



M
a
i
n
t
e
n
a
n
c
e

o
f

t
r
a
d
i
t
i
o
n
a
l

f
r
e

c
u
l
t
u
r
e

F
i
r
e

a
s
s
o
c
i
a
t
e
d

w
i
t
h

o
t
h
e
r

e
n
v
i
r
o
n
m
e
n
t
a
l

i
s
s
u
e
s

(
r
a
d
i
o
a
c
t
i
v
i
t
y
,



l
a
n
d

m
i
n
e
s
,

e
t
c
.
)

E
x
i
s
t
e
n
c
e

o
f

t
h
e

E
u
r
o
p
e
a
n

E
u
r
a
s
i
a
n

F
i
r
e

i
n

N
a
t
u
r
e

C
o
n
s
e
r
v
a
t
i
o
n

S
t
r
u
c
t
u
r
a
l

p
r
o
b
l
e
m
s

a
s
s
o
c
i
a
t
e
d

t
o

t
h
e

t
r
a
n
s
f
o
r
m
a
t
i
o
n

t
o
w
a
r
d
s

N
e
t
w
o
r
k

(
E
F
N
C
N
)

f
o
r

t
e
c
h
n
o
l
o
g
y

t
r
a
n
s
f
e
r

a
n
d

e
n
h
a
n
c
i
n
g

d
i
a
l
o
g
u
e

m
a
r
k
e
t

e
c
o
n
o
m
y

b
e
t
w
e
e
n

c
o
u
n
t
r
i
e
s



P
o
o
r

k
n
o
w
l
e
d
g
e

o
f

t
h
e

e
c
o
l
o
g
i
c
a
l

r
o
l
e

o
f

f
r
e
S
o
u
t
h
e
r
n

E
u
r
o
p
e
a
n

P
B

c
o
n
s
t
i
t
u
t
e
s

a
n

e
f
f
c
i
e
n
t
,

e
c
o
n
o
m
i
c

a
n
d

a
d
a
p
t
e
d

t
o
o
l

t
o

s
o
l
v
e

E
f
f
e
c
t
s

o
f

l
o
n
g

t
e
r
m

f
r
e

e
x
c
l
u
s
i
o
n

p
o
l
i
c
y

i
n

p
u
b
l
i
c

p
e
r
c
e
p
t
i
o
n
c
o
u
n
t
r
i
e
s

p
r
e
s
e
n
t

d
e
m
a
n
d
s

(
e
.
g
.

f
u
e
l

b
u
i
l
d
-
u
p

a
n
d

n
e
w

e
n
v
i
r
o
n
m
e
n
t
a
l

c
o
n
c
e
r
n
s
)

a
n
d

r
e
s
p
o
n
s
i
b
l
e

a
d
m
i
n
i
s
t
r
a
t
i
o
n



D
e
v
e
l
o
p
m
e
n
t

o
f

P
B

p
r
o
f
e
s
s
i
o
n
a
l

s
t
r
u
c
t
u
r
e
s

a
n
d

f
r
s
t

r
e
g
u
l
a
t
o
r
y

C
l
a
n
d
e
s
t
i
n
e

a
n
d

n
e
g
l
i
g
e
n
t

f
r
e

u
s
e

p
r
a
c
t
i
c
e
s

w
h
i
c
h

c
o
n
s
t
i
t
u
t
e

f
r
a
m
e
w
o
r
k
s

e
x
p
o
r
t
a
b
l
e

b
e
t
w
e
e
n

c
o
u
n
t
r
i
e
s

w
i
t
h
i
n

t
h
e

r
e
g
i
o
n

a

h
i
g
h

p
e
r
c
e
n
t
a
g
e

o
f

t
h
e

c
a
u
s
e
s

o
f

f
o
r
e
s
t

f
r
e
s



M
a
i
n
t
e
n
a
n
c
e

o
f

t
r
a
d
i
t
i
o
n
a
l

f
r
e

c
u
l
t
u
r
e

T
h
e

c
h
a
l
l
e
n
g
e

t
o

m
a
k
e

t
h
e

p
r
o
c
e
s
s

o
f

d
e
v
e
l
o
p
m
e
n
t

o
f

P
B



p
e
r
d
u
r
a
b
l
e

(
i
.
e
.

n
o
t

d
e
p
e
n
d
i
n
g

o
n

t
h
e

m
o
t
i
v
a
t
i
o
n

o
f

r
e
d
u
c
e
d



g
r
o
u
p

o
f

p
e
r
s
o
n
s
)

148 Towards Integrated Fire Management
It is evident that the increasingly frequent episodes of large fires and the
development of new fire scenarios create a challenging reality which requires
different political approaches and new management tools. Nevertheless, the
political, ecological and socio-economic diversity of the continent requires
adaptable and fexible proposals able to respond to different problems and to adapt
to different contexts.
The Fire Paradox project is potentially capable to induce changes in the EU and
national fre management policies. In this context a proposal of a Fire Framework
Directive (FFD)
4
has been launched to start a shift in fire policies in Europe
towards integrated fre management (IFM) and fre use regulation. The proposed
Framework Directive would be a basic approach or minimum legal harmonization
of Member State legislation to deal more effciently with the fre problem in Europe.
In particular with regards to fre use practices a series of recommendations are
proposed to be included in the FFD in line with the philosophy of the project:
• There is a need to assume that there are European rural communities with a
fre culture which depend on the use of fre for its welfare. This is not only
benefcial for them but also constitutes a fuel management scheme which
reduces fre risk in the community.
• Fire-relying communities need legislative frameworks which contemplate and
regulate their fre use activities as well as reduce un-wanted ignitions through
the development of social prevention programs.
• A best practice approach for the regulation of traditional use of fre should be
promoted instead of the conventional restrictive regulatory approach that has
been adopted by most Mediterranean countries.
• A general legal framework for the use of prescribed burning in those countries
where an over-restrictive framework impedes the development of this
technique and keeps it at the experimental level. Existing prescribed burning
regulations (e.g. France, Portugal and Spain) can serve as reference for its
defnition since they include common elements to all of them. However
member states should adapt their national legal frameworks to their own
particular demands and needs.
• There is a need to promote the recognition of competences for both traditional
fre use and prescribed burning. Traditional practices share the same skills
than professional use of fre and this should be recognized and supported
by organizational assistance, equipment supply and training sessions for
local populations (Community Based Fire Management). With regard to
the professional technique there is a need for the formalization of required
competences through the development of adequate training systems and
certifcations.
4 For more details, see Agudo and Montiel (2009).
Overview of Prescribed Burning Policies and Practices in Europe and Other Countries 149
References
Agudo, J. and Montiel, C. 2009. Basis to start the process for a proposal of a new legislation
at EU level. Deliverable D.7.1-1.3 of the Integrated project “Fire Paradox”. European
Commission. 45 p.
Botelho, H and Fernandes, P.M. 1997. Controlled burning in the Mediterranean countries of
Europe. European Commission; Environment & Climate Programme; Advanced Study
Courses; Marathon; 6–14 October 1997. Pp. 163–170.
Bradshaw, R.H.W. and Hannon, G.E. 2006. Long-term vegetation dynamics in southern
Scandinavia and their use in managing landscapes for biodiversity. In: Anoletti, M. (ed.).
The conservation of Cultural Landscapes, CAB International. Pp. 94–107.
Bruce, M. 2002. Country report for the United Kingdom. International Forest Fire News 27:
68–76.
Conedera, M., Tinner, W., Neff,C., Meurer, M., Dickens, A.F. and Krebs, P. 2008.
Reconstructing past fre regimes: methods, applications, and relevance to fre management
and conservation. Quaternary Science Review 28: 555–576.
Davies, G.M., Gray, A, Hamilton, A. and Legg, C.J. 2008. The future of fre management in
the British uplands, Int. J. Biodivers. Sci. Manag. 4(3): 127–147.
DEFRA 2007. The heather and grass burning code 2007 version. 28 p.
Di Pasquale, G., Di Martino, P. and Mazzoleni, S. 2004. Forest history in the Mediterranean
region. In: Mazzoleni et al. (eds.). Recent Dynamics of the Mediterranean Vegetation and
Landscape. John Wiley and Sons Inc. Chichester, UK. Pp. 13–20.
Dodgshon, R.A. and Olsson, G.A. 2006. Heather moorland in the Scottish Highlands: the
history of a cultural landscape, 1600–1880. Journal of Historical Geography 32: 21-37.
FAO 1986. Wildland Fire Management Terminology. Food and Agriculture Organization of
the United Nations, FAO Forestry Paper 70. Rome. 257 p.
FAO 2007. Fire management – global assessment 2006. FAO Forestry Paper 151. Rome.
156 p.
Fernandes, P. 2002. Prescribed fre: strategies and management. In: Pardini, G., and Pintó,
J. (eds.). Fire, Landscape and Biodiversity: an Appraisal of the Effects and Effectiveness.
Diversitas 29. Universitat de Girona, Institut de Medi Ambient, Girona. Pp. 187–200.
Goldammer, J.G. 1978. Feuerökologie und Feuer-Management. Freiburger Waldschutz Abh.
1 (2). 150 p.
Goldammer, J. G. 1998. History of fre in land-use systems of the Baltic Region:
Implications on the use of prescribed fre in forestry, nature conservation and landscape
management. First Baltic Conference on Forest Fires, Radom-Katowice, Poland.
Goldammer, J. G., Hoffmann G., Bruce, M, Konsrashov L., Verkhovets S., Kisilyakhov Y. K.,
Rydkvist T., Page H., Brunn E., Lovén L., Eerikäinen, K, Nikolov, N. and Chuluunbaatar,
T-O 2007. The Eurasian Fire in Nature Conservation Network (EFNCN): Advances in the
use of prescribed fre in nature conservation, landscape management in temperate-boreal
Europe and adjoining countries in Southeast Europe, Caucasus, Central Asia and Northeast
Asia. 4
th
International Wildland Fire Conference, Seville, Spain. Pp. 13–17.
Haaland, P. 2003. Five thousand years of burning: The European heathlands, 165 p.
Jensen, H.S. 2004. Denmark: Restoration of Dune Habitats along the Danish West Coast.
International Forest Fire News 30: 14–15.
JRC-IES 2008. Forest Fires in Europe. Report nº 9/2009. JRC Scientifc and Technic
Reports. 88 p.
Lázaro, A., Solana, J., Montiel, C., Goldammer, J.G., Kraus, D. and Rigolot, E. 2008.
Collection, classifcation and mapping of the current prescribed fre and suppression
fre practices in Europe. Deliverable D7.1-3-1 of the Integrated Project “Fire Paradox”,
Project no. FP6-018505. European Commission. 47 p.
150 Towards Integrated Fire Management
Lovén, L. and Äänismaa, P. 2004. Planning of the sustainable slash-and-burn cultivation
programme in Koli National Park, Finland. International Forest Fire News 30: 16–22.
Merou, T.H. and Papanastasis, V. 2002. Legume and grass density under various treatments
in a Mediterranean grassland in Macedonia. In Lowland and grasslands of Europe:
Utilization and development. FAO, Rome. Pp.112–117.
Métailié, J.P. 2006. Mountain landscape, pastoral management and traditional practices
in the Northern Pyrenées (France). In Anoletti, M. (ed.). The conservation of Cultural
Landscapes. CAB International. Pp. 108–124 p.
Montiel, C., Herrero, G., Lázaro, A., Agudo, J. and Aguilar, S. 2009. List and classifcation
of the existing EU and national forest legislation and national policy instruments with
reference to wildland, suppression and prescribed fres. Deliverable D7.1-1-1 of the
Integrated Project “Fire Paradox”, Project no. FP6-018505. European Commission. 110 p.
Naveh, Z. 1975. The Evolutionary signifcance of fre in the Mediterranean Region.
Vegetation 29(3): 199–208.
Pyne, S. J., Andrews, P.L. and Laven, R.D. 1996. Introduction to Wildland Fire, 2nd edition
John Wiley and Sons, New York.
Pyne, S. J. 1997. Vestal Fire. An environmental history told through fre, of Europe and
Europe’s encounter with the world. University of Washington Press.
Ribet, N. 2009. Thèse de 3ème cycle –- Les parcours du feu. Techniques de brûlage à feu
courant et socialisation de la nature dans les Monts d’Auvergne et les Pyrénées centrales.
Rigolot, E. 2005. Bûlage dirigé: Quinze ans d’expérimentation. Espaces Naturales 12: 16–17.
Scott, A.C., Moore, J. and Brayshay, B. 2000. Introduction to fre and the
palaeoenvironment. Palaeogeography, Palaeoclimatology, Palaeoecology 164: 7–11.
SEERAD 2001. The Muirburn Code. Scottish Executive, 21 p.
Van der Zee, F. 2004. Burning of Heathland in Military Areas in the Netherlands
International Forest Fire News 30: 75–76.
Vélez, R. 2005. La población rural en la prevención de incendios forestales, Documento de
Trabajo sobre Protección forestal. FAO, Rome.
Xanthopoulos, G., Caballero, D., Galante M., Alexandrian D., Rigolot, E. and Marzano, R.
2006. Forest Fuels Management in Europe. In: Andrews, P. L. and Butler, B. W. (eds.).
Fuels Management-How to Measure Success: Conference Proceedings. USDA Forest
Service, Portland, OR. 807 p.
4.2 Scientifc Knowledge and Operational Tools
to Support Prescribed Burning: Recent
Developments
Paulo Fernandes
University of Trás-os-Montes e Alto Douro, Vila Real, Portugal
4.2.1 Introduction
Fuel hazard reduction, ecological management or applications in silviculture are
the most common objectives for prescribed burning operations. Prescribed burning
seeks to apply fire to wildland vegetation as exactly as possible, meaning that
specifc treatment objectives are formulated (preferably quantitatively) and burning
is then carried out under well-defned specifc conditions (the prescription) that are
expected to meet those objectives. Other attributes – pre-burn procedures, ignition
patterns, post-burn assessment and monitoring – concur to rationalize and optimize
prescribed burning, as well as to individualize it from traditional burning practices
(Pyne et al. 1996). However, the burn prescription is vitally important because it
translates environmental and fuel conditions into fre behaviour characteristics
and these into fre effects. The ability to prepare and comply with adequate burn
prescriptions is thus central to attain the treatment objectives.
Wildfire science should underpin the decision whenever the use of fire is
considered as a tool to manage vegetation. The fundamental and applied fire
research can directly or indirectly contribute to the development of prescribed
burning technology. Research on fre behaviour and fre ecology issues has provided
the much needed support to implement and maintain successful prescribed burning
programs (e.g. McCaw et al. 2003). As a consistent body of knowledge takes shape
and builds up, increasingly robust and refned burning guidelines leading to more
exact (and desirable) fre effects will hopefully develop.
The initiation and adoption of prescribed burning by managers and management
organizations is relatively new in Europe and dates back to the early 1980s.
Prescribed burning use is geographically restricted and its potential to manage
wildlands is still largely unfulflled, especially in forested areas. While the political
and socio-economic environment is decisive, there is a need for more basic
knowledge and operational guidelines to assist in implementing prescribed burning
programs. The narrow gap that often separates prescribed burning from traditional
burning in Europe testifes to the incipient technical development of the former. This
chapter will provide an overview of the Fire Paradox contribution to advance the
scientifc and technological support to prescribed burning management.
152 Towards Integrated Fire Management
4.2.2 Supporting prescribed burning through increased wildland fuel
knowledge
Dead fuel moisture content usually is the single most relevant variable to plan and
execute a burn operation, as it is directly related with the possibility of fre spread
and is a central element in a prescription. Hence, the ability to produce fuel moisture
content estimates whose accuracy is commensurate with the achievement of specifc
burn objectives is perhaps the most signifcant asset to prescribed burning planning.
Several empirical dead fuel moisture content models based on atmospheric variables
were analysed for needles of pines (Pinus halepensis, P. sylvestris, P. pinea, P.
pinaster and P. radiata), leaves of South American tree species (Nothofagus
antarctica and Astrocedrus chilensis) and for Eucalyptus globulus litter (Vega et
al. 2009). While adequately refecting the pattern of variation in litter and duff
moisture, the examined models fell short in terms of predictive ability. Simple
vapour exchange models (Ruiz et al. 2009) seem suitable for dead fuel moisture
content predictions during periods free from the infuence of rainfall. Laboratory
and feld experiments were carried out to quantify the role of fuel moisture on
ignition likelihood for typical Mediterranean fuel types (Pinus halepensis needles,
grass, Quercus coccifera leaves), highlighting the effect of fuel type and ignition
source on moisture of extinction (Dimitrakopoulos et al. in press).
Being the treatment target, the appraisal and physical description of wildland
fuel is implicit in prescribed burning planning. In particular, the ability to estimate
fuel load and to describe how fuel complexes change with time is of paramount
importance, because those attributes are a major factor in designing a prescribed
burning regime for hazard reduction. A thorough system, outlined in Mårell et al.
(2008), has been devised to collect, organize and exploit fuel data from the point
of view of fire behaviour modelling. Fuel characteristics assessed at different
levels – the fuel particle, the plant, the stand characteristics – from a combination
of destructive and non-destructive sampling methods are stored in a data and
knowledge base (EuroForestFuels) for modelling purposes, including modelling
of fuel dynamics (Krivtsov et al. 2009). The fuel data base interacts with the
Fire Paradox Fuel Manager software (Lecomte et al. 2009), which processes the
vegetation scenarios into 3-D fuel complexes that feed physically-based models of
fre behaviour. The complete process has been applied to Mediterranean France and
resulted in a classifcation based on fre behaviour and vegetation physiognomy and
structure that consists of 10 fuel types (Cassagne et al. 2009a).
4.2.3 Fire-induced pine mortality: setting the limits to prescribed
burning
One important concern when underburning forest stands by prescription is the
amount of damage inficted to the tree canopy. The height of crown scorch is the
basis to assess fre severity and post-burn conifer survival. However, heat transfer
mechanisms linking fre behaviour and tissue necrosis are incompletely understood.
Dupuy and Konovalov (2007) have shown that scorch heights computed from
Scientifc Knowledge and Operational Tools to Support Prescribed Burning... 153
temperature felds generated by the physically-based Firestar 2D model agree well
with plume theory estimates, the current basis to prescribed burning applications.
High flammability and survival to fire intensities in the low to moderate
range justify prescribed burning in pine stands. The fre resistance of European
pine species was ranked on the basis of a combination of morphological and
experimental data with a semi-physical mortality model (Fernandes et al. 2008). All
European pines are amenable by treatment with fre. Pinus canariensis, P. pinaster
and P. pinea possess traits associated to the highest fre survival likelihood, but
species often classifed as fre-sensitive (P. halepensis and P. radiata) do tolerate
low-intensity fre. In South Africa, P. elliottii is the most fre-resistant species of all
commercially grown pines (including P. pinaster), especially when established on
former grassland (De Ronde 2008).
The source of all experimental results that follow is Vega et al. (2008). A study
conducted in a young P. nigra stand (Figure 1) highlighted the role of cambium kill
when crown injury is irrelevant: the survival of individuals with a stem diameter at
1.3-m height (dbh) < 10 cm, corresponding to bark thickness < 0.7 cm, was assured
only when less than half of the bole perimeter at the dbh-level was charred. The
probability of P. pinaster mortality after surface fre was modelled as a function of
crown injury, dbh and season of burn, in the form of three alternative (and similarly
effcient) models where crown damage was expressed as (i) relative scorch height
(RSH), i.e. the crown scorch height as a proportion of tree height, (ii) estimated
crown scorch volume, or (iii) length of the green tip. Nearly all trees with dbh <
7 cm and RSH ≥ 0.90 died, while growing season burns resulted in high (70%)
mortality rates for trees with dbh < 13 cm.
Most results of research on post-burn pine mortality reflect the short-term
outcomes of dormant season, low-intensity fre experiments. The development of
robust, more general models to predict tree mortality implies the need to broaden
the scope of experimental studies in terms of time since fre, burn season and fre
intensity. Data collected in northwest and central Spain revealed that P. pinaster
Figure 1. Tree survival experiments on a young Pinus nigra stand, northern Portugal: during
the fre (left) and one year after (right).
154 Towards Integrated Fire Management
survival 15 months after wildfre was best explained by either crown scorch as
a fraction of crown length or crown scorch volume, and by either the proportion
of bark depth consumed by fre or an exponential function of bark depth. Within
patches or stands of P. pinaster in northern Portugal that experienced single or
multiple fre events, trees with a 12-cm dbh had a 50% probability of surviving
wildfre, the average live tree retaining 64% of its live crown length.
4.2.4 Analysis of prescribed burning treatments: assessing
effectiveness and planning in time and space
Fuel types mapping is the cornerstone of landscape-level analysis of fre hazard,
and hence is vital to the process of selecting areas to treat with prescribed burning,
preferably following an integrated and top-down approach, i.e. from the broad
regional scale to the local level (De Ronde 2009a). The traditional methods used
to recognize and map fuel types are time-consuming and lack accuracy. Several
technologies and software to process high-resolution imagery from remote sensing
have been developed or tested, based on spectral, textural, object-oriented and
3-D methods applied to study areas in France, Morocco and Greece (Borgniet et
al. 2009a). A processing framework and software have been developed to tackle
the various end-user needs and data constraints in order to choose the best mix of
fuel mapping methods; the user can take into account the different constraints and
specifcities of geodata (Borgniet et al. 2009b).
The spatial patterns of prescribed burning are critical to its hazard reduction
effectiveness when extensive landscape treatment is precluded by the available
resources (as well as by other constraints), which very often is the case. After
defning the annual fraction of the landscape that can be treated, optimization is
essentially a matter of defning the spatial arrangement and location of treatment
units. Large, longilinear and irregularly shaped burn units overlapping in the
direction of fire spread will maximize fire growth disruption and fire severity
fragmentation (Loureiro et al. 2006).
Fire simulation is a powerful and flexible tool to assess the effectiveness of
alternative fuel management scenarios and to assist in the temporal and spatial
planning of fuel treatments. Fuel complexes typical of limestone region in southern
France were generated from known post-burn fuel dynamics as an input to the
physically-based Firetec model (Linn et al. 2002). The corresponding fre behaviour
simulations indicate that prescribed burning will mitigate fire hazard effectively
if applied at 2- to 3-year intervals, but this threshold can be seen as a worst-case
scenario, because the spatial resolution used in fuel description underestimates
gaps in the fuel complex (Cassagne et al. 2009b); the annual prescribed burning
efforts required to eradicate high fuel hazard sections were then estimated for two
fuel-breaks. The Farsite fre growth software (Finney 1998) was used in the Marão
Mountains, Portugal (Loureiro et al. 2006) and in the Kassandra Peninsula, Greece
(Cassagne et al. 2009b) to quantify the landscape-level impact of prescribed burning
on the behaviour and size of wildfres conforming to historical patterns. While the frst
case study was applied to an actual treatment in shrubland and the second one was
Scientifc Knowledge and Operational Tools to Support Prescribed Burning... 155
Box 1. A tool to plan and evaluate prescribed
burning operations in pine stands
PiroPinus (Prescribed Burning Guide for Maritime Pine Stands) is an all-in-one tool
based on empirical models to predict fre behaviour and effects for site-specifc stand
and fuel conditions. PiroPinus was conceived for operational use when planning and
evaluating hazard reduction burns and has constantly evolved since its frst version
(Fernandes 2003). It is composed of several inter-related or stand-alone worksheets and
includes the following main functionalities:
• Estimation of fuel loads in each fuel layer, and of moisture contents of surface fne
dead fuels and decomposing litter;
• development of prescriptions by combining restrictions for fre severity on the tree
canopy and on the forest foor, expressed respectively by crown scorch ratio and
duff consumption %;
• assessment of the likelihood of sustained fre spread and of marginal burning
conditions and fre characteristics quantifcation (rate of spread, fame length,
Byram’s fre intensity);
• ignition planning to optimize treatment rate;
• assessment of canopy injury and tree mortality;
• estimation of fuel consumption and post-burn fuel accumulation;
• post-burn evaluation of fre impacts.
PiroPinus was not tested in species other than maritime pine (Pinus pinaster).
Nevertheless, it can be easily adapted by replacing equations or adding options for other
pines.
Figure 2. PiroPinus screenshot illustrating the development of a prescription.
156 Towards Integrated Fire Management
applied to an hypothetical treatment in conifer forest, both highlighted the potential
of prescribed burning to decrease the likelihood of large-scale wildfre, as well as the
crucial role played by the spatial patterns (strategic versus random) of treatments.
Prescribed burning was discussed and proposed as a tool to mitigate CO
2
emissions
to the atmosphere in countries with an acute wildfre problem (Narayan et al. 2007).
This would encourage its expansion provided that the current carbon accounting
practices change (Hurteau et al. 2008). However, the potential of prescribed burning
to reduce emissions depends on how the fre regime is changed, especially in terms
of the overall area burned and fre intensity (Bradstock and Williams 2009). This
would increase even more the relevance of prescribed burning optimization in space
and time, as well as giving more weight to fuel consumption and smoke production
issues when planning burning operations.
4.2.5 Planning prescribed burning operations
Decision-support systems for prescribed burning operations at the treatment unit
level may take several forms, from general guidelines and burning windows to
complex software tools. The feasibility and relevance of prescribed burning in pine
stands, as well as general technical guidelines that consider fuel classifcation and
stand structure, relative fre resistance, silvicultural regimes and fring techniques,
have been compiled by De Ronde (2009b). While derived for the South African
environment most of those guidelines are broad enough to apply elsewhere,
including in the Mediterranean Basin. However, as previously mentioned, tying the
burn objectives to the fre environment is an essential feature of decision support in
prescribed burning operations.
Fernandes (2010) has compiled the current burn prescriptions and technological
solutions used to develop prescriptions in Europe, noting that burning guidelines
are not always consistent (or up to date) with the available scientifc knowledge.
General burning windows consisting of ranges in weather conditions or in fre
danger rating indexes (like in Sweden and Portugal) are commonplace. In Catalonia,
Spain, 12 standard prescriptions are individualized on the basis of fuel availability
and wind speed. USDA Forest Service fre simulation tools (e.g. Scott and Reinhardt
2001) are used in Spain to prepare site-specifc prescriptions, which include sets
of values (minimum, preferred, maximum) for weather conditions, fuel moisture
contents, fre behaviour characteristics and selected fre effects. In Portugal, the
PiroPinus tool (Fernandes, 2003) was developed to assist in planning and evaluating
the results of prescribed underburning in maritime pine (Pinus pinaster) stands.
4.2.6 Conclusion
This chapter has described the contribution of Fire Paradox to expand the scientifc
knowledge and technical application for prescribed burning. Past and current
research efforts on topics that explicitly or implicitly are relevant to prescribed
Scientifc Knowledge and Operational Tools to Support Prescribed Burning... 157
burning have contributed to answer why, when and how to burn selected European
vegetation types. Our understanding of the fire regimes required to perpetuate
species and communities has advanced. Wildland vegetation has been characterized
into fuel parameters, allowing fre hazard classifcation and mapping, for inputs
into operational and theoretical models of fre behaviour characteristics. Links
between the fre environment and indicators of fre severity have been identifed
and described. Decision-support tools for a more effcient and safe use of prescribed
burning are available.
Despite the advances made, the knowledge accumulated up to now is still too
fragmented and incomplete, mirroring the present under-development of prescribed
burning in Europe. In particular, the relationships between fire behaviour and
specifc fre effects are poorly understood for many vegetation types. While this
may not be an obstacle for sound land management using fre as a tool, it will
certainly affect the quality of the achievements of an activity that will likely undergo
increased complexity, public scrutiny and debate. Research and development on
prescribed burning will continue to be required, but scientifc monitoring of the
results of both traditional and prescribed burning practices may well be an additional
key source of data and information.
References
Borgniet, L., Long, M., Capel, A-C., Bouillon, C., Morge, D., Ganteaume, A., Lampin,
C., Jappiot, M., Curt, T., Machrouh, A., Sesbou, A., Mantzavelas, A., Apostolopoulou,
I., Partozis, T. and Gitas, I. 2009a. Method to produce vegetation and fuel maps using
satellite imagery: Deliverable D5.1-2 of the Integrated Project “Fire Paradox”, FP6-
018505. European Commission. 51 p.
Borgniet, L., Long, M., Capel, A-C., Bouillon, C., Morge, D., Ganteaume, A., Lampin,
C., Jappiot, M., Curt, T., Machrouh, A., Sesbou, A., Mantzavelas, A., Apostolopoulou,
I., Partozis, T., Gitas, I., Marell, A., Cassagne, N., Pimont, F., Rigolot, E., Morsdorf, F.,
Koetz, B. and Allgöwer, B. 2009b. Development of an easy to use tool to recognize and
map fuel models: Deliverable D5.1-6 of the Integrated Project “Fire Paradox”, FP6-
018505. European Commission. 107 p.
Bradstock, R.A. and Williams, R.J. 2009. Can Australian fre regimes be managed for
carbon benefts? New Phytologist 183: 931–934.
Cassagne N., Curt, T., Pimont, F., Schaffhauser, A., Rigolot, E., Borniet, L., Dupuy, J.L.,
Estève, R., Ganteaume, A., Jappiot, M., Martin, W., N’Diaye, A., Petit, P., Portier, D.
and Valette, J.C. 2009a. Fuel description and modelling. Internal Report IR3.4-4 of the
Integrated Project “Fire Paradox”, FP6-018505. European Commission. 41 p.
Cassagne, N., Mantzavelas, A., Dupuy, J.L., Lazaridou, T., Pimont, F., Partozis, T.,
Fernandes, P., Loureiro, C. and Rigolot, E. 2009b. Report on the effects of broad scale and
long-term application of prescribed burning on fre risk components. Deliverable D5.3-1
of the Integrated Project “Fire Paradox”, FP6-018505. European Commission. 88 p.
De Ronde, C. 2008. Knowledge base in damage assessment to forests and plantations.
Deliverable D3.2-7 of the Integrated Project “Fire Paradox”, FP6-018505. European
Commission. 21 p.
De Ronde, C. 2009a. Recommendations for prescribed burning mapping in Mediterranean-
type natural vegetation and forests. Deliverable D5.4-2 of the Integrated Project “Fire
Paradox”, FP6-018505. European Commission. 17 p.
158 Towards Integrated Fire Management
De Ronde, C. 2009b. Guidelines for prescribed burning for the specifc vegetation types
assessed within the FIRE PARADOX community. Deliverable D9.1-4 of the Integrated
Project “Fire Paradox”, FP6-018505. European Commission. 63 p.
Dimitrakopoulos, A., Mitsopoulos, I. and Gatoulas, K. 2010. Assessing ignition probability
and moisture of extinction in a mediterranean grass fuel type. International Journal of
Wildland Fire. In press.
Dupuy, J.L. and Konovalov, V. 2007. Simulations of fre impact on tree foliage: 2D results.
Deliverable D2.4-2 of the Integrated Project “Fire Paradox”, FP6-018505. European
Commission. 26 p.
Fernandes, P.M. 2003. Sistema de apoio ao planeamento, execução e avaliação do fogo
controlado em pinhal bravo. In: O Estado da Arte em Modelação da Floresta – o Caso
Português, Actas da I Jornada Técnica de Modelação Florestal. ISG, CIGest, Lisboa. Pp.
33–34.
Fernandes, P.M. 2010. Handbook to plan and use prescribed burning in Europe. Product
P13.1-3 of the Integrated project “Fire Paradox”, FP6-018505. European Commission.
Fernandes, P.M., Vega, J.A., Jiménez, E. and Rigolot, E. 2008. Fire resistance of European
pines. Forest Ecology and Management 256: 246–255.
Finney, M. 1998. FARSITE: Fire Area Simulator-model development and evaluation. USDA
Forest Service, Rocky Mountain Research Station, Res. Pap. RMRS-RP-4. Ogden, UT. 35 p.
Hurteau, M.D., Koch, G.W. and Hungate, B.A. 2008. Carbon protection and fre risk
reduction: towards a full accounting of forest carbon offsets. Frontiers in Ecology and
Environment 6: 493–498.
Krivtsov, V., Fernandes, P., Higham, B., Pezzatti, G.B. and Legg, C. 2009. Release of
prototype software for allometric modelling of fuel structure with provisional case studies.
Deliverable D3.3-5 of the Integrated Project “Fire Paradox”, FP6-018505. European
Commission. 33 p.
Lecomte, I., de Coligny, F., Griffon, S., Pimont, F., Rigaud, E., Rigolot, E. and Vigy, O.
2009. Fire Paradox fuel manager: software and user’s manual. Product P6.1-6 of the
Integrated Project “Fire Paradox”, FP6-018505. European Commission. 85 p.
Linn, R., Reisner, J., Colman, J. and Winterkamp, J. 2002. Studying wildfre behavior using
FIRETEC. International Journal of Wildland Fire 11: 233–246.
Loureiro, C., Fernandes, P., Botelho, H. and Mateus, P. 2006. A simulation-based test of a
landscape fuel management project in the Marão range of northern Portugal. In: Viegas,
D.X. (ed.). Proceedings of the 5th international conference on forest fre research. Elsevier
B.V., Amsterdam. CD-ROM.
Mårell, A., Abdelmoula, K., Allgöwer, B., Borgnier, L:, Cassagne, N., Curt, T., Ganteaume,
A., Ghosn, D., Gitas, I., Jappiot, M., Jimenez, E., Kazakis, G., Koetz, B., Lampin, C.,
Machrouh, A., Morsdorf, F., Pimont, F., Rigolot, E., Sesbou, A., Vega, J.-A. and Wellani,
W. 2008. Fuel description: frst results. Deliverable D3.4-3 of the Integrated Project “Fire
Paradox”, FP6-018505. European Commission. 80 p.
McCaw, L., Cheney, P. and Sneeuwjagt, R. 2003. Development of a scientifc understanding
of fre behaviour and use in south-west Western Australia. In: Abbott, I. and Burrows, N.
(eds.). Fire in Ecosystems of South-west Western Australia: Impacts and Management.
Backhuys Publishers, Leiden, The Netherlands. Pp. 171–187.
Narayan, C., Fernandes, P.M., van Brusselen, J. and Schuck, A. 2007. Potential for CO2
emissions mitigation in Europe through prescribed burning in the context of the Kyoto
protocol. Forest Ecology and Management 251:164–173.
Pyne, S.J., Andrews, P.L. and Laven, R.D. 1996. Introduction to wildland fre. John Wiley &
Sons, Inc., New York, USA. 769 p.
Ruiz, A.D., Vega, J.A. and Alvarez, J.G. 2009. Construction of empirical models for
predicting Pinus sp. dead fne fuel moisture in NW Spain I: response to changes in
temperature and relative humidity. International Journal of Wildland Fire 18: 71–83.
Scientifc Knowledge and Operational Tools to Support Prescribed Burning... 159
Scott, J.H., and Reinhardt, E.D. 2001. Assessing crown fre potential by linking models
of surface and crown fre behavior. USDA Forest Service, Rocky Mountain Research
Station, Res. Pap. RMRS-RP-29. Ogden, UT. 59 p.
Vega, J.A., Ruiz, A.D., Ubysz, B., Guijarro, M., Bianchi, L., Defossé, G. and Todone, J.F.
2009. Dead fuel moisture modelling and validation. Deliverable D3.1-8 of the Integrated
Project “Fire Paradox”, FP6-018505. European Commission. 68 p.
Vega, J.A., Fernandes, P., Conedera, M., Botelho, H., Fernandes, M., Jiménez, E., Loureiro,
C., Lucini, L., Pérez, J.R. and Pezzatti, B. 2008. Tree resistance to fre: frst results.
Deliverable D3.2-2 of the Integrated Project “Fire Paradox”, FP6-018505. European
Commission. 24 p.
4.3 Learning and Training on the Use of Prescribed
Burning Techniques
Conceição Colaço
1
and Domingo Molina
2
1
Centre of Applied Ecology “Prof. Baeta Neves”, Technical University of Lisbon,
Portugal
2
Unit of Forest Fires, University of Lleida, Spain
4.3.1 Introduction
Prescribed burning is often described as a mix of art and science that departs from
traditional burning practices by the existence of formal planning and monitoring
procedures (Pyne et al. 1996). Due to the special nature of this technique, prescribed
burning must be based both on science and experience, and simultaneously, those
who perform it, must be skilful and competent.
A prescribed fire seeks well-defined effects that will fulfil one or more
management goals, which are attained by burning in a specifc fre environment (the
prescription) and following specifc operational procedures (the burn plan) (Pyne
et al. 1996). Fire hazard reduction frequently is the main reason to use prescribed
burning, but this technique is also used to manage habitats for pastures, game and
conservation (Fernandes et al. 2002).
The traditional use of fire by many rural communities in Europe was, and
continues to be, a very important tool in shaping the landscape, together with
other techniques like grazing, and cutting, creating a landscape mosaic with high
ecological and cultural diversity (Goldammer et al. 2007; Silva 1990; Trabaud
1992; Pyne 1984). However, even if in the history of land use in Europe, fre was
an important element with a strong presence in agriculture, forest and grazing
areas, after World War II, fre use in rural areas was almost excluded because of
land-use changes with the arrival of more technologies and also by the increasing
abandonment of rural areas particularly in the Mediterranean. At present, there is a
general neglect of traditional fre uses in the northern countries and central Europe
in contrast with the continued practice in Mediterranean countries but on a smaller
scale than before (see the chapter by Montiel et al. in this book). These traditional
practices can be one of the main causes of wildfres through fres escaping (FAO
2007; DG JRC-IES 2008). Because of this problem, the traditional use of fre often
has strict regulations, while the knowledge acquired on the utilization of fre to
prevent wildfre has started to vanish.
Although in some cases the traditional use of fre can be a cause of destructive
wildfires, prescribed burning performed by experienced practitioners has been
known for a long time to be benefcial to the reduction of forest wildfre hazard.
In the 19
th
century, the use of fre was already recommended by some Portuguese
forestry experts (Varnhagen 1836; Pimentel 1882) as a tool to “control the
162 Towards Integrated Fire Management
undesirable understory, especially in pine forests” to prevent wildfres. Similar
examples were found in France, in the second half of the 19
th
century, and the
long history of the different cycles on the use of prescribed burning in forestry,
including the recent cooperation between Europe and the USA was described in the
documentary “Fire in the Balance” (Manso et al. 2008) produced by Fire Paradox
and shown at the 20
th
anniversary of the Yellowstone Fires.
In the most recent decades the use of prescribed burning for the reduction of
wildfre hazard in Portugal, Spain and France has importantly increased, with some
valuable efforts focused on operational training for prescribed burning.
The frst attempts to promote training in prescribed burning for fre prevention in
European forests date from the 1980s in Portugal (Rego et al. 1989: Silva 1990).
In Spain, the frst prescribed burning training course was conducted in 1995 at the
University of Lleida, which included a detailed prescribed burning plan and its
prescription window using Behave ‘RxWindow’ software (Andrews and Bradshaw
1990). These prescribed burnings (ca. 2 ha total) were accomplished in plots
established in the Tortosa Mountains (Catalonia, northeast Spain). In 1998, the
Spanish Ministry of the Environment developed and disseminated a training and
demonstration video on prescribed burning with the scientifc and didactic support
of the University of Lleida. Since 2000, GRAF (the fire fighting organization
of Catalonia ) has trained on prescribed burning on a regular basis in Catalonia
(Spain). At the beginning of the 1990s fire professionals in France created a
Prescribed Burning Network and the frst “Charter of prescribed burning”, signed
by 74 technicians in 1997. Materials with contents for the use of fre to support
their training program have been permanently updated since then (CIFSC 2004;
Binggeli and Bento 2009). Many of these early efforts followed similar programmes
in the USA, Canada and Australia where the use of prescribed fre was more widely
spread than in Europe and systematic approaches to provide training on the use of
prescribed burning techniques were more developed.
In other parts of the world, prescribed burning courses followed similar patterns.
For example, in Argentina courses started in 1996 promoted by INTA in Santiago
del Estero (Kunst and Moscovich 1996). Before that, several isolated attempts to
carry out prescribed burning based on scientifc knowledge were made only by the
National Park Administration and other organizations. In South Africa, prescribed
burning courses have existed since 2001, mainly attended by foresters and nature
conservationists.
In all cases, many of the technicians who applied the technique of prescribed
burning were forest and range managers, but they could also be fire fighters,
among many other professions. We will use the term fre professional for all of
those trained in fre. In our point of view, agreeing with Kobziar et al. (2009),
fire professionals should have not only specific training on prescribed burning
(professional/continuous training) and practical experience in the feld, but also a
good fre education (university courses).
Following this premise, the objectives of this chapter are to:
• Describe and assess the current context on prescribed burning training in
Europe (Spain, Portugal and France);
• Propose ways to continue the efforts to harmonise qualifcations in Europe and
to permanently exchange experiences on prescribed training;
Learning and Training on the Use of Prescribed Burning Techniques 163
• Describe and assess the current situation with regard to forest education,
university level in Europe (Spain and Portugal) and compare it to the US forest
education;
• Propose new ways for the example of an International Graduate Program in
Fire Science and Management.
4.3.2 Training on prescribed burning in Europe
When addressing the term vocational training, the focus is on “providing the skills,
knowledge and competences needed in the labour market” (Tissot 2004). In this
section we will describe the vocational training on prescribed burning in Portugal,
Spain and France since these are the countries in Europe with the most experience
on this subject. Details on the training systems are provided in Molina et al. (2009).
In Portugal there are several institutions and professionals that work in a direct or
indirect way with fres. To acquire competences on the use and management of fre
(prescribed burning included), these professionals look for specifc training that can
be provided by universities, institutions like the National Forest Authority (at the
base of Lousã) or by some Portuguese Forest Associations (in particular Forestis) in
partnership with the universities.
In Portugal, by law, only an accredited technician can conduct a prescribed burn.
Due to that obligation, the National Forest Authority (AFN) certifes the technicians
and their training by law (Portaria nº 1061/2004 of 21
st
of August). Also, to perform
a prescribed burn, the ‘Burning plan’ must be evaluated by the municipalities and
forest services, with the technicians of those entities also receiving specifc training
to acquire the knowledge to evaluate the burning plans. Finally, there is a specifc
training provided by AFN for the forest workers and fre prevention brigades that
support the use of prescribed fre.
The contents of the three types of training are common in addressing issues as
guidelines for use of fre, fre behaviour and its impacts, and prescribed burning
planning and operational implementation. These issues are addressed at different
depths and the practical modules with feld experience are not required for the group
of the evaluators. The duration of the courses vary accordingly:
1. Prescribed burning training for technicians accreditation
Theoretical modules (duration: 21 hours)
Theoretical-practical modules (duration: 35 hours)
Practical modules (duration: 49 hours)
2. Training for technicians able to evaluate the prescribed burning plan
Theoretical modules (duration: 9 hours)
Theoretical-practical modules (duration: 19 hours)
3. Forest workers and fre prevention brigades
Theoretical modules (duration: 4 hours)
Simulated practice modules (duration: 31 hours)
In Spain, although several vocational courses are offered to professionals by the
Ministry of the Environment, or by fire fighter academies, none of them work
164 Towards Integrated Fire Management
specifcally on prescribed burning except for the region of Catalonia. There, the
Institute for Public Safety (ISPC) has been providing training in the use of fre since
2001. There are fve training levels for different job positions (all end-users in the
fre fghting institute) and two specifc modules for: (i) specialists that will execute
both prescribed burning and backfring), and (ii) managers who plan and supervised
them. Today a total of 120 fre fghters and managers have taken these courses.
Since 2009, 55 of them have a certifcate that ensures the required competence for
a wise use of fre.
Recently (since 2008), the Department of the Environment of Gran Canaria
(Canary Island local government) is developing a prescribed burning course in
collaboration with the University of Lleida. They have accomplished three editions
with more than 50 people. Trainees receive a diploma if they pass the course.
Additionally, the Department of the Environment will latter certify their competence
(for their workers) after successful work on 20 prescribed burning days (equivalent
to 160 hours).
In France, fire is taught through a professional training system that is only
available in some regions. Since the risk of wildfre is only important in part of the
country, only 32 departments (mostly in the south) among the 95 departments are
subject to the wildfre risk legislation. Only fre fghters and fuel managers of this
part of France are to be trained in prescribed burning techniques. Therefore, there
are two schools for fre professionals in France dealing with prescribed burning –
an established one in the southeast, and a more recent one in the southwest). The
schools are managed by the civil protection services but are open to foresters.
French trainees on prescribed burning get a training certifcation if they follow
and complete a specifc course and are endorsed by the trainers. The certifcate
is valid for fve years and they can become prescribed burning bosses if asked
by their employers. The assessment of practices in the feld is only performed by
the employers. Better assessment of the quality of the practices in the feld could
strengthen the development of this technique.
Training in prescribed burning is the same for fre fghters and for vegetation and
fuel managers. They are learning in the same classrooms and practicing together in
the feld. A general certifcation for conventional fre fghting activities, with fve
levels, includes assessment of trainees. However, prescribed burning is a training
option that French fre fghters or foresters are not required to take.
4.3.3 Cooperation, exchanges and the search for harmonization of
European qualifcations on prescribed burning
One of the objectives of Fire Paradox was the creation and sharing of knowledge
between fire professionals of Europe and elsewhere and the promotion of
their mobility between countries. However, with the diversity of educational
qualifcations in different European countries, the mobility of workers in Europe
is very diffcult. To solve this question, the European Commission created The
European Qualifications Framework (EQF) that acts as a translation device to
make national qualifcations more readily understood and compared across Europe,
Learning and Training on the Use of Prescribed Burning Techniques 165
promoting the mobility of workers and learners mobility between countries
and facilitating their lifelong learning (EC 2008). The EQF does this by setting
qualifications into a series of reference levels (1–8), from basic to advanced.
The eight reference levels are described in terms of learning outcomes, split into
knowledge, skills and competence.
In collaboration with Fire Paradox, the EuroFire project (www.euro-fre.eu/),
has developed fre training materials to support the development of similar fre
management skills across the European Union. These materials have been developed
to support fre training and assessment for a Level 2 work environment (members
of a tanker crew, hand crew, or prescribed burning crew) whose personnel are
instructed to do tasks under direct supervision. Six EuroFire competency standard
units (Table 1) for fre crews and personnel involved in the suppression of wildfres
(forest fres and rangeland fres) and in prescribed burning have been developed.
Four training modules to support fre fghters self-learning are also available on the
EuroFire website.
The diversity of fre professional qualifcations in different European countries
is also related to the complexity and diversity of cooperation mechanisms between
departments dealing with fre management at the European and National Level
(Correia et al. 2008). In spite of this complexity there are good examples such as
that of the cooperation in the feld of sharing knowledge, information and practical
experiences between Spain and Portugal. Because of the lack of a global European
framework for continuous cooperation between fre professionals the opportunities
provided by the General Assemblies of Fire Paradox were used to promote
exchanges between these professionals:
• Cagliari, Sardinia, Italy (2007) Seminar on the use of prescribed burning
in Italy for the prevention and fght against forest fres. The objective of
Table 1. EuroFire competency standard units and training modules for self learning.
Competency standard units Training modules for self learning
EF1:Ensure that your actions in the EF1: Ensure that your actions in the
vegetation fre workplace reduce the risks vegetation fre workplace reduce the risks to
to yourself and others. yourself and others
EF2: Apply techniques and tactics to EF2: Apply techniques and tactics to control
control vegetation fre vegetation fre
EF3: Communicate within a team and EF4: Apply hand tools to control vegetation
with supervisors at vegetation fres fre
EF4: Apply hand tools to control EF6: Apply vegetation ignition techniques
vegetation fres
EF5: Control vegetation fres using
EF6: Apply vegetation ignition techniques
pumped water

166 Towards Integrated Fire Management
this seminar was to initiate and stimulate the development of the theme of
prescribed burning in Italy, through the exchange of ideas, knowledge and
experiences between Italian professionals and professionals from other
countries;
• Chania, Crete, Greece (2008) Enriching fre management in Greece with new
tools: suppression fre and prescribed burning. This seminar had the purpose
of initiating a plural debate on the use of fre in Greece, where it is still not
allowed. Some professionals, forestry and fre fghters, and Greek researchers,
gave their contribution, in order to show some initial experiences on prescribed
burning, as well as, the need to legalise suppression fre, often used in fre
fghting without consent. International experts explained the use of fre in
their respective countries, as a way to infuence a serious refection on the
introduction of the use of fre on prevention and fghting Greek wildfres.
• Puerto Madryn, Patagonia, Argentina (2009) 1
st
South American Symposium
on Fire Ecology and Management. This symposium had the purpose of
gathering South American professionals together to debate on fre ecology
and management. Some professionals, forestry and fre fghters, and Argentine
researchers, gave their contribution, describing the diversity of situations on
the use of fre in South American countries (Argentina, Uruguay, Chile and
Paraguay).
These types of meetings are also a very good opportunity to discuss and exchange
views, experiences and actions and to address issues as job hazard abatement during
the use of fre, allowing for the completion of an educational and training report on
“Guidelines to mitigate personal risk in prescribed burning and suppression” (Pous
et al. 2010).
The way forward for the continuation of the exchanges and cooperation between
fre professionals can be based on the experience of these meetings and especially
on the Euro-Mediterranean meeting for prescribed burning professionals held in
Lousã, Portugal, 4–8 February 2009.
Similarly to the objectives of the annual meetings of the French National
Prescribed Burning Network, this Euro-Mediterranean meeting had the main goal
to expand exchanges to a much broader public, encouraging knowledge transfer and
promoting interactions between European prescribed burning professionals. This
was a very fruitful initiative, gathering almost one hundred Portuguese, Spanish,
Italian, French, Moroccan and German prescribed burning professionals (Figure
1). For fve days the participants had the opportunity to visit twelve sites where
prescribed burning had been applied, as demonstration sites.
Future training initiatives at the European level should also make extensive use of
the Network of Demonstration Sites for Prescribed Burning created by Fire Paradox
(Molina et al., 2009b). Documentation of the most relevant sites can be found in the
Fire Paradox information platform – Fire Intuition (freintuition.ef.int/).
A permanent framework for these exchanges of experiences at the European level,
taking advantage of the networks of sites, people and materials gathered in projects
such as Fire Paradox, and building on the experiences of national and international
meetings is fundamental for the future.
Learning and Training on the Use of Prescribed Burning Techniques 167
4.3.4 Fire education in Europe (Portugal and Spain)
The European Universities and Institutes must enter by 2010 the European Higher
Education Area (EHEA) where the university systems will be harmonized under the
Bologna agreements, consisting of three different levels of studies:
• 1
st
cycle, with 3–4 academic years of study (depending on each country the
denominations can be Bachelor of Sciences BSc, engineer, deputy forestry
engineer, technical engineer, among others);
• 2
nd
cycle with 1, 1.5 or 2 more academic years (denominations like Master of
Science MSc, graduate, and forestry engineer are common);
• 3
rd
cycle, consisting of more 3–5 academic years (the denomination of this
level is Doctorate, Philosophy Doctor, PhD).
In Europe, specifc university level courses on fre management are mostly available
in Portugal and Spain, and these were reviewed by Colaço (2005) and by Molina
et al. (2009a). Although France is a country with a tradition of forest science, and
experience in the use of the prescribed burning technique, at present, this formal
education system does not include courses on prescribed burning, even if they are
topics in the PhD programs of some French university schools (Molina et al. 2009a).
In Portugal, fre management education is inserted both in the educational formal
context as a forest engineering graduation major (university level) and in continuous
learning and training programmes, that can also be offered by polytechnic institutes,
Figure 1. Visit to a prescribed burning demonstration site during the “Euro-Mediterranean
meeting for prescribed burning professionals”.
168 Towards Integrated Fire Management
especially with technological specialization programmes (CET). The CET training
programmes in Portugal allow for integration into the working world after the
secondary school or the continuation of studies at university level. The training held
in CET is credited within the university where the student is admitted. In Portugal
there are two approved CET programmes on Forest Defence Against Wildfres. Both
are promoted by Polytechnic Institutes, (Coimbra and Bragança) with 43 European
Credit Transfer System (ECTS) credits each and two semester duration (including
420 to 600 hours of training in a work context). Both schools have the course of
prescribed burning with 3 ECTS.
In Portugal, there are fve schools that offer graduate studies in Forestry science:
two Universities (the University of Trás-os-Montes and Alto Douro - UTAD - in
Vila Real, and the Technical University of Lisbon with the Institute of Agronomy
- ISA/UTL), and three polytechnic institutes (the Agrarian Schools of Bragança
ESAB, Viseu ESAV and Coimbra ESAC). All the schools have courses related to
fres (Table 2).
The PhD degree is given by ISA/UTL and UTAD. On this cycle there are no
specifc courses on fre, however students can do their research in this subject area.
In Spain, there are twelve universities/colleges offering higher education for fre
professionals. On the 1
st
cycle, corresponding to six semesters, 11 schools have
a mandatory course on ‘Natural Resources Protection’ including Wildland Fire
Issues with around 120 hours of instruction. On the 2
nd
cycle, seven schools have
a mandatory course on ‘Natural Systems Management & Protection’ including
Wildland Fire Issues with around 240 hours of instruction (Table 3).
Spanish universities also provide graduate courses or fre management that are
nearly always short (less than 9 ECTS credits), and frequently are partly on-line
courses. These courses are addressed to professionals who want to improve the
knowledge that they have in specifc subjects, to know the advances and the new
contributions and also to be in touch with other specialists. The University of Lleida,
University of Cordoba, and Polytechnic University of Madrid offer different courses
on the subject ‘Wildland Fire Management’.
Taking the University of Lleida as an example, we see that in this school, students
can enrol in any of these courses individually and get credit and a diploma for it.
However, a ‘Wildland Fire Management Master’ degree can be achieved if a fnal
thesis or project is completed and a total of 45 ECTS credits is obtained.
Table 2. Forestry Science higher studies in Portugal and corresponding fre courses.
Courses (with the corresponding European Credit Transfer System – ECTS*)
First cycle Second cycle
ESAB, ESAV Forest Fires (6 ECTS)
ESAC Forest Defense Against Wildfres Planning of Forest Defense Against
(6 ECTS) Wildfres (6 ECTS)
UTAD, Forest Protection (7-7.5 ECTS) Fire Ecology and Management
ISA/UTL (5-6 ECTS)
* 1 ECTS = from 25 to 30 hours (EC 2007)
Learning and Training on the Use of Prescribed Burning Techniques 169
At the European level, the existing cooperation in forestry associates six European
universities (Joensuu, as coordinator, Lleida, Freiburg, Vienna, Wageningen, and
the Swedish Agricultural University) for a Master of Science in European Forestry.
The Master of Science in European Forestry Erasmus Mundus (MSc EF) is a 2-year
double-degree program including various topics. The idea of the MSc-program is
to respond to the increasing number of issues in forest and nature management
that are formulated, implemented, and co-ordinate above the classical nation-states,
providing a whole range of new challenges and demands for policy and management
at the European level. In this context, the participation of the University of Lleida
includes a course on ‘Principles of Wildland Fire Science and Management’ and a
‘Prescribed Burning Laboratory’ with the cooperation of GRAF in Catalonia.
Cooperation between universities of different European countries exist in many
research projects dealing with fre management but no specifc common degree
programmes under this topic have been developed.
4.3.5 Current initiatives in cooperation on fre education outside
Europe
International cooperation developed under the Fire Paradox project outside Europe.
Four international prescribed burning courses were taught between 2007 and
2009 in Patagonia and Santiago del Estero. The courses were organized by the
two Argentinean teams (INTA and CIEFAP) with the assistance of the University
of Lleida. More than 180 fre fghters, forestry students and professionals (most
of them from different Argentinean provinces, but also from Spain, Uruguay and
Paraguay), were trained in the use of prescribed burning techniques. The end
result was the decision taken by the University of Patagonia during the 1
st
South
American Fire Ecology and Management Symposium, to support the creation of
a Masters Degree Program, which will continue with the task of broadening the
knowledge and giving a more academic and professional scope to the dissemination
of fre science, and in particular to training and learning about the use of prescribed
burning techniques initiated during the development of the project.
Table 3. Forestry science higher studies in Spain with fre contents.
Universities First Second
cycle cycle
Universidad de Córdoba No Yes
Universidad de Castilla-La Mancha, Universidad de Extremadura,
Universidad de Huelva, Universidad de León, Universidad de Oviedo Yes No
Universidad Católica Santa Teresa de Jesús de Ávila, Universidad de
Lleida, Universidad Politécnica de Madrid, Universidad de Santiago
de Compostela, Universidad de Valladolid, Universidad Politécnica
de Valencia Yes Yes

170 Towards Integrated Fire Management
New initiatives to enhance international cooperation should be based in the
knowledge of the status of fre education outside Europe. To compare the academic
studies on fre outside Europe we decided to use the case of the USA since it is a
country with strong experience in fre education and training, with a recent and
excellent review on the challenges to educate their fre professionals (Kobziar et al.
2009).
As in Portugal and Spain, fre management and ecology education have been
concentrated in forest and range management programs at universities and technical
community colleges. According to a recent study, only 22 of the thousands of US
universities (providing 2- or 4-year programs), “provide substantial educational
opportunities in wildland fre management and ecology” (Kobziar et al. 2009).
From the 22 universities and colleges that provide at least one certifcate or degree
in fre science, 15 of them offer courses both for undergraduate and graduate studies
(Table 4). Nine universities offer several fre specifc courses at both levels. This
allows for a certain specialization very useful for cooperation.
The need to target mid career professional, technicians or fre professional that are
already employed, meant that both the US and Spanish universities prepared specifc
courses for them. This is the case of the University of Idaho which has an academic
major and eight fre-specifc continuing/distance/ short courses related to fre science.
4.3.6 New challenges and initiatives for the educational profle of a
fre professional
To have a well educated fre professional, the professional needs to learn, and that is
a “process where knowledge is presented to us, then shaped through understanding,
discussion and refection” (Freire 1998).
Table 4. Universities in the USA offering both undergraduate and graduate fre specifc cours-
es (adapted from Kobziar et al. 2009).
US Universities Number of fre specifc courses
Undergraduate Graduate
Level Level
Humboldt State University 6 2
California Polytechnic & State University; University of Idaho 6 1
Colorado State University 4 2
Oklahoma State University; Oregon State University 3 2
Northern Arizona University 3 1
Texas Tech University; University of Florida 2 2
Clark University; Stephen F. Austin St University;
University of California-Berkeley; University of Oregon;
University of Montana; University of Washington 1 1

Learning and Training on the Use of Prescribed Burning Techniques 171
In the last decades, there has been a substantial effort to understand, discuss and
refect about fre and new challenges have appeared for the educational background
of a fre professional. According to Stephens and Ruth (2005), “fre management is
in transition from an era dominated by fre suppression, to one where fre use and
suppression are equally viable resource management options.” Due to this changing
paradigm, the education and training system must evolve to a more integrated way
of looking into the use of fre in the ecosystem.
As referred to in Kobziar et al. (2009) the “new generation of fre professionals
must continuously incorporate new knowledge on fire ecology, fire behaviour,
and social sciences to tackle the multifaceted issues they will face”. Kobziar et al.
(2009) developed the ‘Fire Professional Development Triangle’, which depends on
integrating training, education and experience (Figure 2). None of the sides of the
triangle should be underestimated, but should be complementary.
In spite of these new needs, recognized internationally, there are still no
internationally stable frameworks for cooperation between entities that would
provide the new fre professionals with the knowledge, skills and competences for
the future demands.
In addition, it is widely recognized that modern approaches to education in fre
science should use the concepts of Integrated Fire Management. This implies that a
complete fre professional could be trained to understand and/or use both prescribed
burning and suppression fre. Simultaneously he/she should understand the social
context including the traditional fre uses. This is a new perspective, solving the
apparent paradox of having agents only prepared to use fre and others only prepared
to extinguish it without an integrated perspective. At the moment there are no such
international education programs.
Being aware of this gap at the international level, Fire Paradox launched the
initiative to develop an International Graduate Program in Fire Science and
Management for a Masters Degree (FireMaster). A partnership was established
between the members of the Fire Paradox consortium with more experience in
education in prescribed burning (University of Lleida and UTAD) and two of the
more experienced US universities in this subject (Idaho and Florida) with the intent to
extend the partnership in a second step to other universities (e.g. ISA and Patagonia).
The objective of the partnership is to jointly develop a program to educate European
and American graduate students in understanding, synthesizing and applying science
to support Integrated Fire Management for forests and other natural resources.
Globally, the Graduate Program could involve three complementary parts,
referring to science, applications and projects:
Figure 2. The fre professional development triangle (Kobziar et al. 2009)
T
r
a
i
n
i
n
g
E
x
p
e
r
i
e
n
c
e
Education
172 Towards Integrated Fire Management
• Advanced Fire Science courses offered online and developed jointly for
different topics (Fire Ecology, Fuels Inventory and Management, Fire
Behaviour, Analysis of Real Case Studies, Fire and Landscape Dynamics, and
an International Seminar), with a strong theoretical foundation and case studies
from multiple ecosystems around the world.
• Field/laboratory/hands-on application courses (Prescribed Burning Laboratory,
Restoration and Post-Fire Management, and Integrated Fire Management).
These courses involve learning how the social, cultural and environmental
context infuence the fre challenges and their practical solutions.
• Professional project, including design, implementation and evaluation.
These projects could beneft from the assistance of collaborators at partner
institutions in the USA (e.g. US Forest Service Fire Science Labs), in Europe
(e.g. Centre for Applied Ecology in Portugal), and agencies operating
internationally, including the Global Fire Initiative of The Nature Conservancy,
World Wildlife Fund for Nature, International Union for Conservation of
Nature and the International Association of Wildland Fire.
This Fire Paradox initiative could be the international framework for the fire
education of professionals of the future generation, taking advantage of all the
materials and the network provided by the project. Such an initiative should use
the global concept of Integrated Fire Management, and fully utilize the possibilities
provided by the different nature of the partners, applying the positive spiral started
in the triangle of Education, Training and Experience.
References
Andrews, P.L. and Bradshaw, L.S. 1990. RXWINDOW: defning windows of acceptable
burning conditons based on desired fre behavior. USDA For. Serv. Gen. Tech. Rep. INT-
273.
Binggeli, F. and Bento, L. 2009. Paradox – Living and working with fre – International
News Magazine for Professional Fire Fighters. Edition 1, April 2009. “Fire Paradox”
European Commission. 17 p.
CIFSC 2004. Mémento Permanent – Responsable de Chantier Brûlafe Dirigé. Centre
Interregional de Formation de la Securite Civile. Château de Valabre, Gardanne, France.
Colaço, M.C. 2005. Prevenção de Incêndios Florestais: Um estudo sobre a formação dos
técnicos forestais. Traballo de Investigación Tutelado, Programa Interuniversitario de
Educación Ambiental, Universidade de Santiago de Compostela, Departamento de Teoría
e Historia da Educación, España.
Correia, A., Colaço, C., Rego, F., Solana, J., Herrero, G., Montiel, C., Sesbou, A. and D.
Kraus. 2008. A review of the existing cooperation mechanisms between departments
dealing with fre management at the European and national level. Deliverable D7.1-4-1
of the Integrated Project “Fire Paradox”, Project no. FP6-018505. European Commission.
58 p.
EC 2007. Características principais do ECTS. DG Education and training. http://ec.europa.
eu/education/lifelong-learning-policy/doc/ects/key_pt.pdf
EC 2008. The European Qualifcations Framework (EQF). EC Education and training.
http://ec.europa.eu/education/lifelong-learning-policy/doc44_en.htm
FAO 2007. Fire management – global assessment 2006. FAO Forestry Paper 151.
Learning and Training on the Use of Prescribed Burning Techniques 173
Fernandes, P., Botelho, H. and Loureiro, C. 2002. Manual de formação para a Técnica do
Fogo Controlado. Departamento Florestal, UTAD, Vila Real. 55 p.
Freire, P. 2008. Teachers as Cultural Workers – Letters to Those Who Dare Teach, Translated
by Donoldo Macedo, Dale Koike, and Alexandre Oliveira, Westview Press, Boulder, CO.
Goldammer, J. G., Hoffmann G., Bruce, M, Konsrashov L., Verkhovets S., Kisilyakhov Y. K.,
Rydkvist T., Page H., Brunn E., Lovén L., Eerikäinen, K, Nikolov, N. and Chuluunbaatar,
T.O. 2007. The Eurasian Fire in Nature Conservation Network (EFNCN): Advances in the
use of prescribed fre in nature conservation, landscape management in temperate-boreal
Europe and adjoining countries in Southeast Europe, Caucasus, Central Asia and Northeast
Asia. 4
th
International Wildland Fire Conference, Seville, Spain. Pp. 13–17.
JRC-IES. 2008. Forest Fires in Europe. Report nº 9/2009. JRC Scientifc and Technical
Reports. 88 p.
Kobziar, L.N, Thode, A., Rocca, M., Dicus, C., Hoffman, C., Sugihara, N., Varner, J. M. and
Morgan, P. Challenges to educating the next generation of wildland fre professionals in
the US. Journal of Forestry 107(7): 339–345.
Kunst, C. and Moscovich, F. 1996. Fuego Prescripto. Introducción a la Ecología del Fuego
y Manejo de Fuego Prescripto. INTA and Universidad Nacional de Santiago del Estero.
Santiago del Estero. Argentina
Manso, F., Rego, F.C., and Binggeli, F. 2008. Fire in the balance (video documentary). Fire
Paradox. Lisbon.
Molina, D., Cabré, M., Pous, E.; Valese, E., Colaço, C., Rego, F., Sesbou, A., Kunst, C.,
Defossé, G., Lazaro, A., Herrero G., Solana, J., Rigolot, E., Kraus, D. and Delogu, G.
2009a. Survey of the existing academic and professional training materials related to
forest fre. Deliverable D10.1-1 of the Integrated Project “Fire Paradox”, Project no. FP6-
018505. European Commission. 52 p.
Molina, D., Cabré, M.M., Valese, E., De Ronde, C., Kunst, C., Defosse, G. and Loureiro. C.
2009b. Implemented set of demonstration sites for prescribed burning. Product P9.1-3 of the
Integrated Project “Fire Paradox”, Project no. FP6-018505. European Commission. 41 p.
Pimentel, C.S. 1882. Pinhaes, Soutos e Montados – Cultura, Tratamento e Exploração
D’estas Matas. Lisboa.
Pous, E., Molina, D., Cabré, M. and Blanco, J. 2010. Guidelines to mitigate personal risk in
prescribed burning and suppression. Deliverable D7.1-3-7 of the Integrated Project “Fire
Paradox”, Project no. FP6-018505. European Commission. 80 p.
Pyne, S.J., Andrews, P.L. and Laven, R.D. 1996. Introduction to Wildland Fire. 2
nd
ed. John
Wiley & Sons, New York.
Pyne, S.J. 1984. Introduction to Wildland Fire, John Wiley & Sons, New York.
Rego, F.C., Botelho, H.S., and Silva, J.M. 1989. Workshop on prescribed fre research
(Proceedings). UTAD/DGF/ESF (FERN) Commission of the European Communities. Vila
Real, Portugal.
Silva, J.M. da. 1990. Alterações profundas sofridas pelos ecossistemas forestais portugueses
e a sua relação com a incidência de incêndios - A Técnica do Fogo Controlado, UTAD,
Vila Real, Portugal.
Stephens, S.L. and Ruth, L.W. 2005 Federal forest fre policy in the United States.
Ecological Applications 15:532–542.
Tissot, P. 2004. Terminology of vocational training policy: A multilingual glossary for an
enlarged Europe. European Centre for the Development of Vocational Training (Cedefop),
Luxembourg: Offce for Offcial Publications of the European Communities.
Trabaud, L. 1992. Les Feux de Forêts, Mécanismes, Comportement et Environnement –
France-Selection, Aubervilliers.
Varnhagen, F.L.G. 1836. Manual de Instruções Práticas sobre a sementeira e corte dos
pinheiros, e conservação da madeira dos mesmos: indicando-se os methodos para o clima
de Portugal. Academia Real de Sciencias de Lisboa.
5. Suppression Fire:
a Tool to Control Wildfres
5.1 Overview of Suppression Fire Policies and
Practices in Europe
Cristina Montiel
1
, Pau Costa
2
and Moisés Galán
2
1
Research Group Forest Policy and Socioeconomics, University Complutense of
Madrid, Spain
2
GRAF, Department of Internal Affairs, Barcelona, Spain
5.1.1 Presence, diversity and complexity of suppression fre practices
in Europe
Suppression fre is defned as the use of fre to fght wildfres. This is an ‘intentional
application of fre to speed up or strengthen fre suppression action on wildfres;
types of suppression fring include burning out, backfring, line fring, counter fring,
and strip burning’ (FAO 1986). This technique for wildfre suppression has a long
tradition in Europe. Fire has been used by the local population as a wildfre fghting
tool for protecting their own lives and assets long before fre fghting services were
established in many European countries prone to fre hazards. In the recent past,
fre fghting agencies have also looked more in detail towards the use of fre as a
complementary tool to other fre fghting techniques because of the evolution of
wildfre hazard conditions.
Indeed, the current situation of forest ecosystems (overgrowth of vegetation) in
conjunction with unfavourable weather conditions (high temperatures, very low
humidity and strong winds) means that, in the worst situations, the magnitude of
wildfres overwhelms the extinction capacity of conventional means for fre fghting.
The velocity at which fames propagate and and the high intensity make wildfres
unstoppable. In this context, suppression fre is one of the methods of conducting
indirect attack, by planning the control of the fre fronts by surrounding the fre
with fuel-free strips, fre and control lines. Once the control stage is achieved, if
fre intensity allows, direct attack is continued otherwise performing backfring
techniques is advisable (Martínez 1997). Suppression fire is therefore not the
solution to extinguish a wildfre, but in the right circumstances suppression fre
supports containment and helps control the fre. Hence it is one more contribution to
extinction strategies.
Thus, suppression fre is a potentially powerful and very effcient technique,
especially in the event of large wildfres. In the Mediterranean area, it is used for
facing quick and intense fre fronts, and in Northern-Central Europe, suppression
fre is mostly directed to anchor extensive fanks that are diffcult to access or fanks
which overcome the capacity of the extinction means. Nevertheless, this technique
to stop wildfre propagation is not suffciently used in Europe. Its development had
an earlier start in Portugal and Spain (in the 1970s and 1980s), and more recently
also in southern France and a number of other European countries. Currently, the
use of suppression fre is mainly concentrated in southern Europe.
178 Towards Integrated Fire Management
The assessment of suppression fre practices is a very complex issue. It is often
difficult to distinguish between traditional use of suppression fire by the rural
population and its implementation by forest and civil protection services; in many
European countries suppression fre is used in a clandestine way. Suppression fre
therefore raises a complex paradox in Europe: it is an ancient and very effcient
technique in wildfre extinction, especially in the event of large fres, but its use is
limited, and often clandestine. The origin of this paradox has sociopolitical roots and
hence tackling and solving the paradox must involve social and political channels.
It is necessary to have a previous knowledge of the problem, identifying where and
under which conditions suppression fre is used, as well as the positive and negative
factors which have an infuence on its use. This has been the approach of Fire
Paradox to obtain an overview of suppression fre practices in Europe at national
and regional level, allowing identifcation of existing conficts and opportunities for
the use of suppression fre framed in an integrated fre management system.
The purpose of this chapter is to show the results obtained through the analysis
of policies and practices of suppression fre in Europe. Firstly, the spatial pattern
of suppression fre practices is displayed, as well as the diverse legal and political
approaches existent at the national and regional scale regarding suppression fre.
After that, the key elements and challenges for the use of suppression fire in
Europe are identifed, highlighting the importance of the local scale in planning,
management and decision making. Finally, recommendations are formulated
in order to promote the use of suppression fire according to the socio-spatial
circumstances of the European territories.
5.1.2 Introduction of suppression fre techniques in legal and policy
instruments
Large wildfres have been a major problem for Mediterranean countries since the
1990s. Simultaneous large wildfres occurring in “windows” of severe fre weather
takes turns from one country to another throughout summer. The episodes of large
wildfres recorded in Portugal, southeastern France and northeastern Spain (2003),
Portugal (2004, 2005), Greece (2000, 2007 and 2009) show how the behaviour
of large fres exceeds suppression capacity. Nowadays, large wildfres spread so
fast and with such energy, that it is necessary to have an extinction system enough
fexible to easily allow a change of priorities. Nevertheless, wildfre suppression
tactics are still almost completely dependent on direct attack methods, based on
water use and aerial support.
The system of fre prevention-extinction must adapt to the real necessities of the
problem. It must change from a total-attack strategy reacting to wildfre, towards
a confinement strategy based on the anticipation of fire behaviour, in order to
maximize opportunities in those places where fre is vulnerable. In this context, the
use of fre is coming back and claiming its role as a safe, effcient tactic, dependent
on time and place. Fire used as a suppression tool has been very successful in
fghting large fres, as it broadens the range of opportunities and speed in which
successful tactics can be performed. Also it offers a wide range of tactical options,
Overview of Suppression Fire Policies and Practices in Europe 179
from backfres or burnout operations, to perimeter defnition or linking suppression
actions up to an anchor point or a strategically defensible area.
There are many possibilities for using fre to fght fre that are at present misused
or neglected because of the loss of fre culture and knowledge in European countries,
and also because of the fre suppression policies and regulations. Suppression fre
practices are used extensively but insuffciently regulated in southern Europe. On
the other hand, it is diffcult to distinguish between the use of suppression fre by
rural population as a fghting tool to defend their territory
1
, and the professional use
by fre fghting services as an extra tool for wildfre suppression. This is one of the
reasons why it is so diffcult to fnd information about the professional, regulated
use of suppression fre in Europe (Lazaro et al. (2008) (see Figure 1). Furthermore,
even though suppression fire is a well known practice in Europe, national fire
polices have not taken this technique into account from the beginning and not all
the countries have a legal or technical defnition for it. That means that fre use
in wildfire suppression is generally not regulated, except in Catalonia (Spain),
Portugal, France and recently in Sardinia (Italy) (Herrero et al. (2009).
Obviously, each country has a different fire problem and needs different
regulations on fre suppression, fre prevention, or the use of fre in suppression
or in prescribed burning. In general, the existing regulations endorsing the use of
technical fre as both a tool for management and wildfre extinction require the
necessity of setting out a change as for large wildfre management. The concepts on
which technical use of fre is based, and the qualifcations required to perform duties
involving these techniques need to be laid down. Additionally, these regulations
also usually defne the main applications for suppression fre. For instance, the
Catalonian Decree
2
includes legal defnitions for two types of suppression fre:
• Backfring as ‘an extinction manoeuvre carried out with technical fre with
the aim of eliminating and/or shifting oxygen, verticalizing the column and
achieving that secondary spots fall within the burned area hence halting the
advance of the main fre front’.
• Burning out as ‘an extinction action performed by means of technical fre,
pursuing the aim of removing vegetation fuel in those areas in potential danger
of being burnt by the uncontrolled advance of wildfres’.
5.1.3 Key elements and challenges for suppression fre practices
European countries and regions must have their own fre task forces ready to deploy
all parts of their territories with a complete set of tactics in the case of wildfre,
including indirect attack using suppression fre. There are however a number of
constraints which hinder the use of fre as an extra-tool for wildfre suppression,
amongst which we can fnd the following:
1 This practice is forbidden in European countries, meaning the igniter of a fre is identifed as an arsonist who sets fre
during the summer period. Lighting a fre, even when the intention is to extinguish another one, may be considered a
crime.
2 Decree no. 312/2006. BOGC nº 4685 of 27 July 2006.
180 Towards Integrated Fire Management
a) This technique requires support from local forces, which usually are neither
reliable nor fexible enough to perform joint manoeuvres with personnel of
higher administrative levels;
b) The loss of culture of fre or the lack of social consensus to carry out the use
of fre can result in counterproductive social behaviour;
c) The conficting position of land-owners regarding backfring in emergency
situations;
d) The civil responsibility of the Incident Commander leading wildfre
suppression actions and taking the decision to use suppression fre techniques
or not.
Therefore, the opportunity and capacity to perform indirect attack methods for
wildfre suppression does not only depend on regulations and their implementation,
but also on developing planning and management actions at the local level and
beyond.
Figure 1. Suppression fre use in European regions and year when the professional practice
began.
Overview of Suppression Fire Policies and Practices in Europe 181
Planning and management at the local level:
The effective and socially accepted use of suppression fre in Europe depends on a
string of spatial and socioeconomic factors that infuence the existence and nature of
fre use practices in the different countries and regions. These include:
a) Landscape conditions: the traditional use of fre by rural population as a
fre fghting tool was based on a mosaic landscape with a well-maintained
path network and clustered houses, that is, the maintaining of the traditional
Mediterranean agro-silvo-pastoral socio-spatial system (see Figure 2).
b) Know-how: suppression fre practices are to a large extent sustained in
traditional fre use knowledge. Fire culture and collective experience were
essential for the traditional use of fre. Nowadays we can speak about
knowledge of fre behaviour and professional experience.
c) Territory appraisal and awareness: knowledge of the territorial
characteristics-, such as topography, vegetation structure, distribution of
habitats and infrastructure, path conditions, etc. (see Figure 3).
Figure 2. Cork oak slope with sweet chestnuts stands in Maures mountains (municipality
of Collobrières, France). The application of suppression fire techniques require certain
territory conditions in order to manage fre by using anchor points and areas with different
combustibility: mosaic landscape structure (mountain topography and discontinuity in forest
fuels), absence of isolated settlements and a well-maintained path network. (photo by C.
Montiel).
182 Towards Integrated Fire Management
Consequently, the introduction and implementation of professional and technical use
of suppression fre requires a protocol of key actions which should be implemented
at the local level. Suppression fre is a local practice, in fact one that people only
used within the boundaries they were familiar with and which constituted their
heritage. That is the reason why it is so diffcult to frame suppression fre practices
within general policies.
In those regions where fre culture survived, local experts have knowledge of the
territory (relief, paths, local wind, vegetation, assets, etc.) and historical knowledge
of fire behaviour in that territory, but their objectives are not the same as fire
fghting services. Their focus is to defend local assets; they do not have the holistic
strategic approach of suppressing wildfres, though the intention to reduce damage
to the landscape is common to both. In such regions there are the most numerous
and the most permanent experts in fre use with a kind of legitimacy in the local
territory because of their property and land tenure rights. Therefore it is necessary
to: (i) establish a dialogue between the stakeholders involved in suppression fre
use; and (ii) normalize the use of fre at the local scale, defning the duties of each
stakeholder and reaching a consensus to allow collaboration. The fnal aims are
to regularize suppression fre practices and to defne a successful way for wildfre
suppression cooperation at the local level.
Figure 3. Second front of the backfre practiced in La Malière (municipality of Collobrière,
France) during the wildfre of 1990: the asphalt road and the herbaceous area of the valley
were the anchor points of the backfre which progressed up-hill, under the chestnut stand,
towards the fre front (photo by C. Montiel).
Overview of Suppression Fire Policies and Practices in Europe 183
Governance mechanisms and civil responsibility exercise
Fire fghting strategies in Mediterranean countries have evolved through the second
half of the 20th century from a defence system based on the organization of local
populations and the use of their own resources to a professional system run by
forest and civil protection services (Castellnou 2010). This enabled the adoption of
a strategic approach and means of suppression to address the increasing signifcance
of wildfires (Dujas and Traimond 1992). However, socio-spatial evolution and
changes in fre scenarios mean that new ways of collaboration are needed which
consider the organized participation of certain sectors of the population in
prevention and suppression duties (Aguilar and Montiel 2009).
Suppression fre techniques are a very effective and powerful fre fghting tool, but
its local and technical application require the availability of responsible, qualifed
people with technical skills and good relations with the local population, as well as
to an extent the maintenance of traditional social-spatial structures. Therefore, fre
policies should consider these aspects when regulating suppression fre use and also
the establishment of the corresponding governance mechanisms.
In those municipalities with a forest culture, there is a need for empowerment
of local groups (Aguilar 2003). Land owners are willing to collaborate with fre
fghters and demand their right to take part in extinction duties, on the basis of their
knowledge and legitimacy on the territory. Moreover, by demanding their right
to get involved in fre fghting, locals also claim a certain control of fre, which
somehow is a symbolic expression of territorial power.
Social participation in policy-making and application of policy is less frequent
in Mediterranean countries than in other EU member states. Another important
element associated with governance, which also applies to forest policy, are learning
processes that operate through diffusion mechanisms or networks. Expertise in
fre use is a crucial resource, but also the revival of old and wise, yet abandoned,
fre practices that were in hands of rural communities (Aguilar and Montiel 2009).
Furthermore, learning processes and cooperation should be carried out in a context
of confdence, and with a will to cooperate; this will allow the legal and policy
framework for a regulated and effective action to be formulated (see Figure 4).
On the other hand, suppression fre practices entail a responsibility which slows
down its application in operations due to the risks associated with it (risk of
fatalities, material losses, etc.). Current urban societies are not willing to forgive
mistakes, and hence it is necessary to develop a legal framework as well as an
appropriate suppression fre plan. The effectiveness of technical fre depends on
determination and immediate action, which means that it is not susceptible of
detailed prior planning. In fact, professionalization and normalization of suppression
fre practices will likely reduce the effectiveness of traditional practices, although
appear to increase safety. In any case, it is necessary to clearly regulate the civil
responsibility of duties involving technical fre, and to do that defning the roles
played by different stakeholders is essential.
Professional training:
Suppression fre is a very challenging practice which requires in-depth knowledge
of fre, ecosystems and the relations between them. Fire users need a high level of
knowledge about fre behaviour and also about fre fghting. Theoretical knowledge
184 Towards Integrated Fire Management
must be complemented with practical experience which enables prediction of fre
propagation behaviour, including the interaction between fre fronts.
Expertise and experience are needed for the responsible and effective use of
suppression fre. Regarding that, professional training concerning the use of fre as a
tool for fghting fre involves the formalization of required competences (theoretical
and practical), adapted to each territorial context. For instance, the use of fre in
North Atlantic areas should face medium-low intensity burnouts without convection
processes while, in Mediterranean areas, training must focus on convection
processes. So, training to perform fre suppression actions means different levels of
capacity for different types of fre.
If Europe is to respond to the fre problem in a forceful way, it should take into
account all the possibilities offered by different extinction systems. This includes
training professionals to acquire appropriate qualifcations in order to offer to the
different countries qualifed teams that are fexible and ready to give a measured,
organized response where they are needed. That is, a strong training programme must
be developed, with progressive skill levels, which enables a smooth harmonization of
the different national and regional existing training programmes in Europe.
It would be therefore interesting to develop a European integrated qualifcation
system, which could be further developed in each country to match with the
different political, administrative or cultural systems, customs and values.
This would enable certification of the acquired qualifications and would make
professional exchange and cooperation easier between countries and regions. The
Fire Paradox initiative ‘recording system and identifcation for the professionals in
the fre use’ aims to promote a European qualifcation system for fre management,
based on the formalization of training protocols that would facilitate international
cooperation and would enhance the global interoperability of emergency managers.
Figure 4. Scheme of an innovative and participative governance mechanism for the use of
suppression fre.
Overview of Suppression Fire Policies and Practices in Europe 185
5.1.4 Conclusions and recommendations
The use of suppression fre is a social, political and technical challenge. It enables
responsible agencies in cooperation with actors at the local level to act upon areas in
which the use of conventional means of fre fghting would involve certain danger,
as well as a rapid intervention upon areas where there is diffcult access and limited
human resources. It constitutes a complementary strategy to other fre fghting
techniques and an opportunity to fght large wildfres, although its implementation
requires previous social acceptance as well as regulation of fre use.
The constraints on the use of suppression fre are social acceptance and legal
regulation, and are very complex: on one hand, the confictive position of land
owners regarding backfiring in emergency situation; on the other hand, the
civil responsibility of the Incident Commander, when taking the decision to use
suppression fire techniques, based on the existing legal framework. Moreover,
there is often confusion between the traditional use of suppression fre performed
by rural population and its implementation by forest and civil protection services,
which strengthens the clandestine character associated with this technique in many
European countries.
Regarding the basic elements to be considered for the formulation of the legal
framework for this practice, the following aspects are highlighted: (a) suppression
fre should be carried out even if the land owner does not provide authorization, due
to the emergency situation; (b) the Incident Commander is responsible for taking the
Table 1. Proposal for a European fre fghting training framework.
Level Available operations
1 Basic Level Competencies:
• Backfres and burnout operations.
• Defnition of a special perimeter by means of burnout operations.
• Safety improvement in fre management operations.
• Combined operations with fre and other tools.
2 Additional Competencies:
• Carry out ignitions to manage vegetation fres. That is a greater ability to
manage each single drip torch, with a broader range of objectives and techniques
to manage wildfres. Such as limit suppression actions to an anchored point or a
design unit.
• Greater ability in combined operations.
• Track opportunities associated to anticipated changes of fre behaviour to
propose tactics to prevent and fght wildfres in Mediterranean ecosystems.
3 Additional Competencies:
• Control of operations with technical fre to manage wildfres in Mediterranean
ecosystems. So, greater ability to manage groups of drip torches combined with
other tools, with a broader range of objectives.
• Direct or delay fres spread to reach specifc resource management objectives.
• Prioritize opportunities associated with anticipated changes of fre behaviour
to propose strategies and tactics to prevent and fght wildfres in Mediterranean
ecosystems.

186 Towards Integrated Fire Management
decisions, and his liability should be covered by the responsible administration; (c)
the execution of a technical fre should have some minimum and general regulation,
as well as the establishment of common techniques, but it cannot be pre-conditioned
to a fxed and detailed method; and (d) technical knowledge is required and it must
entail both research and training efforts.
Cooperation between different stakeholders through learning processes and
participatory mechanisms are also badly needed for the use of suppression fre.
Learning and drawing lessons, know-how transfer and training schemes are required
in order to solve the social conficts that are constraining the use of fre as an extra
tool to fght wildfres.
The use of fre requires the availability of responsible, qualifed people with
technical skills, and well established communication with the local population, as
well as an insight into traditional socio-spatial structures. Therefore, fre policies
should consider the need to regulate the use of suppression fre and also to develop
professional training and to establish the corresponding governance mechanisms.
Furthermore, the use of fre as a tool for wildfre prevention or suppression needs
to be designed, applied and evaluated. All this makes widespread application of
suppression fre techniques diffcult. It is a matter for experts on fre behaviour and
fre fghting systems, and consequently the use of fre must, by defnition, be in few
hands.
With regard to the authority of landowners and other people who hold rights over
land affected by wildfres, it must also be determined whether any of the persons
holding property rights may prevent or restrict the use of suppression fre. As for
the intervention of the authorities, the examination of the regulations reveal that
at least the reference to administrative reports is meaningless because the Incident
Commander cannot wait for an answer to act. In those circumstances where
recourse to suppression fre is necessary it does not seem feasible to depend on
the issue of any offcial permission. Likewise, the need to notify the corresponding
authority about the execution of suppression fre seems rather unfortunate, though
understandable, because the participating agents are responsible for the intervention
and act as representatives of that authority. This is the reason why, in the few cases
in which this matter has been regulated, the execution of suppression fre techniques
is considered a power that rests with the Incident Commander. In fact, because
of the speed needed to act in these cases, the state of necessity that may lead to
the performance of such a risky and dangerous operation and the unpredictable
and changeable nature of fre behaviour, it should be the Incident Commander who
makes the decision to use suppression fre, without prejudice to a legal provision,
and with timely notifcation to the authorities.
Without doubt the issue is not only about legislation and application of laws.
Regulations must be accepted by the whole society and, in order to achieve that,
political efforts are required to understand the problem from the viewpoint of
professional forums and along with other social agents. It is likewise necessary to
set measures involving education, social awareness and above all, to establish a
standardized training system for all the professionals of the sector (extinction task
forces, personnel involved in prevention duties, etc.), which integrates each and
every method offered by extinction techniques.
Overview of Suppression Fire Policies and Practices in Europe 187
References
Aguilar, S. and Montiel, C. 2009. New governance strategies for wildland fre management
in Mediterranean countries. In: Buttoud, G. (ed.). Pre Conference Proceedings of the
International Symposium Change in Governance as Collective Learning Process:
Management, Politics and Ethics in Forestry, Nancy, France. Pp. 9–18.
Aguilar, S. 2003. Environmental confict resolution: litigating patics having to face
environment decisions. Observatorio Medioambiental 6: 25–33.
Castellnou, C. 2010. EL Programa Catalan de gestión del Fuego: El GRAF en el Cuerpo
de Bomberos de la generalitat de Catalunya. In: Montiel, C. and Kraus, D. (ed.). Best
practices of fre use. Prescribed burning and suppression fre programmes in selected case
studies regions in Europe. EFI Research Report 24. European Forest Institute. In press.
Dujas, J.M. and Traimond. 1992. Le maitre du feu. Les incendies dans les landes de
Gascogne. Terrain 19: 49–64.
FAO 1986. Wildland Fire Management Terminology. Food and Agriculture Organization of
the United Nations, FAO Forestry Paper 70. 257 p.
Herrero, G., Montiel, C., Agudo, J. and Aguilar, S. 2009. Assessment document on the main
strengths and weaknesses of the legislation and policy instruments concerning integrated
wildland fre management in the EU, in European Member States and in North African
countries. Deliverable D7.1-1-2 of the Integrated Project “Fire Paradox”, Project no. FP6-
018505. European Commission. 117 p.
Lázaro, A., Solana, J., Montiel, C., Goldammer, J.G., Kraus, D. and Rigolot, E. 2008.
Collection, classifcation and mapping of the current prescribed fre and suppression
fre practices in Europe. Deliverable D7.1-3-1 of the Integrated Project “Fire Paradox”,
Project no. FP6-018505. European Commission. 47 p.
Martínez Ruíz, E. 1997. Manual del contrafuego. El manejo del fuego en la extinción de
incendios forestales. TRAGSA, Madrid.
5.2 Suppression Fire Use in Learning Organizations
Marc Castellnou
1
, Daniel Kraus
2
, Marta Miralles
1
, Giuseppe Delogu
3
1
GRAF, Department of Internal Affairs, Barcelona, Spain
2
Fire Ecology Research Group, MPI Chemistry, Freiburg, Germany
3
Corpo Forestale e di Vigilanza Ambientale, Cagliari, Italy
5.2.1 Introduction
Fire is a natural part of European landscapes both because of natural lightning
ignitions and by human activity for hunting, fghting or land management. However,
fre has recently started to be regarded as an economic and social problem. The
cost of fre suppression is higher than ever, and large fres are becoming larger,
with faster spread rates and higher fre intensities. The landscape is changing to a
continuous mosaic of fuels, as most small and medium-size fres are extinguished
as soon as they develop. The consequences of this fast growth in fire fighting
organizations only focused on fre suppression are increased fuel accumulation
and increased fuel continuity at the landscape level. As a result wildfires can
become non-containable and the proportion of area burned in such big wildfres
will continue to grow. During the past century, this problem has developed in
different ways throughout Europe. However, if we look at this problem from a
global perspective, we can only come to the conclusion that fre is becoming a more
important problem. There is a clear increase in the number and proportions of so
called megafres in countries such as Portugal, France, Spain, Italy, and Greece, but
even in England, Sweden and Norway large wildfres are occurring simultaneously.
The trend observed in Europe is that wildfres larger than 50 ha affect more than
75% of total surface burned in Europe (San-Miguel and Camia 2009).
Europe is now faced with a new fre-related problem all over the continent, and
trying to enforce a fre suppression policy is not going to solve the problem. In this
sense, suppression fre techniques are not only an opportunity to improve effciency
to fght these large fres that are beyond the capacity of extinction, but they are also
linked to skillful, experienced and trained groups of fre specialists that are able to
introduce learning in organizations. The aim of this chapter is to explore the deep
changes in fre management organizations that are associated to the introduction
of suppression fire techniques, and this calls for new concepts not only in fire
management but also in organizational learning.
5.2.2 Key concepts and a new strategy for wildfres
Since most EU states are still trying to enforce a fre suppression policy, it is evident
that the situation all over Europe is simply not going to be solved by the use of more
190 Towards Integrated Fire Management
resources or a more aggressive direct fre suppression strategy. There is a need to
change the strategy towards reinforcing fre as a suppression technique. It is necessary
to focus not only on improving individual training, but especially on improving the
way organizations are learning and applying lessons learned from experience. So, two
new concepts arise as a part of a global solution all over the world: suppression fre
and learning organizations are both key concepts in the future process of change.
Any fre used as a suppression technique during uncontrolled fres is considered
suppression fre, where the main objectives and tactical options can be summarised
as follows:
• Directing or delaying fre spread;
• Mitigate re-ignition risks;
• Improving the safety of the public and fre fghters;
• Always to be more effcient and safer in direct suppression fre;
• This type of fre application can provide a range of applications, from burning
out or backfring operations to limiting suppression actions to an anchor point.
Sometimes the most effective fre tool that will strengthen suppression of a large
fre is to reduce the fre intensity of the fre front, so that an anchor point can be
established (see Figure 1d). This can be the most successful tactic to mitigate re-
ignition risks by avoiding an increase of fre intensity through a fring operation near
a critical point.
A learning organization is skilled in creating, acquiring, interpreting, transferring
and retaining knowledge, and adapts to reflect new learning (Alexander and
Thomas 2003). This means to become a learning organization; new knowledge, new
opportunities and new diagnoses must be constantly included in the organization
(Nasiatka 2008). This is an important prerequisite for the use of suppression fre
where the scale of consequences of a high risk operation precludes learning through
experimentation and the potential for failure is high.
5.2.3 A new strategy for wildfre
In order to cope with a worsening wildfre situation, there is a need to change the
strategy towards the following:
• changing the landscape by managing fre regime itself to solve the problem,
i.e. a change from suppressing all fres that lead to a regime of high intensity
fres towards a new approach based on tolerating a fre regime with low
intensity fres;
• adapting fre-fghting organizations to cope with low frequency, high impact
large fre situations through building of knowledge and experience .
In this context, introducing fre use techniques in fre fghting organizations is the
frst step towards a change of the core values in fre services. Managing fre use
techniques are basic to managing landscapes, and as high risk techniques these must
be grounded in a modern learning organization.
Suppression Fire Use in Learning Organizations 191
5.2.4 Evolution of fre management
The historical evolution of fre management in Europe is key to understanding what
policies and paradigms have been used by countries as the fre scenario evolved. In
landscapes with a high proportion of large fres, the process of understanding fre
and reaching a fre management situation is a step by step process.
Phase 1. Fire exclusion
Traditional use of fire has been substituted by prohibition, centralized from
governments during the 20
th
century. Initially fire was used by shepherds,
gamekeepers, rangeland managers or villagers for local interest (Lázaro et al. 2008).
Figure 1. Multiple-use of suppression fre: it does not only stop the fre before it reaches
critical points from where fre behaviour would be worsening, suppression fre can also
reduce spotting distance through raising the smoke column of the advancing fre with its
convection. In 1a and 1b, suppression fre reduces the perimeter of the head of the fre through
drawing the one head into the other. It can also be used to ensure that no fare-ups develop, by
means of anchoring the head fre front. This can be done by igniting a freline from the anchor
line (1c), or by means of reducing fre intensity before the fre reaches the anchor line, (1d).
1a + b. On the left, the fre has two main heads following the ridges. On the right, a burn out
has called the lower head into the upper one. The result is a single head of the fre and a fank,
where before there were two main heads.
1c. Ignition from anchor line where fre will
be held.
1d. No ignition is done in a fre front close to
the anchor line, but also it is not extinguished.
The only action performed is controlling
intensity; by doing that, the front is well-
anchored, reducing the risk of re-ignitions or
fare ups.
192 Towards Integrated Fire Management
As the socio-economic context changed, the techniques used by rural villagers to
deal with productivity, regeneration of pastures, or fre risk management changed.
The perception that society had of fire changed. Fire became the enemy that
threatened the landscape and the ecosystem. Consequently, the use of fire was
prohibited (Seijo 2009),
However, policies based on extensive fre suppression began to fail in different
parts of Europe, and were leading to an increase of relative importance of large fres.
Several recent cases substantiate this paradox: Catalonia (1986 to 1993), Galicia
(1994 to 2005) and Greece (2001 to 2006) were regarded as examples of success
and in each case catastrophic fre seasons followed those periods (Rigolot et al.
2009). Slowly, fre´s role in managing the ecosystem and in creating the landscape
began to be understood.
Phase 2. Prescribed burning is introduced
Prescribed burning is not only an important tool for nature conservation, but also
a technique useful for wildfire prevention in different ways: as a basis to train
and maintain the know-how of using suppression fre techniques, and to create a
fuel mosaic in the landscape, which, when strategically placed, is able to provide
opportunities to fght wildfres. This especially happens when the mosaic is formed
by patches with less fuel to burn and with forest stand with less capacity to develop
crown fres. (Finney et al. 2005; Rigolot et al. 2009; Cassagne et al. 2009). The
number of countries performing prescribed burning is increasing, both for fire
prevention purposes (Portugal, France, and north of Spain) or other land management
objectives (Sweden, Scotland, Germany, and parts of Spain) (Lázaro et al. 2008).
Phase 3. Suppression fre is re-introduced
When large wildfres overcome extinction capacity (become uncontrollable), they
can have a serious impact on the economy and may put populations at risk. Old
ways of fghting fre are remembered, and fre application can become a “technique”
Figure 2. The slow descending, low intensity fre (dotted arrow) allowed the survival of trees,
but also provided the opportunity for fre extinction. Subsequently, another fre started at the
bottom of the valley, and exposed every tree to signifcantly higher fre intensity (continuous
line). The suppression of the low intensity fre front which the trees survived allowed the
development of a high intensity fre which killed all the trees. Source: GRAF, Catalonia.
Suppression Fire Use in Learning Organizations 193
to fght and control wildfres. In Portugal, Spain or Cyprus suppression fre was
introduced during the mid 20th century as a tool to fght continuous and large fre
perimeters. The large fres in the decade since 2000 have added some countries
to this short list, but in those cases fre is used by highly trained fre specialists:
Catalonia (2000), France (2005), Sardinia (2006) and Portugal (2004) (Lázaro et al.
2008; Rifà and Castellnou 2007).
Phase 4. Managing natural ignitions
In areas very prone to large fres, the introduction of medium fres in a signifcant
proportion in the landscape reduces the proportion of large fres (Piñol et al. 2007).
The extension of surface area to be burned is recognized as a necessity, but also
as very diffcult and expensive to achieve (State of Victoria 2003; Nasiatka 2003;
Van Wilgen 2002; Cassagne et al. 2009). As the objective of fre extinction is not to
reduce the surface area burned, but to reduce the negative impacts of wildfres, the
opportunity to manage natural fres for that purpose seems a logic step. However,
there are socio-political constraints. Wildland Fire Use policies appeared associated
to natural ignitions in large natural landscapes, when the fre suppression cost is
perceived as higher than the perceived risk associated with managing unplanned
ignitions. Examples can be found in the USA in the 1960s and 1970s (Van
Wagtendonk 2007) and in South Africa in 1992 (Van Wilgen 2002).
Phase 5. Managing all ignitions
In a densely populated Europe, where natural and cultural landscapes cannot be
separated, Portugal is initiating a new path. Each part of a wildfre is managed
according to its resource objectives, whatever caused the fre,. The purpose of fre
fghting operations goes beyond "minimizing surface area burned", and considers
also land use objectives or ecosystem sustainability. This shift from fre fghting
to fre management has clear ecological and economic benefts in landscapes with
a high proportion of large fres and with prescribed burning and suppression fre
programmes in place. It will be a very slow process, due to perceived risk, social
acceptance and the need of different services working together: emergency services,
forest services, fre services and landscape planning services.
5.2.5 From resources to knowledge
This shift from focusing on fre exclusion towards fre management implies also a
shift in investment, from "more resources" to "knowledge". Now, in the ‘large wildfre
era’, resources are still needed, but knowledge of fre behaviour and of operations
is fundamental (Sagarzazu Ansaldo and Defossé 2009). Identifying the main factor
that is allowing fres to escape control in each region (see Box 1), and the lessons to
be learned in each experience allows fre-fghting organizations to be effcient and to
progress towards appropriate fre management policies, strategies and tactics.
Fire Reports and Case Studies
These are essential to learn from own experience and past history (Garvin 2000).
Each fire provides an opportunity to explore and adjust new operations, new
194 Towards Integrated Fire Management
strategies, new tools, and to pass this information on to others, supplementing and
strengthening a person’s experience (Alexander and Thomas 2003). So, it is not only
an organization’s task to create a platform to keep records, it also means to track
the results of decisions taken, and analyze these results in ways that reveal their key
lessons (Alexander 2002; Sagarzazu Ansaldo and Defossé 2009).
Transferring knowledge throughout the organization (Garvin 2000).
The objective is to anticipate problems before a wildfre occurs and to discuss
lessons learned after a fre (same day and during the next season), or after a whole
Figure 3. The same synoptic weather situation implied different patterns of fire spread
depending on the macroscopic relief. The frst two fres occurred near the coast in Banyoles
(a) and Masquefa (b). They are displaying a rounded shape, and their perimeters are affected
by the daily shift of land and sea breezes. Meanwhile the fres near Les Avellanes (c) and Sant
Pedor (d), occurred in the interior. Their perimeter patterns are not affected by sea breezes
and they have an elongated shape. Understanding these patterns has a great infuence on the
selection of tactics.
a
b
c
d
Suppression Fire Use in Learning Organizations 195
fre season (Castellnou et al. 2007). This is particularly applicable when promoting
fre analysis training through case examples throughout the whole organization.
Exchanging information and personnel with other countries, regions and
organizations.
This should include especially decision-makers, stakeholders and fre specialists
of countries, regions or organizations with similar present or future problems. This
will not only improve operational procedures, but will also enable learning from
best practices of other organizations allowing to include local or foreign expertise,
and to increase the range of possible solutions to similar problems (Garvin 2000;
Castellnou et al. 2008). The creation of platforms to exchange information (e.g. web
platforms and web forums for fre professionals) is fundamental.
5.2.6 From individuals to organizations
The last step after understanding the road towards management and understanding
the new focus on knowledge, is looking at people not as workers, but as information
keepers. Putting all this together allows a global vision of what is going on, what
is coming and what is needed. The whole team should take part of the process of
decision-making, each one at its level of responsibility. Individuals are keepers
of the information which they acquire and which they create and this is the most
valuable asset.
Accordingly, including fre intelligence at different levels of the organization
(Garvin 2000) means to train all personnel at different levels of fre behaviour
analysis. Furthermore, it also means consolidating Fire Specialist groups with a high
capacity of analysis (see chapter by Molina et al.). Thus new job positions must be
promoted, i.e. the Firing Specialist and Firing Operations Leader at the fre front,
the Firing Tactics Commander in a fre division, and the Fire Behavior Analyst for
the whole of a fre (see chapter by Miralles et al.). The ability of any organization to
listen to experts is the key: experts who can apply new techniques in new situations,
who are able to recognize patterns, to separate noise from a real signal, with long-
time, landscape vision, recognized by their own people, recommended by other
mentors (Saveland 2008).
The use of fre during fre suppression operations can become a complex task,
since it adds another risk factor to the regular suite of risks in fre fghting. In
complex emergencies, any organization must respond in a fexible, scalable way
through a standard management hierarchy and standard procedures in order that
there is a common framework within which people from different organizations
can work together effectively (such as the Incident Command System, developed
in the USA). The attitude and mental model of providing standards, policies and
directives is an important part of the issue, but it is only part of the story (Saveland
2008). In complex systems – operating in a chaotic environment where the need to
manage the unexpected arises – the organization does not only require the structure
of command and control. It also needs a learning culture, to enhance and sustain
safe and effective work practices.
196 Towards Integrated Fire Management
Box 1. Identifcation of different types of wildfres escaping
extinction control (fre generations) and best strategies and
response from different organizations.
Large fres can be uncontrollable for different reasons: from continuous surface
fres, to fast spotting surface fres, to crown fres and wildland-urban interface
crown fres, to simultaneous wildland urban interface large-fres (Table 1).
Obviously different types of large fres, with different socio-economic contexts,
need different responses, and in Europe we have seen and experienced multiple
extinction systems, with different regulations, different competences (forest/fre
services), and even different fre fghting tools (e.g. suppression based on use
of hose lines, or on use of defence lines). These differences in response to fre -
used by different organizations and regions working together in fre suppression
applications - provide a great capacity to adapt to a changing environment.
This is the most important decision: to identify exactly what is the fre situation
(fre generations: Table 1) and to adapt a response to it in a local context (but
taking into account the lessons learned by others that have experienced similar
situations).
New large wildfres and global seasonal change are forcing the identifcation
of specific requirements for each area and to respond to them, with
organizations that are capable of adapting to local conditions but with the
capacity to learn and share experiences with others.
The response and strategy to fght fres, and thus the type of fre specialist
requires some changes. These specialists have to adapt to a range of large fre-
types, and the socio-economic pressure of extinction of such fres is serious.
Thus expert decisions need to be taken at different scales (front, sector or whole
fre area), from different levels of expertise (managing one drip torch, or a
group of torches, or managing fre interaction as a whole).
However, the more investment in ‘response capacity’, the more fuel will
accumulate and a new generation of large fires will be created. So the fire
extinction response must be accompanied by measures to reduce fuel load,
continuity, vulnerability of the forest, and measures to promote self-protection
of people and infrastructures.
Suppression Fire Use in Learning Organizations 197
Table 1. Description of generations of large fres with potential to escape the capacity of
extinction due to: fuel continuity, rate of spread, maintained crown fre behaviour, high in-
tensity wildland urban interface (WUI) spread and simultaneously occurring large fres. Best
strategy and response from different organizations and the type of fre specialist (see chapter
by Miralles et al.) involved.
Landscape
Fire behaviour
Strategy
Response
Fire specialist
groups involved
Operations
available
Landscape
Fire behaviour
Strategy
Response
Fire specialist
groups involved
Operations
available
1 – Fuel continuity
Continuous fuel distribution allows long perimeters. Farmlands
abandoned.
Surface fres consume fne fuels, and spread on dry grasses and/or
litter. Mainly wind-driven fres with lack of anchor points.
Direct attack, based on local resources.
• Local response, reinforced with seasonal fre fghters.
• Linear infrastructures, accessibility and building water points.
BRIF (Spain). Basic fre specialists (technical fre support, level 2)
with non-specialized unit commander (level 4):
• Backfres and burn out operations, especially perimeter defnition
with burn out operations;
• Improve safety in fre management operations;
• Combined operations.
2 – Rate of spread
Fuel build-up allows faster spread rates of fres and spotting.
Wind- and topography-driven fast surface fres, mainly spreading in
shrubs and tall dry grasses, with occasional crown fres.
The speed of the fres overruns the holding lines.
Fast dedicated and effcient attack, based on local expert input and
aerial support.
• Dense detection and suppression net to ensure a fast and powerful
dispatch of fre engines/crews and helicopters.
• Wider range of combined suppression techniques: fre, hand tools
and machinery, hose lines and aerial attack, combined operations.
• Fire analysis to anticipate small-scale opportunities
Example GRAFF (France). Wildland fre ignition specialists
(level 3) with technical fre support (level 2) and non-specialized unit
commander (level 4).
Additional competencies:
• Tracking opportunities associated to windows of opportunities;
• Greater ability to operate drip torches;
• Broader range of objectives and techniques to manage wildfres;
• Greater ability of combined operations.
198 Towards Integrated Fire Management
Table 1. Continued.
Landscape
Fire behaviour
Strategy
Response
Fire specialist
groups involved
Operations
available
3 – Maintained crown fre.
Fuel build-up to crown continuity results in high intensity active
crown fres with convective plumes that are beyond capacity of
control.
Active crown fres and long spotting distances during extreme heat
waves, with very few opportunities to control.
Fire changes behaviour faster than information is passed on in the
chain of command.
Slowing down the fre front and confnement strategies.
• Reinforcement of logistics units.
• Directing or delaying fre spread to reach specifc resource
management objectives.
• Introduce fre regime concept in forest planning.
• Include fre analyst to anticipate opportunities, increasing the
adaptability of the organization to changes of fre behaviour.
Example GAUF (Portugal), GRAF (Catalonia). Wildland ignition
specialists (level 3) with technical fre support (level 2) and technical
fre specialist (level 4). Supervised by a Fire Analyst.
Additional competencies:
• Directing or delaying fre spread to reach specifc resource
management objectives;
• Prioritizing windows of opportunities;
• A greater ability to manage groups of drip torches combined with
other techniques, with a broader range of objectives.
Landscape
Fire behaviour
Response
Strategy
Fire specialist
groups involved
4 – WUI fres
The WUI becomes involved in the forest fre environment.
Residential and industrial areas are increasingly affected by wildfres.
Fires that can start and be stopped inside the WUI.
New landscape situations are forcing a change from attacking fres to
defending houses and people in a new defensive situation.
• Fire specialist groups as mobile units inside the residential areas,
trained to take tactical decisions on the basis of general strategies.
• Simulators, GPS and mapping technologies to track resources in
real-time.
• Professional logistics units.
Lowering tactical decision level, and command taking strategic
decisions
Same fre specialist as for maintained crown fre generations.
Suppression Fire Use in Learning Organizations 199
Table 1. Continued.
Landscape
Fire behaviour
Response
Strategy
Fire specialist
groups involved
5 –Large fres.
Zones at risk are faced with simultaneous large and fast spreading,
extremely intense wildfres.
Simultaneous crown fres involving wildland urban interfaces, mainly
during heat waves.
• New skills are needed to respond to simultaneous large fres.
• The answer is sharing resources, on a larger scale. This implies:
• standardized training systems;
• logistics organization for regional assistance.
• Cooperation and exchange of information and experience is
required.
Pool resources distribution to help on a regional scale
Same fre specialist as for maintained crown fre generations.
5.2.7 Conclusions
Creating a set of procedures for fire extinction within fire management means
incorporating knowledge into the system, and to transform from a hierarchical
organization to an organization adapted to learning. This is a complex, time-
consuming and expensive task. European policies should rather strengthen
investments in:
• Training fre analysis, fre behaviour and fre ecology for all staff fghting fres,
as in fast-spreading fres, tactical decisions are taken by lower ranks in the
structure, and must be based on knowledge.
• Exchanging information and experts in promoting learning organizations that
currently are clearly insuffcient. Experience must be easily transferrable and
any operation performed should form part of the knowledge-base.
• Appointing fre analysts to help detecting and prioritizing windows of
opportunity.
• Creating fre specialist groups to perform suppression fre operations. This
means not only long training and exchange programmes for individuals, but
also for those that can form part of teams which can extrapolate the results for
a wider audience.
Large fres are an increasing problem, which must be addressed at a landscape scale:
• Fire as an operational technique can be regarded as a frst step, but fre
management ideas must also be developed further, from being a technique to
manage fuel to become a technique to manage the landscape.
• Finally, promoting a wise use of fre means initiating, promoting and
expanding public awareness policies, which stop “selling” all fres as
emergencies to the public, and rather attempt to explain fres as an integral part
of landscape dynamics.
200 Towards Integrated Fire Management
References
Alexander, M.E. and Thomas, D.A. 2003. Wildland fre behavior case studies and analyses:
value, approaches, and practical uses. Fire Management Today 63: 4–9.
Alexander, M.E. 2002. The staff-ride approach to wildland fre behaviour and fre fghter
safety awareness training: a commentary. Fire Management Today 62: 25–30.
Cassagne, N., Mantzavelas, A., Dupuy, J.L., Lazaridou, T., Pimont, F., Partozis, T.,
Fernandes, P., Loureiro, C. and Rigolot, E. 2009. Report on the effects of broad scale and
long-term application of prescribed burning on fre risk components. Deliverable D5.3-1
of the Integrated Project “Fire Paradox”, Project no. FP6-018505. European Commission.
88 p.
Castellnou, M., Iglesias, N., Guarque, J., Perna, J., and Lleonart, S. 2007. Experiencias
de formación continuada en Cataluña durante las campañas 2003, 2004, 2005 y 2006.
Proceedings of the IV Internacional Wildfre Conference. Sevilla. Spain.
Castellnou, M., Mira, N., Castellarnau, X., Miralles, M., Loureiro, C., Lorca, M., Salgueiro,
A., Binggeli, F., Gibson, S., Lambert, B., Grau, G., Solivellas, P., Cendra, J. and Quílez,
R. 2008. Guideline for exchanges between fre managers. Deliverable D10.4-1 of the
Integrated Project “Fire Paradox”, Project no. FP6-018505. European Commission. 58 p.
Finney, M., McHugh, C.W. and Grenfell, I.C. 2005. Stand- and landscape-level effects of
prescribed burning on two Arizona wildfres. Canadian Journal of Forest Research 35:
1714–1722.
Garvin, D.A. 2000. Learning in action. Harvard Business School Press. Boston, MA, USA.
Lázaro, A., Solana, J., Montiel, C., Goldammer, J.G., Kraus, D. and Rigolot, E. 2008.
Collection, classifcation and mapping of the current prescribed fre and suppression
fre practices in Europe. Deliverable D7.1-3-1. of the Integrated Project “Fire Paradox”,
Project no. FP6-018505. European Commission. 47 p.
Nasiatka, P. 2003. Southern California Firestorm 2003. Report for the Wildland Fire Lessons
Learned Center. 70 p.
Nasiatka, P. 2008. Building the foundation for a learning culture. Fire Management Today.
68: 5–7.
Piñol, J., Castellnou, M. and Beven, K.J. 2007. Conditioning uncertainty in ecological
models: Assessing the impact of fre management strategies. Ecological Modelling 207:
34–44.
Rifà, A. and Castellnou, M. 2007. El modelo de extinción de incendios forestales catalan.
Proceedings of the IV Internacional Wildfre Conference. Sevilla, Spain.
Rigolot, E., Fernandes, P. and Rego, F. 2009. Managing Wildfre Risk, Prevention,
Suppression. In: Birot, Y. (ed.). Living with wildfres, what science can tell us. EFI
Discussion Paper 15. European Forest Institute. Pp. 49–52
Sagarzazu Ansaldo, M.S. and Defossé, G.E. 2009. Study of a 2
nd
set of largest past fres.
Deliverable D8.3-1 of the Integrated Project “Fire Paradox”, Project no. FP6-018505.
European Commission. 82 p.
San-Miguel, J. and Camia, A. 2009. Forest Fires at a Glance: Facts, Figures and Trends in
the EU. In: Birot, Y. (ed.). Living with Wildfres: what science can tell us. 15. European
Forest Institute. Pp. 11–18.
Saveland, J. 2008. Making sense of organizing for high reliability and learning. Fire
Management Today. 68: 8.
Seijo, F. 2009. Who framed the forest fre? State Framing and peasant counter-framing
of anthropogenic forest fres in Spain since 1940. Journal of Environmental Policy and
Planning. 11: 103–128.
State of Victoria. 2003. Report of the Inquire into the 2002–2003 Victorian Bushfres.
Australia.
Suppression Fire Use in Learning Organizations 201
Van Wagtendonk, J.W. 2007. The history and evolution of wildland fre use. History 3:3.
Van Wilgen, B.W. 2002. Changing paradigms in conservation: Lessons from six decades of
fre management in a large African national park.
5.3 Improving Suppression Fire Capacity
Marta Miralles
1
, Daniel Kraus
2
, Domingo Molina
3
, Carlos Loureiro
4
,
Giuseppe Delogu
5
, Nadine Ribet
6
, Oriol Vilalta
1
1
GRAF, Department of Internal Affairs, Spain
2
Fire Ecology Research Group, MPI Chemistry, Freiburg, Germany
3
Unit of Forest Fires, University of Lleida, Spain
4
University of Trás-os-Montes e Alto Douro, Vila Real, Portugal
5
Corpo Forestale e di Vigilanza Ambientale, Cagliari, Italy
6
Forest Technology Centre of Catalonia, Barcelona, Spain
5.3.1 Introduction
The use of fre during fre suppression operations can become a complex task, since
it adds another risk factor to the regular suite of risks in fre fghting: if not carefully
executed it may ‘fuel’ the main fre instead of containing it. However, it defnitely has
the potential to create new opportunities in front of large wildfres that are beyond the
control of any available fre fghting system when using a direct attack approach. In
order to reduce risk and improve effciency, fre use is a management tool to be placed
in few skilful hands only, with extensive and on-going training and broad experience.
Fire has been used as a technique to control wildfres for a long time. However,
given the dramatic changes of European landscapes during the last half of the 20
th

century, fre behaviour and associated fre risk has been worsening, and using fre
to change the behaviour of a wildfre in our 21
st
century landscapes, does not only
mean to light a fre. It means predicting fre behaviour, using fre as a management
tool and conducting combined fre fghting operations. The capacity of applying
these three main pillars of fire use qualifications, is the key to an efficient fire
management, and should therefore be the main focus of training.
5.3.2 Types of Suppression Fire
There are several ways in which fre is used to fght fre – collectively these are
termed suppression fre.
• Burn out or burning out is defned as removing fuel between a constructed
freline and the edge of a fre. Often this is done at the rear or on fanks of the
fre continuously with freline construction
• Backfre is similar to a burn out except it is ignited to take advantage of
the convective indraft ahead of an oncoming fre. Backfring is relatively
uncommon because it is typically performed directly ahead of the advancing
head of an intense fre. Timing is critical to allow for the reversal of wind
direction associated with the blocking of ambient wind by the main front and
indrafts to the headfre.
204 Towards Integrated Fire Management
• Back burn is defned as an ignition intended to spread in a backing direction
into the wind or down slope. Back burning may be done in prescribed fre or
on wildfres.
5.3.3 Combined fre fghting operations
When executing backfring operations during a large wildfre, the key to success
is the appropriate combination with other fre fghting techniques, since fre use
has by far a broader range of applications when combined with other techniques.
Anchoring an operation with fire involves significant changes in hose line or
defence line construction, and even requires distinct aerial support. Consequently,
safety protocols must be established accordingly, and organizational procedures
must be integrated in these operations. All organizational changes and adaptations
to a suppression fire operation must be controlled by the fire chief or Incident
Commander and advised by specialized fre analyst technicians.
5.3.4 Predicting fre behaviour: logical analysis language
The frst Fire Analyst in Europe was contracted in 1999 in Catalonia, and was
completely linked to the use of fre as a technique to fght fres. Fire was introduced
as a tool to reduce the spread of the head of large and fast fres moving through the
crowns of the trees, with massive long distance spotting. This allowed more time
for hose line construction and hand tools to contain the fanks. However, to safely
achieve a decrease in fre spread and to apply effciently a confnement strategy, an
appropriate anticipation of fre behaviour and a deep knowledge of fre operations
are needed (Castellnou and Miralles 2009).
Consequently fre analysis became a useful tool not only in slowing down head
fre spread and applying confnement strategies, but also to increase effciency
and safety in backfring and burn out operations. More recently, other regions and
countries were introducing fre as a technique and hiring fre analysts (in Gran
Canaria and in Portugal), or training ignition specialists in fre analysis (in Sardinia).
Slowly, even in wildfre scenarios in which suppression fre is not used, fre analysis
has been introduced to direct suppression efforts in the most effcient and safe way
(Castilla La Mancha, Galicia, Valencia, Aragon) (see Molina et al. in this book).
In order to change wildfire behaviour with a suppression fire, it is not only
important to understand what the fire will do, but also how operations might
change the behaviour. Every minor decision involving suppression fres must be
constantly adapted to minor changes in topography (slope, aspects, canyons), fuel
characteristics (fammability, fuel load, fuel distribution), weather forecasts (winds,
temperature, humidity) and changes in fre spread pattern related to fuel availability,
fre front and smoke plume dynamics.
Some quantitative and qualitative methods exist to anticipate fire behaviour
(Andrews 1986; Finney 1996; Clarke et al. 1994; Piñol et al. 2005). Quantitative
Improving Suppression Fire Capacity 205
methods are mainly used for political decision-making, training, prevention and
pre-suppression. During suppression operations, however, they are far from fast
or accurate enough. “During a real wildfre, if you are watching the computer you
are losing what fre is doing: there is too much information to be reviewed and
processed, and you can’t lose it” (Richard Rothermel, pers. com.).
Fighting fres does not only mean following protocols and applying training; it
goes beyond behaving like a “drill-ground commander”(Douglas Campbell, pers.
com.). Every fre situation is unique, and therefore no training or protocol will be
an exact ft to every fre situation. When confronted with a real fre under various
circumstances, a fre fghter must intelligently adapt to the situation. Experience
helps the fre fghter to recognize when the situation is unique. Consequently a rather
large base of experience is required since any new size of fre, in a new fuel type, in a
different landscape and with a different meteorology is a new set of information to be
stored and to be used under similar conditions (Alexander and Thomas 2003). This is
a lot of information to be in one’s mind, and a way to organize this mental database
of every fre fghter is fre analysis language (Campbell 1995; Castellnou 2000).
This knowledge is not only empirical, since in addition to experience it is also
based on the acquisition of scientifc knowledge and its integration in the analysis of
situations. Fire analysis is a semi-empirical qualitative tool that integrates scientifc
knowledge and experience. Its main purpose is to anticipate fre behaviour changes.
A fre analysis language system (Campbell 1995) becomes a key tool during
suppression fre use because it helps to:
• learn to see and observe;
• communicate information and tactics; and
• increase safety.
Figure 1. A deep understanding of plume-fre dynamics is needed to perform fre operations
in a plume dominated fre; the fre is not a continuous front but many different plots joining
in a large convection air mass. In plume dominated fres, without a deep understanding of
fre behaviour and plume dynamics, suppression fre operations can have the same effect as
additional spot fres, and exacerbate, instead of reduce, the effects of the wildfre.
206 Towards Integrated Fire Management
Learn to see and observe. Observing a fre from the air allows a unique view of the
fre environment. Fires as a whole, however, do not need to be watched from the
air to be understood and to take strategic or even small tactical decisions. Everyone
making decisions in a fre situation needs to learn to extract all possible information
of fre behaviour from every scene watched and a lot of information must be quickly
registered: not only relevant fame characteristics for parts of the fre affected by
Box 1. Fire anthropology
The technical use of fre mainly consists in knowing the behaviour of the fre, in order
to understand its logic and to manipulate its behaviour to serve one’s objectives.
Formalizing the fire control skills of traditional and institutional practitioners is a
challenge, because the technical nature of these skills is invisible, since they basically
rest on a sensory culture carried through the memory and the body.
The art of hazards management: geography and wildfre knowledge
In similar conditions and in familiar places, the behaviour of the fre shows certain
patterns. Repetition of wildfres provides traditional and institutional practitioners the
opportunity to gain experience of signals related to different fre behaviour in a certain
area. The hazards, the knowledge of geography and of the previous wildfres is part of
the technical skills, as far as the relative stability of geography and fre behaviour in
these particular places makes up for meteorology instability and uncertainty.
The traditional techniques can be described as a ‘dry’ culture of fire (meaning
techniques that do not use water) with few people using very simple ignition devices.
So, ignition patterns traditionally used are less complex than the ones carried out by
institutional practitioners, and are adapted to a more limited wildfre behaviour range and
to a limited geographical area.
Sharp-eye: having seen, seeing, foreseeing, reacting
The ability to control a fre does not rely on sleight-of-hand but on sharp-eye. Vision is
at the heart of technical knowledge, yet vision is not only the use of the visual faculty, it
also requires a specifc intellectual ability. It means observing at the time and establishing
comparisons between previous situations and the current situation (Weick1993; Hutchins
1995).. The behaviour of an ignition refects the consequences of the practitioners’
actions, so they need to observe many times these consequences to infer the choices to
be made. While the important things are invisible to the eyes of the neophyte, seeing
and having seen are essential skills for the practitioner to be able to foresee, and thus to
ponder and act in an appropriate way.
In a moving reality such as a wildfire, “we need to recognize places, ways and
opportune moments; and that knowledge becomes the key to an efficient action”
(Trédé 1992). Knowing how to catch the opportune moment to act is about being able
to anticipate the right circumstances, recognize them, immediately update previous
deliberations and use them.
All these skills cannot be trained, must be experienced. These experiences can be
acquired through decades following an expert, as done by traditional practitioners, or
through some years of prescribed burning some training and some years of following
teams of experts as done by institutional practitioners.
Improving Suppression Fire Capacity 207
different factors, but especially where these factors will change. Based on observed
fre behaviour and foreseen changes of factors, forecasts of where and how fre
behaviour will change are done. The decisions taken are only as good as the quality
of data available: the quality of the information extracted from the observed fre
behaviour, past observations of fre behaviour in similar situations, one’s ability
to identify key factors, as well as the quality of weather forecasts and maps. Large
fres offer very few opportunities for containment and to forecast opportunities in
advance, a key will be the ability to analyse which factors affecting fre behaviour
will change in the fre front. Consequently, anyone taking (small or large) decisions
at the fre front, should know how to observe and extract information and not only
to describe changing factors (relevant factors affecting fre behaviour and their
alignment, relevant aspects of an smoke column, of fuel availability, identify the
key factor that is speeding up a fre, the fre spread). Additionally, fre fghters have
to relate changing factors with opportunities. This means relating real fre behaviour
affected by a current set of factors, and predicting fre behaviour under forecasted
combination of the same factors, as well as relating fuel availability forecasts to
changes in extinction capacity of a fre and relating changes in main factors driving
the fre to changes in existing opportunities. A decision should also be based on
the quality of data available, especially in weather forecasting, mapping, and fre
behaviour associated with fuels.
Communicate information and tactics. Intuition cannot be explained, but
an adequate fire analysis allows explanation of and discussion about how an
opportunity of control at the fre perimeter is related to forecast fre behaviour
changes, and therefore, how to relate tactics to effciency. This has an added beneft
in prevention: the use of a common fre analysis language helps to explain and
discuss best fre prevention plans to fght the expected fre behaviour type and
prepare the predicted opportunities to fght fre or to use fre.
Increase safety. Safety is a primary consideration in any wildfre operation, and
that is particularly important when managing fire as a technique. Fire analysis
provides key information not only about real time fre behaviour, but especially
about forecast fre behaviour.
However, the behaviour of large wildfires is a complex phenomenon. The
relations between fre environment factors change in a complex, non-linear manner
as the spatial and temporal scales change, due to complex interactions at different
scales of fre environmental variables (Holling 1981). Fire behaviour has too many
components to account for the behaviour and interactions of all the parts, but too
few components to permit the assumption of uniform behaviour of the whole
system (McCormick et al. 2000). So, subtle shifts in environmental variables cause
qualitative changes in fre behaviour, and the system’s behaviour changes its scale
(McCormick et al. 2000).
5.3.5 Capacity of analysis
The study of historical fires provides a vocabulary that assists in identifying
phenomena individuals may not have seen before. Fire analysis language adapts to
208 Towards Integrated Fire Management
fre behaviour complexity through describing multiple factors that can be relevant
at different scales, and through describing the most frequent system behaviour at
each scale (examples in Campbell 1995; Castellnou et al. 2009). However, when
predicting fre behaviour, extrapolating fre analysis factors up-scale or down-scale
results in losing predictive power. In a wildfre with complex fre behaviour, sharing
a language is not enough, just as existing quantitative fre models are not enough.
What is needed is the capacity to identify the critical factors that explain most of
observed fre behaviour, to predict changes and identify indicators to re-evaluate fre
behaviour. Only real fre observation can help predict when a component or process
may enter into feedback and comes to dominate system behaviour (Alexander
and Thomas 2003). Real experience provides tools to decide which factors are
selected, when based on a personal mental database of experience. This is what
North American professionals call ‘slide-tray’. Only a mental database of thoughtful
experience provides analysis capacity. Wildfre is a clear case of complex behaviour.
Experience allows a better approximation to compare, scale and understand the role
of the different factors considered when analysing a fre. When just counting on
fre suppression experience without a sound learning and training framework, the
learning-period for the individual will become longer. Instead, only an unrealistic
insight based on similar experiences to face past-experience situations is available.
Adding a language-based logic, the fire fighter is forced to organize their own
mental database. This serves as a tool then to share knowledge, benefting from the
experience of others as well as their own experience, and we need fre professionals
able to cope with new types of fre spread, sometimes in different geographical
regions than their own home environment.
For good capacity of analysis it is important to have some personal skills such as
being able to read a map easily, to move in a landscape without getting lost, to store
and maintain a large mental database with the signifcant information on previous
wildfres observed, to identify key factors in such a complex landscape where a
large fre is burning and identify models or patterns between situations. Thus, spatial
reasoning, systematic thinking or conceptual reasoning might enable fre analysts to
build the most suited model for the fre they eventually face.
Balance between language of analysis, technical knowledge, experience and
know-how provides capacity of analysis, that is vital in increasing effciency and
safety in any operation with fre.
5.3.6 Managing the ignition: experience as a base
The ability to manage a single drip torch or a team of fire ignition specialists,
defnes the ability to adjust the ignition pattern in order to control fre behaviour.
When a suppression fre operation must be performed, many decisions related to
fre ignition are to be taken: number, size and distribution of the ignitions at any
moment. Experience is required to get trained in the management of drip torches or
other ignition devices.
The following objectives should be promoted to have trained personnel ready to
perform technical fre operations:
Improving Suppression Fire Capacity 209
Encourage training on European specifics. Training in the specifics of the
European situation should be made available through joint training courses attended
by future trainers, and through available e-learning processes. This includes training
a group of trainers from different countries and services in logical fre analysis
to improve knowledge transfer between professionals, and to increase the safety
and effciency of fre operations. Training fre managers for the fastest and largest
fres is of major importance. Some past large fres can already serve as a model of
what can be expected in future large fres, and some lessons must be learned from
the past and shared (Alexander and Thomas 2003). Studying past fres from other
regions is strengthening this kind of analysis (Sagarzazu Ansaldo and Defossé 2009)
and when a similar situation is experienced by other people, whose training was
based on these fres, the probability of a successful operation increases signifcantly.
Also joint training during fre suppression demonstrations has allowed techniques
and experiences to be exchanged between different European Fire Specialists
(Castellnou et al. 2008).
Work towards a common qualifcation system. Europe is facing enormous fres,
simultaneously-occurring large fires showing fast, extreme fire behaviour in
wildland urban interfaces as we have seen in, for example, Greece in 2007 and in
Portugal in 2003. This is a reality that we know for sure will happen again in not too
distant future. No country can successfully face such fres alone since no resources
of a single fre service will be enough for such a ‘large fre’ event. Furthermore,
large fre situations are experienced only once in a while in each country. These
events must be faced from a European perspective. However, nowadays in Europe,
every region and country uses different tools, different training techniques and
different qualifcations for its fre professionals. A frst step towards standardizing
training involves a common qualifcation system for Europe.
Accordingly, the Fire Paradox project has developed core competences for fre
specialists, levels 2 to 5 of the European Qualifcation Framework.
Level 1: Introduction – Assist
Level 2: Operator – Support
Level 3: Basic – Do/Apply
Level 4: Team boss – Manage
Level 5: Commander – Decide
This common qualifcation system serves to:
• standardize qualifcations without interfering with national/local decision
making about training;
• make exchange and collaboration between countries easier through shared
competences;
• develop performance based training with shared competences as a critical
component.
Thus a list of abilities, skills, competences and qualifcations for every job role,
should be directly related to the management of fre (see table 1). This common
qualifcation system is:

• a guidance for new organizations;
• a guidance for every organization in Europe, following a recommendation
210 Towards Integrated Fire Management
Table 1. Specifc Job Positions proposed directly related with technical fre:
Level Title
2 Technical Fire Support
3 Prescribed Burning Ignition Specialist In rangelands, heathlands
4 Prescribed Burn Boss and understory burning
3 Firing Specialist
4 Firing Operations Leader
5 Firing Tactics Commander

Table 2. Scheme of proposal of competences for Level 3 and Level 4 wildland fre specialists.
Level 3. Firing Specialist (FS) Level 4. Firing Operations Leader (FOL)
• L2 – Ensure that your actions in The competences of a Firing Specialist plus:
vegetation management reduce the risk • L4 – Prioritize opportunities associated to
to yourself and others changes of fre behaviour to propose
• L3 – Apply techniques and tactics to tactics and strategies to prevent and fght
suppress wildfre, in relation to fre wildfres in Mediterranean Ecosystems
behaviour • L4 – Control of operations with technical
• L3 – Improve safety in fre fre in a prescribed burning in rangelands
management operations and heathlands
• L3 – Help fre contain operations with • L4 – Control of operations with technical
suppression fre use (following a fre in a prescribed burning in understory
specifed fre plan). burning
• L3 – Carry out ignitions in a prescribed • L4 – Control of operations with technical
burning in rangelands and heathlands fre, to manage wildfres in Mediterranean
• L3 – Carry out ignitions in a prescribed Ecosystems
burning in understory burning • L4 – Coordinate look-out operations, in a
• L3 – Carry out ignitions to manage team performing technical fre
wildfres in Mediterranean Ecosystems • L4 – Coordinate data recording and
• L3 – Identify opportunities associated monitoring in a prescribed burning
to changes of fre behaviour to propose
tactics to prevent and fght wildfres in
Mediterranean Ecosystems

Optional competences (Minimum 4)
L2 – Use water with light weight portable pumps to suppress and or control wildfres
L2 – Use of heavy machinery to suppress wildfres
L2 – Use of aerial support to suppress wildfres
L2 – Use water from engine pumps and hoses to suppress and/or control wildfres
L2 – Use of hand tools to suppress and or control wildfres
Recommended experience Recommended experience added to FS
Prescribed Burning or Wildfre Hours 100 h Prescribed Burning or Wildfre Hours 100 h
Prescribed Burning (PB) 30 h Prescribed Burning (PB) 30h
PB in Rangelands 10 h PB in Rangelands 20h
PB in Understory 10 h PB in Understory 20h
Wildfre Hours as Fire Specialist 50 h PB boss 50h
TOTAL FS 200 h Wildfre Hours as Fire Specialist 180 h
TOTAL FOL 400 h
TOTAL FS + FOL 600 h

Improving Suppression Fire Capacity 211
to develop these qualifcations and competences in the frame of National
qualifcation systems;
• a basis for exchanges and help between countries (to be included when writing
mutual assistance agreements).
However, there are many different kinds of fre behaviour experienced all over
Europe. The skills needed to perform fre operations may vary depending on the
complexity of the operation and the fre regime characteristics. Thus, some added
competences are needed to be competent in high intensity, plume dominated fres
with independent crowning, characteristic of Mediterranean summers.
Additionally, it is clear that performing as a Firing Specialist (FS) in a specifc
ecosystem, it is required to acquire some extra know-how specifc to other fre
regimes when performing under different conditions.
Each qualifcation has a series of related competences (see Table 2)
Create opportunities for maintaining and sharing expertise. The main problem
with the use of fire for suppression is to maintain capacity of managing fire.
Consequently, if we cannot keep operational teams active, they cannot keep adding
experience hours, which holds a rather expensive problem, since not in all fres,
burn out operations can be performed. This is especially important in countries
with low-frequency, but fast and intense (large) fres threatening forest areas. Many
fre professionals are only exposed once or twice in their professional career to
the highest intensity – fastest and largest – fre experiences. People based in state
fre services (both structural and forest service based) cannot accumulate enough
experience to be prepared for such big events. It will always be a new situation,
which needs a different response (more anticipation and working in advance) than
the usual direct attack approach.
Consequently, emphasis is put on creating opportunities for accumulating and
transferring knowledge and experience, allowing the sharing of expertise and
training capabilities. However, providing experience in all different types of fres,
fre intensities and fre behaviour is sometimes diffcult if we only look at countries
or regions individually. Of course, working and experience hours under different
situations are required to have the ability to perform burn outs afterwards. In this
context it is sometimes diffcult to achieve the high expectations, especially if we
don’t look at Europe as a global wildfre area. The exchange of practitioners in large
fre events in Europe allows more people to have the opportunity of participating
in backfiring actions. Thus, the promotion of exchanges, both of people and
information from all over Europe should be targeted and a series of best exercises
and best practices are proposed to enrich exchanges.
Best exercises during exchanges. Exchanges with the purpose of sharing exercises
have proven to be very useful because they are a great opportunity to share different
approaches to a common problem. Best exercises shared during exchanges have
been suppression fre demonstrations, analysis and priorities of opportunities in
small scenarios and tracking opportunities in a landscape.
Best practices during exchanges. Best practices are the ones that allow learning
from the tools and operations that the host organization is operating and performing,
e.g. participating in a prescribed burning or constructing a line with the host tools,
including the ones that allow learning from the type of fres, fuels, opportunities and
capacity of extinction in the host organization (e.g. analysis of past fres).
212 Towards Integrated Fire Management
Box 2. Lessons learned through exchanges
Fire fghters only experience a large wildfre episode in their area once to three times in
their career. Only exchanging people during periods of risk in other areas can guarantee
having experts ready to face a large fre episode, such as those occurred in Portugal
(2003, 2005) or Greece (2007). Exchanging people and information have also benefts
for the professionals who are hosting their colleagues, because both groups can share
new approaches to the same old problems.
Figure 2. Zone where a burnout
was started in l’Aude (France) in
August 2007 in Armissan. The
opportunity to control this flank
could be used only through a fre
suppression operation, and the
presence of a Catalan fre specialist
(in a Fire Paradox exchange)
provided the French fire fighters
the needed expertise to perform
successfully this operation, and to
acquire needed experience in their
own ground.
Figure 3. Catalan fire specialist
team working in a north Atlantic
European ecosystem in Scotland,
to get more experience in a light
fuels fire, to be prepared for a
wildfire in their region in the
Pyrenees.
Figure 4. Training the use of
the torch in prescribed burning:
learning to look back (Catalan
example). Regulating the distances
between ignitions and distances
of discontinuous fire fronts, with
the same fuel means different
behaviour.
Improving Suppression Fire Capacity 213
5.3.7 Prescribed burning as a basic tool to train in the use of
suppression fre
The most effcient way to increase opportunities of learning a technique to manage
fire is participating in the performance of prescribed burning. Performing fire
operations during a wildfre in Mediterranean areas requires many skills that are also
necessary to perform prescribed burning both in rangelands and heathlands (without
limitation of avoiding crown scorch but with larger fre fronts to control), and also
the ones needed to perform understory burning (with more control of ignition
patterns in a detailed scale due to crown scorch limitations).
Improving individual capability of analysis and drip torch management is a long
process, but allows:
• to increase the capacity of lower levels to take informed tactical decisions;
• to improve the quality of strategic decisions, increasing the effciency in the
use of resources when there is a real opportunity;
• to improve the leadership towards other organizations participating in fre
extinction;
• to plan in advance fre suppression operations.
5.3.8 Conclusions and recommendations
By defnition, fre suppression techniques are used in a changing and unstable context.
As a consequence, in any wildfre scenario, and especially in a scenario involving fre
management, fre services must use all techniques and abilities available to read the
clues given by fre behaviour, and to anticipate moments, places and ignition patterns,
critical for the success of the operation. That means that building up an experienced
fre service and maintaining knowledge and experience of fres in all its different
conditions, is as important as a good training. This implies investing in well trained
fre specialists, providing opportunities for experienced crews and fre analysts to
experience large fre events, and investing in tools to increase and share knowledge
and experience. During the large fres in Spain, Portugal, France and Greece, these
investments in fre-experienced fre services had not been in place. In fact, policies
of the European Union were repeatedly impeding factors to this development, since
a request for massive aerial means for cooperation and assistance is still the common
procedure, rather than asking for well trained crews and experienced fre analysts. To
achieve the required changes in Europe, the needs are:
• Training, training and training about each fre. We need to ensure that everybody
knows a good range of past fres and to create thereby a ‘common experience’;
• Use analysis to determine where to attack and when, so we use our resources at
their maximum capacity;
• The same knowledge can be used in prevention during a no fre situation;
• Using fre as a suppression technique means getting experience through
prescribed burning;
• Send fre professionals working outside their own fre service area and comfort
zone, so they can pick up more experience to be used back home.
214 Towards Integrated Fire Management
Box 3. Prescribed burning to improve fre specialist abilities in Portugal.
It is important that the fire professionals become more involved in forest fuel
management, outside the fire season, mainly during prescribed burning operations.
Prescribed burning is the most suitable scenario to provide training opportunities outside
a wildfre. Some large prescribed fres can be used as training situations for the use of
fre and to test best ignition patterns in suppression fre. This has been done in Portugal,
where fre analysts and fre specialists must have prescribed fre training, certifcation
and practice before starting to participate in wildfre analyses and suppression fre use.
The prescribed burning program included shrublands (1 318 ha in the last two years)
but also forested stands (173 ha). Such a complete prescribed burning program is a
needed step to train technicians and burning teams.
Some of these technicians are part of the fre analysis and fre use team (GAUF –
Grupo de Análise e Uso do Fogo) that provide support to wildfre fghting during the fre
season.
The activity of this group during the prescribed fre season provided the possibility for
different actors to be involved in fre fghting, creating experience and knowledge that is
useful during the fre season.
This strategy was also used to establish international exchange. The Lousã Summer
Bases in 2007 and 2008, and the Prescribed Burning Practitioners Meeting, in February
2009, are examples of this exchange.
The participation of this highly experienced fre analysts and fre specialists was also a
beneft for the prescribed burning program, namely through:
• Support to local burning teams needing the support of more experienced elements.
• Support to local burning teams, in their training, allowing to extend and to increase
the effciency of burning operations, using more wide prescription windows, and
burning in more diverse situations,
• Application of this management technique in regions where there are not qualifed
technicians or hand crew teams with prescribed burning training.
These actions allowed a better expansion of prescribed burning, a larger implementation
and more efficient fire management actions. This were also good opportunities to
develop experience and exchange between teams, and also initial attack and suppression
fre training, in situations close to the wildfre environment.
References

Alexander, M.E., and Thomas, D.A. 2003. Wildland fre behavior case studies and analyses:
value, approaches, and practical uses. Fire Management Today 63: 4–9.
Andrews, P.L. 1986. Behave: fre behavior prediction and fuel modeling system - BURN
subsystem Part 1. USDA Forest Service, Intermountain Research Station, Ogden, UT,
USA. 130 p.
Campbell, D. (ed.) 1995. The Campbell Prediction System: A Wild Land Fire Prediction
System & Language. 129 p.
Castellnou, M. 2000. Nuevas Metodologías de Prevención de Grandes Incendios. Congreso
Forestal Ibérico. Castelo Branco, Portugal.
Improving Suppression Fire Capacity 215
Castellnou, M., Mira, N., Castellarnau, X., Miralles, M., Loureiro, C., Lorca, M., Salgueiro,
A., Binggeli, F., Gibson, S., Lambert, B., Grau, G., Solivellas, P., Cendra, J. and Quílez,
R. 2008. Guidelines for exchanges between fre managers. Deliverable D10.4-1 of the
Integrated Project “Fire Paradox”, Project no. FP6-018505. European Commission. 58 p.
Castellnou, M and Miralles, M. 2009. The great fre changes in the Mediterranean - the
example of Catalonia, Spain. Crisis Response 5: 56–57.
Castellnou, M., Miralles, M., Pages, J. and Pique, M. 2009 Tipifcación de los incendios
forestales de Cataluña. Elaboración del mapa de incendios de diseño como herramienta
para la gestión forestal. Proceedings of the 5th Congreso Forestal Español. Ávila, Spain.
Clarke, K.C. Brass, J.A. and Riggan, P.J. 1994. A cellular automation model of wildfre
propagation and extinction. Photogrammetic Engineering & Remote Sensing 60: 1355–
1367.
Finney, M. 1996. FARSITE: Users Guide and Technical Documentation Systems for
Environmental Management. Missoula, MT, USA.
Holling, C.S. 1981. Forest insects, forest fres and resilience. USDA Forest Service General
Technical Report WO 26: 445–464.
Hutchins, E. 1995. How a cockpit remembers its speed. Cognitive Science 19: 265–288.
McCormick, R.J. Brandner, T.A. and Allen, T.F.H. 2000. Toward a theory of meso-scale
wildfre modeling - a complex system approach using artifcial neural networks. In: LF
Neuenschwander, K.C., Ryan, G.E., Gollberg and Greer J.D., (eds.). Proceedings from the
Joint Fire Science Conference and Workshop. University of Idaho and the International
Association of Wildland Fire. Moscow, ID, USA. Pp. 3–15.
Piñol J., Beven K and Viegas D.X. 2005. Modelling the effect of fre-exclusion and
prescribed fre on wildfre size in Mediterranean ecosystems. Ecological Modelling 183:
397–409
Sagarzazu Ansaldo, M.S. and Defossé, G.E. 2009. Study of a 2
nd
set of largest past fres.
Deliverable D8.3-1 of the Integrated Project “Fire Paradox”, Project no. FP6-018505.
European Commission. 82 p.
Trédé, M. 1992. Kairos : L’à-propos et l’occasion. Editions Klincksieck, Paris. 338 p.
Weick, K.E. 1993. The collapse of sense making in organizations: the Mann Gulch Disaster.
Administrative Science Quarterly 38: 628–652.
6. Solving the Fire Paradox – Regulating the
Wildfre Problem by the Wise Use of Fire
6. Solving the Fire Paradox – Regulating the
Wildfre Problem by the Wise Use of Fire
Francisco C. Rego
1
, Joaquim S. Silva
1
, Paulo Fernandes
2
, Eric Rigolot
3
1
Centre of Applied Ecology “Prof. Baeta Neves”, Technical University of Lisbon,
Portugal
2
University of Trás-os-Montes e Alto Douro, Vila-Real, Portugal
3
National Institute for Agricultural Research, Avignon, France
6.1 Introduction
“Fire is a bad master but a good servant” is a Finnish proverb which illustrates the
paradox of fre, as it refects the detrimental effects when it runs out of control and
the benefcial effects when it is mastered suiting human interests. However fre
should be understood beyond an anthropocentric perspective, as it is associated with
many natural ecosystems in the world, therefore playing a role in their conservation.
Despite this general understanding shared by many ecologists, we are far from
knowing what is ‘good for nature’. On the contrary we have a quite clear perspective
of what is ‘good’ for human societies. However, where fre is concerned there are
contradictory perspectives of master and servant, which lead to the paradox of fre.
Understanding the detrimental and beneficial effects of fire is fundamental
for a more comprehensive view of the Fire Paradox and constitutes the optimal
conceptual basis for integrated fre management policies and practices. Considering
this integrated perspective, the set of coherent actions required to minimize the
net social costs of fres include: the mitigation of fuel hazard, the reduction of
unwanted ignitions and the suppression of wildfres. In turn, this set of actions must
be translated into policies and disseminated to all participating agents. The structure
of the present chapter follows this framework, including different developments and
solutions to solve the paradox of fre.
6.2 Minimizing the net social costs of fres
The objective of minimizing the social impacts of forest fres implies in the frst
place, the need of quantifying the social costs and benefts of fre regimes. This is
a problem which is common to many aspects of forests and wildland areas, given
the diffculties to determine the value of externalities, for example. Despite the
methodologies which were developed to tackle this problem (see Mavsar et al. in
this volume) there are major drawbacks which have to be solved or minimized,
such as the absence of common criteria for data collection on costs and expenses
associated with wildfres.
220 Towards Integrated Fire Management
Minimizing the social costs of fire implies the assessment of fire effects on
vegetation, people and infrastructures. With this purpose, research efforts were
developed in order to determine fire effects on trees using experimental and
numerical studies of fre behaviour and effects (Guijarro et al. (2009), including
a review on the state of the art of pine resistance to fre (Fernandes et al. 2008).
Another contribution to the assessment of fre effects was performed by Hostikka
et al. (2008) and Torero (2008) including the simulation of fre brands (i.e. burning
particles such as twigs or leaves carried by wind and capable of starting fres)
travelling into houses (Sikanen 2009). Wildland urban interfaces are both an
important source of ignitions and an important concern due to the values at risk in
terms of human lives and property. The strategy to tackle this important problem
includes the characterization and the mapping of wildland urban interfaces and the
assessment of vulnerability and potential damages (Lampin et al. 2008).
Another important aspect regarding the social costs of fres has to do with the
traditional use of fire in some European regions. Traditional fire use practices
constitute an infuencing factor which has to determine and guide the strategies
and recommendations to be applied within a given social context. Recognizing the
benefts of traditional fre by integrating these practices instead of systematically
suppressing them, may contribute to fuel management and may save the resources
spent in unnecessary fire fighting (Lázaro et al. 2008). Nevertheless European
countries present contrasting situations in terms of traditional fre use: in Central
Europe and Baltic countries there has been a general abandonment of rural fre
practices but in other countries, particularly in the Mediterranean Basin, fre is a
deeply rooted tool for rural activities.
Minimization of wildfre costs is also infuenced by the characteristics of the
vegetation which burns. Different works have approached the problem of fire
selectivity (Bajocco and Ricotta 2008; Moreira et al. 2009) allowing the conclusion
that land use may definitively influence fire incidence and therefore the social
costs from wildfres. Another related issue is the capacity of vegetation to continue
assuring the benefts and services to the society and its role in the ecosystems, after
the fre has passed. Some vegetation types are more resilient than others, which
means they are capable of contributing to reduce the negative effects of wildfres
by quickly recovering after burning. This may contribute to strongly reduce the cost
of wildfres from both economic and ecological perspectives (see Mavsar et al. in
this volume and Krivtsov et al. 2008). Therefore, integrated fre management should
consider not only the use of fre to modify fuels and to fght fres, but also to drive
the ecological succession towards less fre prone and more resilient vegetation, in
order to minimize the costs of wildfres.
6.3 Managing fuels with prescribed burning to reduce wildfre hazard
and severity
Proper fuel management requires sound knowledge of fuel characteristics for the
different fuel types. Accurate characterization and mapping of fuels can provide
useful information to be used in pre-fre planning, and therefore in the construction
Solving the Fire Paradox – Regulating the Wildfre Problem by the Wise Use of Fire 221
of forest fre management plans. Remotely sensed data have already been used to
directly identify and map fuel types or to obtain vegetation maps, which were later
translated into fuel models. The objective of fuel type classifcation is to defne
vegetation associations with similar potential fre behaviour. With this purpose
several recent studies were produced aiming at developing fuel typology and fuel
characterization systems, using remote sensing (Mårell et al. 2008). Additionally,
fre hazard maps are needed to assist the spatial planning of prescribed burning. A
sequence of mapping exercises was conducted in study areas in South Africa and
Portugal for integrated fre management on a regional level (De Ronde 2009a).
These can be used with advantage in other European countries to develop optimized
fre protection systems, and to select where to apply prescribed burning and fuel
treatments in general.
The study of fuels has different aims, one of the most important being the use
of fuel characteristics as inputs to fire behaviour models and systems. Recent
works have allowed the development of a fuel editor aiming to be, on one hand, a
management tool for manipulating fuel complexes and on the other, an application
that enables fre simulations and the generation of vegetation post fre succession
steps (Lecomte et al. 2009). Since fuel conditions are constantly changing, the
dynamic side of fuel characterization is undoubtedly a challenging domain, given
the diffculties of forecasting scenarios of fuel development, and the usefulness of
having this information for fuel and fre management planning, including the use of
prescribed fre. Therefore some attempts have been performed in order to model the
temporal variation of fuel accumulation and fuel structure, with the general aim of
evaluating prescribed fre regimes (Krivtsov et al. 2008; Krivtsov et al. 2009a,b).
The use of prescribed burning has to be thoroughly assessed in order to allow
evaluation of its effectiveness in reducing fre hazard. In the absence of the required
long-term information, fre modelling and analysis of the fre regime dependency on
fuel age, can be used as a surrogate to ascertain the relevance of prescribed burning
(Cassagne et al. 2009).
A positive side effect of prescribed burning is the possibility to mitigate CO
2

emissions by reducing the potential fre hazard of wildlands. However, prescribed
burning in Europe can only make a signifcant contribution in those countries where
fre incidence is high (Narayan et al. 2007). Over a 5-year period, the emissions
from wildfres in Europe were estimated to be approximately 11 million tonnes of
CO
2
per year, while with extended prescribed burning programs this was estimated
to be 6 million tonnes, a potential reduction of almost 50%. This means that for
countries in the Mediterranean region it may be worthwhile to account for the
reduction in emissions obtained when such techniques are applied.
As a fnal consideration regarding the importance of prescribed burning as a key
issue to solve the Fire Paradox, we should add the need of international cooperation.
This has been achieved through the knowledge exchange between different
European countries but also with non-European countries. A good example was
the establishment of guidelines for the integration of prescribed burning in Europe,
using the experience from South Africa (De Ronde 2009b). Although very different
from Europe, the South African situation can be extrapolated to some European
countries, especially those which share similar tree species used in industrial
plantations such as Pinus spp. and Eucalyptus globulus.
222 Towards Integrated Fire Management
6.4 Decreasing unwanted ignitions and integrating the traditional use
of fre
The study of ignitions starts with the basic characterization of forest fuel pyrolysis,
and the process of ignition and combustion. From the fuel perspective, the processes
controlling ignition and fame spread are the degradation chemistry and the different
heat and mass transport processes. Given the complexity of these processes new
developments have recently been introduced by comparing different methods for
studying and modelling the pyrolysis process of forest fuels (Rein et al. 2008).
Fuel moisture content (FMC) is a critical parameter that affects fre ignition, fre
behaviour and fre severity. However, FMC modelling is still a challenge. In fact,
FMC is a necessary input for all existing physical, semi-physical and empirical
fre behaviour prediction models as well as ignition probability models. Besides
that, FMC, or its surrogate in the form of an index, is a key factor in most fre
danger rating systems. From the operational point of view, FMC estimates are also
required for prescribed burning planning and for suppression activities. Recent
studies have provided new data for the prediction of fuel moisture using remotely
sensed data (Todone et al. 2009), methods for estimating ignition probability of
fne fuels (Kazakis et al. 2007), and relationships between ignitability and fuel bed
characteristics (Curt et al. 2007).
Besides the physical and biological mechanisms related to ignitions, the basic
problem to solve, in order to decrease the number of unwanted ignitions, deals
with the relationships between fre starts and land use, particularly in the European
context. According to different studies it is known that ignition probability is closely
related with human presence and human activities (see Catry et al. in this volume).
Agricultural practices are known to be an important source of ignitions in different
European regions. Also road networks and wildland urban interfaces have a positive
relationship with ignition probability. Therefore the identifcation of these situations
is of crucial importance for decreasing the number of unwanted ignitions (Bajocco
et al. 2008).
Consequently the development of public awareness strategies is a fundamental
aspect if we want to mitigate the problem of unwanted ignitions and to have
a conscious use of fre. The achievement of these objectives depends on a proper
use of recent information and communication techniques adapted to the different
specifc situations (Badillo and Bourgeois 2008). In this feld we are far from having
satisfactory solutions in terms of outcomes obtained from the communication process.
Therefore it is important to analyze the communication process, in order to develop a
more effective communication and dissemination strategy (Badillo et al. 2009).
6.5 Reducing the extent of wildfres through suppression fre
The further objective is to decrease wildfre sizes, costs and damages. One main
approach is the use of suppression fre. This is a very challenging practice requiring
in-depth knowledge of fre behaviour, including the interaction between fre fronts,
which has been poorly addressed by past research. Numerical simulations on
suppression fres can be performed with existing 3D physical-based models such
Solving the Fire Paradox – Regulating the Wildfre Problem by the Wise Use of Fire 223
as Firetech and WFDS (Linn et al. 2002). Numerical simulations of a fully-physical
model are an alternative to obtain information on fre propagation. These models
provide detailed information such as the temperature of the gaseous phase, the
temperature of the different components of the vegetation, fuel moisture content
of the vegetation, heat fuxes, as a function of spatial position and time. Given that
these quantities cannot be determined through experiments, numerical simulations
are a powerful tool to study the mechanisms of fre behaviour and effects. Recent
developments have contributed to, and improved, the scarce body of numerical
studies of wildfre behaviour using full-physical models, addressing the interaction
between fre fronts and fre impact on trees.
An additional source of knowledge is the recording of large past fires. It is
worth noting that the implications of wildfre case studies done in one country can
be extended to others. The reconstruction and analysis of fres can be based on
historical fre databases, reports, journals, satellite images and interviews, including
the description of time and cause of ignition, meteorological and topographic
conditions, main vegetation types present in the region, fre suppression strategies
and tactics applied, fre behaviour characteristics (rate of spread, length of fame,
spotting, etc.), and diffculties encountered by fre suppression. Detailed information
about past wildfres are valuable not only in assisting training but also as sources of
research data (Sargarzazu and Defossé 2009).
Temporary feld bases located in areas historically subject to wildfre are also an
interesting source of knowledge and experience. These multidisciplinary summer
laboratories were a meeting point for fre research in Europe with the objective of
accelerating the collection and sharing of knowledge for subsequent dissemination.
The overall objectives of feld bases were the follow-up and monitoring of wildfres
and the analysis and interpretation of aerial documentation. Aerial observations
recorded by airplanes above wildfres allowed for a continuous follow-up of fre
propagation. The use of infrared and laser technologies provided data allowing
deeper understanding of the mechanisms involved. This information together
with ground-based observations was shared between scientists, students and fre
managers from different countries (Binggeli 2009).
The availability of information on fre propagation issued with the support of
simulation software is another crucial aspect of fre fghting. Image analysis software
is capable of extracting fre lines and spotting from infrared images, and a ‘regional
centre of resources’ is able to put them in a geographical system and to send this
information. Moreover, the simulation of fre spread running on a remote dedicated
server can supply extra information about fre propagation and this information
can be used to generate a prevision fre map in real time. The transmission of this
data can be used by fre commanders (on a laptop or PDA) in real time, to run fre
simulations and fre spotting assessment (Orabona et al. 2007).
6.6 Policies and knowledge dissemination
The general concept of integrated fre management is particularly relevant in the
dissemination of knowledge and recommendations to policy makers, fre managers
and university students, given the integrated perspective which has necessarily to
224 Towards Integrated Fire Management
be introduced in these different approaches to the paradox of fre. These approaches
may come in the form of scientific outputs, technical guidelines, and policy
documents to support the implementation policies and practices, and the academic
and professional training.
It is important to assess the main benefits and limitations of the legislation
and policy outputs in order to consider their contributions to integrated fire
management, before making proposals for new legislation. Although the European
legislation has contributed to homogenize national legal frameworks, there are still
important differences among countries. On the other hand most forestry-related
documents hardly ever mention wildfre management. Also, there is clearly a lack
of an effective coordination among the different units dealing with fres. Further,
problems of coordination are exacerbated by the fact that many countries have
federal systems or are undergoing decentralization trends. Regarding community-
based cooperation, organized groups of local stakeholders are emerging especially
in Mediterranean countries. These groups contribute to fre management as a result
of instrumental motivation, or self-interest. Furthermore, some common ecological
and socio-economic patterns have been recognized at the regional level, which can
be used to provide recommendations for the future and to set the basis for new
legislation and policy measures relative to integrated fre management, adapted to
each territorial context (Agudo and Montiel 2009; Herrero et al. 2009). Apart from
sector policies which have the responsibility for national fre management (forest
and civil protection), it is considered necessary to analyze other public policies,
which are outside of the sector scope. The assessment performed by Lázaro et al.
(2009) has identifed some common fndings about the role that territorial policies
should play concerning integrated fre management at the national and European
level.
Integrated fre management also deals with cooperation policies. Cooperation
mechanisms, from the regional to the international scale, emerge as an evident
answer to the problem of wildfres and fre use. In most countries the coordinating
institutions in fire prevention are the agricultural, forest and environmental
government services, with only a few countries assigning that responsibility to civil
protection related institutions. The majority of cooperating entities in prevention
are public institutions, but in some cases private entities, such as volunteers, private
companies, and forest owners associations also participate in prevention programs.
In most countries the coordination of fre extinction activities is the responsibility of
the civil protection, but in some countries the coordination is the responsibility of
forest or environmental services. International cooperation is more often associated
with mutual assistance protocols for fre extinction, such as bilateral agreements, and
European Union’s Community Civil Protection Mechanism. Bilateral agreements
are usually related to aerial means, but some countries also have agreements for
the intervention in common border areas. European Union’s Community Civil
Protection Mechanism is the most important mechanism at the European scale,
conveying wildfre disaster relief requests to the community countries (Correia et
al. 2008).
With regard to knowledge dissemination, we must distinguish the professional
training of managers and fre fghters and the academic training level. Both should
include an integrated fre management approach in order to know how to analyze and
Solving the Fire Paradox – Regulating the Wildfre Problem by the Wise Use of Fire 225
understand fre in its dual perspective: i.e. as both master and servant. Prescribed fre
and experimental fres must be considered as an obligatory part of any training for fre
professionals. The extensive practice of prescribed burning is a very effective training
for fre professionals who have an important role in fuel management, simultaneously
with their training for fre fghting. Formal knowledge dissemination is far from being
suffcient to prepare fre professionals. Training is a fundamental aspect in assuring
effectiveness in fre fghting efforts. The training of fre professionals strongly benefts
from the exchange of knowledge between crews from different countries during the
occurrence of large wildfres (Castellnou et al. 2008).
Concerning university and post-graduate level training, not many initiatives have
been particularly focused on integrated fre management. In Europe, fre academic
training is not responding to the increasing number of issues in forest and nature
management that are formulated, implemented, and co-ordinated at a level above
that of the nation state. This situation provides a whole range of new challenges and
demands for policy and management at the European level. This situation calls for
the implementation of post-graduate studies on fre management. The educational
goals of such a post-graduate program should provide the students with an effective,
quality learning experience. Science-based education must be balanced with
application of that science, and providing advise that is tailored to help students to
meet their long-term career and educational goals. Further, such graduate programs
should seek to attract and support students who have the potential to become leaders
within the fre community, and to foster a culture of diversity, community and global
citizenship. An initiative with these characteristics has recently been launched under
the form of an International Fire Master Program based on an agreement between
four universities from Europe and the USA, as an outcome of the Fire Paradox
project.
6.7 Final considerations
As a closing remark, it is important to retain a message which is crucial for solving
the paradox of fre: there is no single solution to this complex problem. Solving
the paradox of fire should make use of a whole range of attributes, from fuel
management to public awareness; from professional training to science informing
policy. The Fire Paradox project has produced a variety of fre science goods and
services including: scientifc papers, technical reports, technological tools, training
programs and training materials. This diversity of outputs is necessary if we want
to introduce changes in fre management, rather than just stating principles. The
outcomes of the Fire Paradox project provide an integrated solution to a complex
problem in fre management. In the present day the complex trade-offs between
fre as a bad master and as a good servant, have necessarily to be played by agents
that have the knowledge and the training required under the principles of integrated
fre management and wise use of fre. The preparation of these agents, from policy
makers to fre managers, under this integrated approach is the core of our proposal
to solve the Fire Paradox.
226 Towards Integrated Fire Management
References
Agudo, J. and Montiel, C. 2009. Basis to start the process for a proposal of a new legislation
at EU level. Deliverable D.7.1-1.3 of the Integrated project “Fire Paradox”. European
Commission. 45 p.
Badillo, P.-Y. and Bourgeois, D. 2008. Report to the European Commission: synthesis of the
state of the art. Deliverable D11.1-5 of the Integrated Project “Fire Paradox” , Project no.
FP6-018505. European Commission. 51 p.
Badillo, P.-Y., Bourgeois, D., Marciano, J-P. and Asdourian, B. 2009. Seminar with experts.
Internal Report 11.3-5 of the Integrated Project “Fire Paradox” , Project no. FP6-018505.
European Commission. 18 p.
Bajocco, S. and Ricotta, C. 2008. Evidence of selective burning in Sardinia (Italy): which
land-cover classes do wildfres prefer? Landscape Ecology 23(2): 557–567.
Bajocco, S., De Angelis, A., Rosati, L., Ricotta, C., Moufis, G.D., Gitas, I.Z., Silva, J.S.,
Moreira, F., Vaz, P., Catry, F., Ferreira, P.G., Pezzatti, G.B., Torriani, D. and Conedera,
M. 2008. Publication describing wildfre regimes distribution in selected European study
areas. Deliverable D4.1-1b of the Integrated Project “Fire Paradox”, Project no. FP6-
018505. European Commission. 79 p.
Binggeli, F. 2009. Bulletins of the bases. Internal Report IR8.1-1 of the Integrated Project
“Fire Paradox”, Project no. FP6-018505. European Commission. 7 p.
Cassagne, N., Mantzavelas, A., Dupuy, J.L., Lazaridou, T., Pimont, F., Partozis, T.,
Fernandes, P., Loureiro, C. and
Rigolot, E. 2009. Report on the effects of broad scale and long-term application of
prescribed burning on fre risk components. Deliverable D5.3-1 of the Integrated project
“Fire Paradox”, FP6-018505. European Commission. 88 p.
Castellnou, M., Mira, N., Castellarnau, X., Miralles, M., Loureiro, C., Lorca, M., Salgueiro,
A., Binggeli, F., Gibson, S., Lambert, B., Grau, G., Solivellas, P., Cendra, J. and Quílez,
R. 2008. Guidelines for exchanges between fre managers. Deliverable D10.4-1 of the
Integrated Project “Fire Paradox”, Project no. FP6-018505. European Commission. 58 p.
Correia, A., Colaço, C., Rego, F., Solana, J., Herrero, G., Montiel, C., Sesbou, A. and Kraus,
D. 2008. A review of the existing cooperation mechanisms between departments dealing
with fre management at the European and national level. Deliverable 7.1-4 – 1 of the
Integrated Project “Fire Paradox”, Project no. FP6-018505. European Commission. 58 p.
Curt, T., Ganteaume, A., Jappiot, M., Lampin, C., Borgniet, L., Baudel, J., Esteve, R.,
Martin, W., N’Diaye Boubacar, A., Madrigal, J., Díez Galilea, C., Marino, E., Guijarro,
M. and Hernando, C. 2007. Empirical modelling of fuel fammability and initial fre
propagation: First results. Deliverable D3.1-6 of the Integrated Project “Fire Paradox”,
Project no. FP6-018505. European Commission. 23 p.
De Ronde, C. 2009a. Recommendations for prescribed burning mapping in Mediterranean-
type natural vegetation and forests. Deliverable D5.4-2 of the Integrated Project “Fire
Paradox”, Project no. FP6-018505. European Commission. 17 p.
De Ronde, C. 2009b. Guidelines for the integration of prescribed burning in the Fire
Paradox environment: within natural vegetation, industrial forest land and the urban
interface. Deliverable D13.1-2 of the Integrated Project “Fire Paradox”, Project no. FP6-
018505. European Commission. 26 p.
Fernandes, P.M., Vega, J.A., Jiménez, E. and Rigolot, E. 2008. Fire resistance of European
pines. Forest Ecology and Management 256: 246-255.
Guijarro, M., Madrigal, J., Marino, E., Hernando, C., Díez, C., Mendes-Lopes, J.M.,
Ventura, J., Santos, N., Meléndez, J., López, F., Aranda, J.M., Díaz, F., Defossé, G.,
Godoy, M.M., Sagarzazy, M.S., Bianchi, L., Bravo, S., Kunst, C., Ledesma, R. and
Godoy, J. 2009. Fire impact on trees and shrubs: methods and frst results. Internal Report
Solving the Fire Paradox – Regulating the Wildfre Problem by the Wise Use of Fire 227
IR2.3-3 of the Integrated Project “Fire Paradox”, Project no. FP6-018505. European
Commission. 45 p.
Herrero, G., Montiel, C., Agudo, J. and Aguilar, S. 2009. Assessment document on the main
strengths and weaknesses of the legislation and policy instruments concerning integrated
wildland fre management in the EU, in European Member States and in North African
countries. Deliverable D7.1-1-2 of the Integrated Project “Fire Paradox”, Project no. FP6-
018505. European Commission. 117 p.
Hostikka, S., Mangs, J. and Mikkola, E. 2008. The effect of three-dimensionality on the
building response to a forest fre. Deliverable D2.5-2 of the Integrated Project “Fire
Paradox”, Project no. FP6-018505. European Commission. 53 p.
Kazakis, G., Ghosn, D., Xanthopoulos, G. and Dimitrakopoulos, A. 2007. Probability of
ignition modelling in forest fuels – First results. Deliverable D3.1-5 of the Integrated
Project “Fire Paradox”, Project no. FP6-018505. European Commission. 42 p.
Krivtsov, V., Fernandes, P., Higham, B., Pezzatti, G.B. and Legg, C. 2009a. Release of
prototype software for allometric modelling of fuel structure with provisional case studies.
Deliverable D3.3-5 of the Integrated Project “Fire Paradox”, Project no. FP6-018505.
European Commission. 33 p.
Krivtsov, V., Legg, C., Vigy, O., Rigolot, E., Lecomte, I., Curt, T., Jappiot, M., Lampin, C.,
Fernandes, P. and Pezzatti, B. 2008. Prototype of biophysical models of fuel accumulation
using empirical data. Deliverable D3.3-3 of the Integrated Project “Fire Paradox”, Project
no. FP6-018505. European Commission. 30 p.
Krivtsov, V., Vigy, O., Legg, C., Curt, T., Rigolot, E., Lecomte, I., Jappiot, M., Lampin-
Maillet, C., Fernandes, P. and Pezzatti, G. 2009b. Fuel modelling in terrestrial ecosystems:
An overview in the context of the development of an object-orientated database for wild
fre analysis. Ecological Modelling 220:2915–2926.
Lampin, C., Long, M., Jappiot, M., Morge, D., Bouillon, C., Mantzavelas, A.,
Apostolopoulou, I., Lazaridou, T., Partozis, T., Topaloudis, T., Galiana, L., Herrero, G.
and Solana, J.L. 2008. Guide for end-users to characterize and to map wildland urban
interfaces. Deliverable D5.2-2 of the Integrated Project “Fire Paradox”, Project no. FP6-
018505. European Commission. 18 p.
Lázaro, A., Galiana, L., Aguilar, S., and Sesbou, A. 2009. Assessment document on the main
strengths and weaknesses of the legislation and spatial policy instruments concerning
wildfre integrated management in the EU, in European member states and in North
African countries. Deliverable D7.3-1 of the Integrated Project “Fire Paradox”, Project
no. FP6-018505. European Commission. 81 p.
Lázaro, A., Solana, J., Montiel, C., Goldammer, J.G., Kraus, D. and Rigolot, E. 2008.
Collection, classifcation and mapping of the current prescribed fre and suppression
fre practices in Europe. Deliverable D7.1-3-1 of the Integrated Project “Fire Paradox”,
Project no. FP6-018505. European Commission. 47 p.
Lecomte, I., de Coligny, F., Griffon, S., Pimont, F., Rigaud, E., Rigolot, E. and Vigy, O.
2009. Fire Paradox fuel manager: software and user’s manual. Product P6.1-6 of the
Integrated Project “Fire Paradox”, Project no. FP6-018505. European Commission. 85 p.
Linn, R., Reisner, J., Colman, J.J. and Winterkamp, J. 2002. Studying wildfre behavior
using FIRETEC. International Journal of Wildland Fire 11:233-246.
Mårell, A., Abdelmoula, K., Allgöwer, B., Borgniet, L., Cassagne, N., Curt, T., Ganteaume,
A., Ghosn, D., Gitas, I., Jappiot, M., Jimenez, E., Kazakis, G., Koetz, B., Lampin, C.,
Machrouh, A., Morsdorf, F., Pimont, F., Rigolot, E., Sesbou, A., Vega, J.-A. and Wellani,
W. 2008. Fuel description: frst results. Deliverable D3.4-3 of the Integrated Project “Fire
Paradox”, Project no. FP6-018505. European Commission. 80 p.
Moreira, F., Vaz, P., Catry, F. and Silva, J.S. 2009. Regional variations in wildfre
susceptibility of land-cover types in Portugal: implications for landscape management to
minimize fre hazard. International Journal of Wildland Fire 18:563-574.
228 Towards Integrated Fire Management
Narayan, C., Fernandes, P.M., van Brusselen, J. and Schuck, A. 2007. Potential for CO
2

emissions mitigation in Europe through prescribed burning in the context of the Kyoto
protocol. Forest Ecology and Management 251:164–173.
Orabona, V., Mazzoleni, S. and Giannino, F. 2007. Report on the type of data collection.
Deliverable D6.4-1 of the Integrated Project “Fire Paradox”, Project no. FP6-018505.
European Commission. 12 p.
Rein, G., Torero, J., Guijarro, M., Esko, M. and Castro, A. 2008. Methods for the
experimental study and recommendations for the modelling of pyrolysis and combustion
of forest fuels. Deliverable D2.1-1 of the Integrated Project “Fire Paradox”, Project no.
FP6-018505. European Commission. 83 p.
Sagarzazu, M.S. and Defossé, G.E. 2009. Study of a 2º set of largest past fres. Deliverable
D8.3-1 of the Integrated Project “Fire Paradox”, Project no. FP6-018505. European
Commission. 82 p.
Sikanen, T. 2009. Fire safety analysis around targets using FDS: Final achievements -
Transport of frebrands and attack on buildings. Deliverable D2.5-4 of the Integrated
Project “Fire Paradox”, Project no. FP6-018505. European Commission. 24 p.
Todone, J.F., Bianchi, L.O. and Defossé, G.E. 2009. First feld sampling results of dead
and live fuel moisture prediction in Patagonia. Internal Report IR3.1-7b of the Integrated
Project “Fire Paradox”, Project no. FP6-018505. European Commission. 24 p.
Torero, J.L. 2008. Criteria for building material ignition and assessment of burn injuries in
wildland fres: model development. Internal Report IR2.5-3 of the Integrated Project “Fire
Paradox”, Project no. FP6-018505. European Commission. 58 p.
Important Concepts and Terms
Integrated fre management: A concept for planning and operational systems that
include social, economical, cultural and ecological evaluations with the objective of
minimizing the damage and maximizing the benefts of fre. These systems include
a combination of prevention and suppression strategies and techniques that integrate
the use of technical fres and regulate traditional burning.
Fire management: All activities required for the protection of burnable forest and
other vegetation values from fre, and the use of fre to meet land management goals
and objectives.
Traditional burning (or traditional fre use): The use of fre by rural communities
for land and resource management purposes based on traditional know-how.
Appropriate traditional fre use: The use of traditional burning under legal regulations
and good practices.
Technical fre: The controlled use of fre carried out by qualifed personnel under
specific environmental conditions and based on an analysis of fire behaviour.
Technical fres are divided into prescribed fres, wildfres within prescription and
suppression fres.
Fire within prescription: A prescribed fre or a wildfre that burns within prescription.
Prescribed burning (or prescribed fre): The application of a fre under specifed
environmental conditions, which allow the fre to be confned to a predetermined
area and to attain planned resource management objectives.
Wildfre within prescription: A wildfre that is confned to a predetermined area and
produces the fre behaviour and the fre effects required to attain the planned fre
treatment and/or resource management objectives.
Wildfre: Any unplanned and uncontrolled vegetation fre which, regardless of the
ignition source, may require suppression response or other actions according to
agency policy.
Suppression fre: The application of a fre to accelerate or strengthen the suppression
of wildfres.
23
European Forest Institute Research Report 23
The approach taken in the Fire Paradox project was based on the paradox
that fire can be “a bad master but a good servant”, thus requiring the
consideration of the negative impacts of current wildfire regimes
(understanding fire initiation and propagation) and the beneficial impacts
of managed fires in vegetation management and as a planned mitigation
practice (prescribed burning together with some traditional fire uses)
and for combating wildfires (suppression fire). These were the four
integration pillars of the project.
This Research Report reflects the structure of the project, corresponding
to its integration pillars – initiation, propagation, prescribed burning and
suppression fires – and including a closing chapter which synthesizes
and combines the main project outcomes. The book provides science
based knowledge that can assist policy makers to develop the necessary
‘common strategies’ to elaborate and implement integrated fire
management policies. It makes extensive use of the science and technology
findings from the Fire Paradox project, focusing on policies and best
management practices, as well as providing guidelines for the future.
The Fire Paradox project (2006–2010) was funded by the European
Commission Research and Development 6
th
Framework Programme.
The project included 30 partners from eleven European countries and
six partners from Africa, South America and Asia, with close support
from an International Advisory Committee formed by nine specialists
from the USA, Canada and Australia. Fire problems and solutions are
found all over the world, and we see the knowledge exchange and benefits
of Fire Paradox will extend far beyond Europe.
Joaquim Sande Silva
Francisco Rego
Paulo Fernandes
Eric Rigolot
(editors)
The European Forest Institute – EFI – is an international organisation established
by European States. EFI strengthens and mobilises European forest research
and expertise to address policy-relevant needs.
www.efi.int
ISSN: 1238-8785
ISBN: 970-952-5453-48-5
Towards Integrated Fire Management –
Outcomes of the European Project
Fire Paradox
S
i
l
v
a

e
t

a
l
.



T
o
w
a
r
d
s

I
n
t
e
g
r
a
t
e
d

F
i
r
e

M
a
n
a
g
e
m
e
n
t

–

O
u
t
c
o
m
e
s

o
f

t
h
e

E
u
r
o
p
e
a
n

P
r
o
j
e
c
t

F
i
r
e

P
a
r
a
d
o
x

Towards Integrated Fire Management – Outcomes of the European Project Fire Paradox

Joaquim Sande Silva, Francisco Rego, Paulo Fernandes and Eric Rigolot (editors)

European Forest Institute Research Report 23, 2010

Scientific Advisory Board of the European Forest Institute 2009–2010 Prof. Dr. Bas Arts, the Netherlands Prof. Kristina Blennow, Sweden (as of 1 January 2010) Prof. Americo Carvalho Mendes, Portugal Dr. Emil Cienciala, Czech Republic Prof. Dr. Carlos Gracia, Spain (as of 1 January 2010) Prof. Dr. Hubert Hasenauer, Austria, Chairman of the SAB Dr. Antoine Kremer, France (until 31 December 2009) Prof. Michael Köhl, Germany Prof. Marco Marchetti, Italy Prof. Leena Paavilainen, Finland Prof. Jean-Luc Peyron, France Dr. Victor Teplyakov, Russian Federation (until 31 December 2009)

Editorial Office Prof. Dr Hubert Hasenauer, Editor-in-Chief Ms Minna Korhonen, Managing Editor Contact: publications@efi.int European Forest Institute Torikatu 34, 80100 Joensuu, Finland www.efi.int Cover photo: Pedro Palheiro Layout: Kopijyvä Oy Printed at WS Bookwell Oy, Porvoo, Finland Disclaimer: This report was prepared for the project Fire Paradox under EU Sixth Framework Programme (contract no. FP6-018505). The content of this publication is the sole responsibility of the authors and does not necessarily reflect the views of the European Union or the European Forest Institute. © European Forest Institute 2010 ISSN: 1238-8785 ISBN: 970-952-5453-48-5

To José Moreira da Silva. . Pau Costa. and all those who dedicated their lives to fire management.

.

3 Learning and Training on the Use of Prescribed Burning Techniques .............................................4 The Importance of Economics in Fire Management Programmes Analysis ............................3 An Overview of Policies and Practices Related to Fire Ignitions at the European Union Level ................. 161 5................ 3 2.. 177 5...................................... 61 3................. 229 .. 189 5........................................................................................................ Climate and Land Cover as Controls on European Fire Regimes .................................................5 Improving Fire Management Success through Fire Behaviour Specialists ........................... Wildfire Propagation: Knowledge and Search for Solutions 3................................ 121 4............3 Improving Suppression Fire Capacity ................................... Mapping and Assessment .......... 71 3.6 Wildfire Scenarios: Learning from Experience ........................................................1 Overview of Prescribed Burning Policies and Practices in Europe and Other Countries......3 Wildland Urban Interfaces..1 Humans....... 105 3................ 137 4...........................1 Fire Starts and Human Activities ...........2 Fire Modelling and Simulation Tools .....................Table of Contents Foreword .........................................2 Flammability: Influence of Fuel on Fire Initiation ................................................................. 219 Important Concepts and Terms ......................1 Overview of Suppression Fire Policies and Practices in Europe ............................................................................................... ix 1............. Suppression Fire: a Tool to Control Wildfires 5...2 Scientific Knowledge and Operational Tools to Support Prescribed Burning: Recent Developments .................. Prescribed Burning: Preventing Wildfires and More 4......... Solving the Fire Paradox – Regulating the Wildfire Problem by the Wise Use of Fire ...................... 23 2............................................................................................................................ 203 6.............................................................. Fire Behaviour and Vulnerability: Characterization................ 35 3..... 151 4....vii Acknowledgements ............... 93 3....... Introducing the Fire Paradox ......................................................................................................... 49 3....2 Suppression Fire Use in Learning Organizations ............................ 9 2............ Wildfire Initiation: Understanding and Managing Ignitions 2..........................................................................................................

.

I am convinced that it will be a useful contribution to the implementation of fire management policies and measures in Europe.Foreword One of the most important challenges for European forests. Another cornerstone addressing the importance of the fire issues in the EFI’s activities was the EFI Discussion Paper 15 “Living with wildfires: what science can tell us”. European Forest Institute . shrublands and crops every year. I would also like to thank the European Commission for supporting this publication through Fire Paradox. This was the justification for the establishment in 2005 of the EFI Project Centre PHOENIX concentrating on fire ecology and post-fire management. Risto Päivinen Director. with the aim of gathering the main results achieved by the project in one publication. In terms of knowledge dissemination. This knowledge is fundamental to support adequate political decisions and to implement the most appropriate technical solutions in order to mitigate the problem. provided the perspective that societies should learn to live with wildfires by controlling the circumstances causing forest fires. The European Forest Institute is committed to support initiatives which help generation and dissemination of scientific knowledge on forest fires. Besides the necessary acknowledgements to all experts involved. Yves Birot and his co-authors from the Regional Office EFIMED. is fire. This publication by Dr. especially in Southern Europe. This general concept of Fire Management was also the leading principle of the EC-funded project “Fire Paradox”. 2006–2010. in which EFI was part of the consortium. Fire has been affecting on average half a million hectares of forests. The result of this effort is the present EFI Research Report. key researches in the Fire Paradox project have been compiled into a report.

.

Dr. since this book is the direct result of the work developed during the four years duration of the project. Editor-in-Chief of EFI's Research Report series. January 2010 . In particular we have to thank three persons who gave precious assistance to the publication of this Research Report: Minna Korhonen. We would also like to thank all of the experts who participated in the review process: Martin Alexander Margarita Arianoutsou Mark Beighley Yves Birot David Christenson Miguel Cruz Neels De Ronde Mark Finney Armando Gonzalez-Caban Jim Gould Chad Hoffman Justin Leonard Vittorio Leone Rodman Linn Peter Moore José Manuel Moreno Penelope Morgan Jeffrey Prestemon Douglas Rideout Kevin Ryan Jim Smalley Scott Stephens Ramon Vallejo Gavriil Xanthopoulos The editors are also grateful to all authors who accepted the challenge to participate in this Research Report. Finally we would like to thank the whole Fire Paradox Consortium. The editors appreciate the recommendations sent by Prof.Acknowledgements The editors would like to thank all those who have contributed to the present publication. The Editors Joensuu. Hubert Hasenauer. Tim Green and Luís Queirós.

.

1. Introducing the Fire Paradox .

.

and European levels. Paulo Fernandes2 and Eric Rigolot3 (the editors) 1 Centre of Applied Ecology “Prof. Integration of the activities and the use of the capabilities of the rural populations (communities. Vila Real. Canada and Australia which formed the International Advisory Committee. vegetation (forest) protection. and we see the knowledge exchange and benefits of Fire Paradox will extend far beyond Europe. practices and policies under integrated fire management in Europe. 2. Avignon.1. national. South America and Asia. being defined by FAO/GFMC as “fire management systems”. and prescribed fire technology”. and smoke management (community-based fire management). The project included 30 partners from eleven European countries and six partners from Africa. These include one or both of the following concepts of integration: 1. Portugal 2 University of Trás-os-Montes e Alto Douro. Francisco C. The main objective of the Fire Paradox project was to produce and use science to inform decisions. Fire problems and solutions are found all over the world. Silva1. Portugal 3 National Institute for Agricultural Research. values-at-risk. and the development of strategies for their implementation at the regional. detection. individual land users) to meet the overall objectives of land management. Integration of natural or human-caused wildfires within prescription and/ or planned application of fire in forestry and other land-use systems in accordance with the objectives of prescribed burning. suppression. Introducing the Fire Paradox Joaquim S. involving “the strategic integration of such factors as knowledge of fire regimes. with close support from ten specialists from the USA. The approach taken by Fire Paradox includes the development of the science to inform the use of fire as a main component of the fire management solutions proposed. Baeta Neves”. probable fire effects. Technical University of Lisbon. Integrated Fire Management involves further integrating the different components of fire management (fire prevention. 1 1999 Wildland Fire Management Terminology produced by the Global Fire Monitoring Centre for the FAO . France The Fire Paradox project (2006–2010) was funded by the European Commission Research and Development program. cost of fire-related activities. The Wildland Fire Management Terminology produced by the Global Fire Monitoring Centre for the FAO1 defines Fire Management as including “all activities required for the protection of burnable forest and other vegetation values from fire and the use of fire to meet land management goals and objectives”. Rego1. level of forest protection required. and use) with consideration of fire ecology relationships.

The chapters were written by leading authors recruited from key researchers from Fire Paradox Modules and Work Packages. It makes extensive use of the science and technology findings from the Fire Paradox project. recognizing that “managing beneficial aspects of fires may involve various forms of fire use”. Within each chapter there are a number of text boxes.4 Towards Integrated Fire Management In other documents. as in many regions of Europe fire has been traditionally used and continues to be used in rural practices whereas. The approach taken in Fire Paradox was based on the paradox that fire can be “a bad master but a good servant”. prescribed burning and suppression fires – and including a closing chapter which synthesizes and combines the main project outcomes. The book has four different topic sections. there may also be more potential solutions to solve the paradox. In this section authors present the current knowledge and the project findings on fire . propagation. and outputs in order to illustrate key messages and recommendations. focusing on policies and best management practices. the European situation offers a very special context to the use of the Fire Paradox philosophy. In fact. This apparent paradox is more evident in Europe than in many other regions of the world. The book provides science based knowledge that can assist policy makers to develop the necessary ‘common strategies’ to elaborate and implement integrated fire management policies. at the same time. and by facilitating the implementation of “cost-effective approaches of preventing detrimental fires and maintaining desirable fire regimes”. and the traditional fire use. In order to complement the information provided in the different chapters. containing 15 chapters: Section 1. corresponding to its integration pillars – initiation. The concept of integrated fire management was never intended for Europe. When compared to other less populated parts of the world. and the philosophy of Fire Paradox offers an innovate approach for its implementation. These were the four integration pillars of the project. many of which describe specific project activities. integrated fire management is viewed as “an approach to addressing the problems and issues posed by both detrimental and beneficial fires”. and by using fire to suppress wildfires (see Figure 1). the present publication includes a CD containing the deliverables and the internal reports from Fire Paradox Modules and Work Packages. such as those produced by the Global Fire Initiative. the vast majority of damaging wildfires have human causes. Fire Paradox proposes the full integration of fire use in the prevention and suppression strategies by promoting the possibilities of the beneficial use of fire by prescribed burning. Wildfire initiation: understanding and managing ignitions. by “evaluating and balancing the relative risks posed by fire with the beneficial or necessary ecological or economical roles that it may play”. These documents are directly or indirectly referred to in the chapters and were included in order to provide the outputs of the project on which the conclusions and recommendations were essentially based. as well as providing guidelines for the future. because of this. This Research Report reflects the structure of the project. thus requiring the consideration of the negative impacts of current wildfire regimes (understanding initiation and propagation) and the beneficial impacts of managed fires in vegetation management and as a planned mitigation practice (prescribed burning together with some traditional fire uses) and for combating wildfires (suppression fire).

Section 3. Integrated Fire Management. This section presents a review on the policies and practices of suppression fire in Europe and other regions. Section 4.Introducing the Fire Paradox 5 Figure 1. Section 2. the knowledge on fire behaviour for fire fighting operations and the use of wildfire scenarios for fire suppression purposes. fuel flammability and fire occurrence. . wildland urban interfaces. and an assessment of policies and practices related to fire ignitions. the scientific knowledge and operational tools to support prescribed burning. from a fire use perspective. Suppression fire: a tool to control wildfires. This section reviews recent developments to evaluate prescribed burning policies and practices in Europe and other regions. fire modelling and simulation. Wildfire Propagation: knowledge and search for solutions. This section addresses issues related with different factors influencing European fire regimes. ignitions related to human activities. fire economics. and discusses the learning and the training on the use of prescribed burning. Prescribed burning: preventing wildfires and more.

In addition. this book presents science-based knowledge and recommendations for changing the paradigm of fire management in Europe and elsewhere. As a concluding text. the present Research Report provides a comprehensive synopsis of science. Overall.6 Towards Integrated Fire Management the application and practice of suppression fire as an operational tool in Europe and discusses the requirements of training for improving suppression fire capacity. products and services resulting from the Fire Paradox project. technology. there is a standalone chapter presenting some of the project developments which may contribute to solve the Fire Paradox in the context of integrated fire management. .

2. Wildfire Initiation: Understanding and Managing Ignitions .

.

1 Fire Starts and Human Activities Filipe X. Francisco Moreira1. Anthropogenic and natural causes There is evidence that the earliest use of fire by hominids occurred more than one million years ago (Pausas and Keeley 2009). This is especially true for Europe and particularly for the Mediterranean Region with its long history of human intervention in the ecosystems (Thirgood 1981). 1998). Technical University of Lisbon. The ability to understand and predict the patterns of fire ignitions will help managers and decision makers to improve the effectiveness of fire prevention. Then we present a more detailed analysis of key factors driving fire ignitions based on some case studies. Switzerland 2. Portugal 2 Institute for Environment and Sustainability.1. Italy 4 Ecosystem Boundaries Research Unit. . The EFFIS database only includes wildfires that affected.1. so that in most regions of the world. Swiss Federal Research Institute WSL. detection and fire fighting resources allocation. 2. Catry1. at least partially. Joint Research Centre of the European Commission.1 Introduction The knowledge about where. Carlo Ricotta3 and Marco Conedera4 1 Centre of Applied Ecology “Prof. Joaquim S.2 Overview of fire ignitions occurrence in Europe Unless otherwise specified the information presented in this introductory section results from the analysis of data extracted from the European Forest Fire Information System (EFFIS) wildfire database concerning the period 1998–2007 (21 countries – not all the countries have data for the whole period). and we conclude by making some remarks concerning implications for management. Rego1.2. forest and other wooded land areas. University of Rome “La Sapienza”. Italy 3 Department of Plant Biology. human activities have become much more important than natural sources of ignition (Goldammer and Crutzen 1993). when and why do fires start is essential to assure appropriate fire policy and management. Francisco C. although solid evidence date from less than 300 000 to 400 000 years (Weiner et al. Baeta Neves”. Natural fire regimes have been increasingly altered by humans for many thousands of years at least. Andrea Camia2. Silva1. This chapter starts by giving an overview of the patterns of fire occurrence in Europe.

there are many other factors that are also known to cause fires. Concerning the question of how do fires start. such as the burning of agricultural and forestry residues. 95% of the fires in Europe are directly or indirectly caused by human behaviour and activities. there are some common aspects which characterize the human behaviour and activities which result in fire starts. 44% started by negligence or accident. Bassi and Kettunen 2008. Main general causes of wildfires in Europe. accidents with electric power lines and railways or the fire use associated with forest recreation (e.10 Towards Integrated Fire Management Figure 1. Figure 1 shows the main general causes of wildfires for the 13 countries that had at least 500 fires that were successfully investigated.g. Currently. accident). the investigation of ignition causes is crucial if we want to understand the important role of humans in fire regimes. Independently of the general designations (intentional. with intentional causes representing 2% to 79% of all fires. However. negligent and accidental causes representing 10% to 98%. land burning for pasture renovation or the use of machinery. including arson. there are enormous differences between countries. However there is a considerable difference among the different . Among the fires with known causes. and natural causes representing 0% to 32%. However. countries are ordered by decreasing proportion of deliberately caused fires). Some of the major anthropogenic causes of wildfires in Europe are associated with land management. Vélez 2009). which can vary among countries. negligent. As we saw there are big differences among countries with regard to what are the main fire causes. Data extracted from the fire database of the European Forest Fire Information System (EFFIS) (Period 1998–2007. and only 5% had natural causes (mainly lightning). only countries with a minimum of 500 fires successfully investigated are included. 51% were intentionally caused.

in Portugal uncontrolled agricultural burnings are classified as negligence. European countries with respect to the percentage of fires for which the cause of the fire is determined. Moreover the criteria for investigating and classifying the cause of a fire are not uniform among European countries. There are several factors that can potentially explain the spatial variation of ignition density. Although the distinction between natural and human-caused fires may be more or less straightforward.g. the different countries have not used the same criteria. main economic activities and land use. Several studies worldwide have identified numerous factors influencing the spatial patterns of fire ignitions (e. based on province-level data (NUTS 3). it is much more difficult to classify the different human activities responsible for a fire ignition. For example. Fire density (number of fires/year/10 km2) in Europe: average annual distribution in the period 1998–2007. Even at the very coarse level of classification which only separates negligent/accidental from intentional (arson) causes. Spatial and temporal patterns of fire ignitions The higher densities of wildfires in Europe are mainly concentrated in the southern regions with the Mediterranean countries accounting for most of the fires reported. Loboda and Csiszar . These percentages range from nearly 100% in Latvia to only 5% in Portugal. making international comparisons difficult. whereas in Spain this cause for fire starts is classified as intentional (APAS 2004). Mercer and Prestemon 2005. precluding the possibility of reliable comparisons. such as climate. population density. However in some central and northern countries fires are also frequent (Figure 2).Fire Starts and Human Activities 11 Figure 2.

including the increasing recreational use of forests and land accessibility contribute to this trend. source: EC 2008). For example the EFFIS database shows an annual average of 89 000 fires in the last decade (1998–2007) which is lower than the average reported for the 1990s in a UN report (UN 2002). the number of fires has apparently stabilized or is even beginning to decrease. EC 2008). and until recent years scarce information was available for Europe. However. looking at the trends of the five largest EU Mediterranean countries (those having a longer time-series fire database). UN (2002) reported an increase from about 40 000 fires in the 1970s to 95 000 in the 1990s (decade averages for a group of 31 European countries). Some additional information is presented in this chapter (see sections 2.1. and even if they present very different situations.3 and 2. In the last decade. Evolution of the number of fires in five southern European countries in the period 1980–2007.12 Towards Integrated Fire Management Figure 3. we can also see that the overall tendency in the current decade is towards a stabilization or to a decrease in the number of fires (Figure 3). while the UN report refers to 31 European countries. however the EFFIS database refers to a group of 21 European countries. 2007). the effects of different factors on fire ignition occurrence can vary a lot among ecosystems and across spatial scales (e.4). the improvements on fire detection methods and in the communication/reporting systems are likely to have also contributed to the detection and reporting of more fires. 2007).1. and we do not know to what extent each factor has influenced the observed trend. thus limiting direct comparisons. Several socio-economic factors. but some more time is necessary to clearly confirm this trend. . However. Yang et al. On the other hand. The number of wildfires in Europe has been increasing in recent decades (UN 2002. despite the large annual variability. using moving averages (n = 5 years) (based on country-level data.g.

In the Mediterranean countries the majority of fires occur during the summer.00h (53% of all ignitions). corresponding to the summer months (July to September). and only 7% in autumn (October to December).Fire Starts and Human Activities 13 Figure 4. while in most of the remaining countries the majority of fires occur in spring. On average. 27% of the wildfires occur during the spring months (April to June). The majority of fires start between 13. Data extracted from the fire database of the European Forest Fire Information System (EFFIS) (Period 1998–2007.00h.00h and 07. The majority (51%) of all wildfires occurs during the third trimester of the year. .00h and 19. 21 countries). Fire occurrence in the European region is mostly concentrated in summer and spring (see Figure 4). On average the hourly distribution of wildfires peaks between 14. 14% occur during winter (January to March).00h and 17. and the period of the day when fewest fire starts is between 03. Average monthly (top) and hourly (bottom) distribution of wildfires in Europe. with 30% of all fires starting in this period (Figure 4).00h (only 3%). However the monthly distribution of wildfires is not the same for all the countries. There are few differences among European countries with regard to the distribution of fire ignitions during the day.

less than 1% of all fires are larger than 100 ha. 2001). The work was focused on characterizing and identifying the main factors driving the spatial occurrence of wildfire ignitions. The Mediterranean countries have the highest number of fires and the largest area burned in Europe. Population density was found to be the most important variable in the multivariate models. 2. the vast majority of wildfires are small. In Portugal. traditional burning to eliminate agricultural residues). 70% of all fire ignitions occurred in municipalities with more than 100 persons/km2 although they only represent 21% of the territory. meaning a higher probability of ignition in the more populated areas.g. On average 70% of all European wildfires burn less than 1 ha and 95% burn less than 10 ha. Cardille et al. 2008. The selected explanatory variables were all highly significantly related with the spatial distribution of ignitions. and only a few have major consequences. The vast . Several logistical models were developed and an ignition risk map was produced for the entire Portuguese mainland.2% are considered to be large fires (more than 500 ha). and less than 0. Recently a research team studied about 130 000 fires that occurred in Portugal during a five-year period (Catry et al.3 Analysis of key factors driving fire ignitions The Portuguese case study Portugal is the European country with the highest density of wildfires. the larger fires are almost exclusively concentrated in the Mediterranean countries. Cardille et al. like those implicating an active of fire (e.g.g. Land cover was also hypothesized to be a determinant factor causing ignitions because different kinds of human activities (land uses) lead to different levels of risk. similar to other author’s findings (e.14 Towards Integrated Fire Management Fire ignitions and resulting burned area In the European region. In several other studies worldwide population density was also found to be positively related to wildfire ignitions (e.1. Although in general these patterns are quite similar in all countries. Several socio-economic and environmental variables hypothesized to be related to the spatial distribution of fire ignitions were selected and analyzed. 2009). flammability). Modelling fire ignitions Human presence and activity are the key drivers of fire ignitions in Portugal. In our studies land cover showed a strong influence on the probability of fire ignition. demonstrating that it is possible to predict the likelihood of ignition occurrence with good accuracy and using a small number of easily obtainable variables. 2001). 2007. moisture. and because different land covers have different fuel characteristics (e. This variable was positively correlated with ignition occurrence. On average about 500 000 ha are burned every year in these countries. In total. and thus it was selected as a case study for a more detailed analysis. both in terms of ignitions and of burned area.g. which can determine fire ignition and initial spread.

According to our results. becoming negative when predicting the occurrence of fires larger than 500 ha. 98% of all ignitions occurred at less than 2 km from the nearest road and 85% were within a distance of 500 m. shrublands and sparsely vegetated areas also showed a positive influence on the multivariate models. which are known to cause frequent wildfires in the Iberian Peninsula (see Box 1). Elevation positively influenced ignition distribution. The high ignition incidence in urban-rural areas may also be explained by the co-occurrence of agricultural activities and a high human presence.g. and that the herbaceous vegetation in many Mediterranean agricultural areas is easier to ignite and the fires propagate more easily than in other fuel types. Although most fires start in areas of higher population density. The influence of land cover type is different for ignitions resulting in small or large fires. MMA 2007). Additionally. this result has also been found in other studies (e.g. 1996). Vazquez and Moreno 1998). Forests and shrublands registered three times less fire ignitions than if they were randomly distributed. This effect may be because some human activities are more likely to occur at higher elevations. but their influence was considerably lower. results also show that the probability of large fire occurrence is significantly higher in woodlands and lower in urban-rural areas. especially during the summer when fuel moisture is very low. Forests. Fire ignitions and resulting burned area The spatial patterns of ignition are not the same for all fire sizes (Catry et al. and the modelling work performed taking into account the final burned area.g. and only 15% started in forested or uncultivated areas although they cover 50% of the country. there are important differences in the origin of ignitions that resulted in smaller versus larger burned areas. Results indicate that agriculture is a very important factor influencing fire starts. Probability of fire ignition occurrence decreased with increasing distance to main roads. Considering all fire sizes. Vega-Garcia et al. the class of mixed urban-rural areas had nine times more ignitions than would be expected if they occurred randomly in the territory. 2008). However. The strong positive influence of population density observed in the global multivariate predictive models rapidly decreases for increasing fire sizes. This is in accordance with previous investigations of fire causes. such as the renovation of pastures for livestock using traditional burnings. which concluded that a large proportion of wildfires is due to agricultural activities (e. The frequency of ignitions also depends on distance to roads. However we verified that the frequency of larger fires starting at longer distances . This conclusion is based on both the comparisons between the frequency of ignitions resulting in fires of different sizes. with most larger fires starting in areas with comparatively lower population density (see Figure 5).Fire Starts and Human Activities 15 majority of fires started in agricultural and mixed urban-rural areas (85%). and ignitions are more likely to occur closer to the main roads independent of the resulting fire size. when compared to small fires. lightning-caused ignitions have been described to be more likely to occur at higher elevations (e. Also contributing to this high ignition incidence are the higher population densities usually present in these areas (in comparison to forested and uncultivated areas). Badia-Perpinyà and Pallares-Barbera (2006) also found a higher ignition frequency at higher elevations in a rural area of north-eastern Spain.

05. distance to roads and elevation. land cover. Expected fire frequencies (corresponding to a random occurrence) are shown as reference (adapted from Catry et al.001) in relation to population density. ** p<0. . 2008). Comparison between fire ignitions that resulted in burned areas smaller or larger than 500 ha (significant differences * p<0.16 Towards Integrated Fire Management Figure 5.

such that the probability of fire ignition will be affected to a large extent by the nature of the different land cover types. The models produced had reasonable to good predictive ability. Models built for different years performed similarly. if several land cover types were equally fire-prone. the actual number of fires in each land cover class was compared with the results of 999 random simulations. The null hypothesis is that fires occur randomly across the landscape so no differences between the relative abundance of fires in each land cover class and the relative extent of each class within the analyzed region are expected. Land cover and fuel characteristics are characteristics that are often closely related (Turner and Romme 1994). and (iii) 1331 fire records in the Cantons of Ticino. agricultural practices. .1. 2002). To determine whether the number of fires in the examined cover classes is significantly different from random.g. In a different study a research team quantified coarse-scale patterns of fire ignition in three selected European study areas in order to determine to what extent ignition occurrence is dependent on land cover in different European regions (see Bajocco and Ricotta 2007). with ROC (Receiver Operating Characteristics) analysis indicating 74% to 87% agreement between predicted probabilities and observed outcomes. Monte Carlo simulations were used. fires would occur randomly in the landscape. 2. as stated in the Portuguese study case. Graubuenden and Uri (Switzerland) during the period 1982–2005. Authors identified land cover types where fire ignition probability is higher or lower than expected by a random null model assuming that. understanding the role of land cover in shaping coarse-scale wildfire regimes has become a major concern. At each study site.4 European comparisons – Relationships between ignitions and land cover Due to increasing anthropogenic fire activity. (ii) 3023 fire records in the Coimbra district (Portugal) during the period 2001–2005. the composition of which slightly varies across the study areas (see Table 1). larger fires are significantly more likely to start at higher elevations than smaller fires (Figure 5). where the original classes were aggregated into 12 coarser macroclasses. This is because human presence and impact (e. population density. but also that the spatial patterns of ignitions are different for larger or smaller wildfires.Fire Starts and Human Activities 17 is significantly higher than for smaller fires. which indicates little spatial variability in a five-year period. Methods The analyzed data sets are composed of (i) 13 377 fire records in Sardinia (Italy) during the period 2000–2004. Land cover maps with a common legend were used for all selected study sites derived from CORINE Land Cover data. grazing pressure) together with the amount of fuel and its spatial distribution are basic factors in explaining fire ignition and propagation (Lloret et al. The urban class includes mixed urban-rural areas. Results of this study clearly showed that fire ignitions are strongly related with human presence and activity. Concerning elevation.

By contrast. .GR-UR) + (-) NP + + + + + (+) - Results The results of the analysis on fire ignitions in selected European study areas are shown in Table 1.01). natural grasslands and shrublands exhibit the opposite behavior. Overall. respectively. due to the presence of two different fire seasons (spring and summer). Land Cover Types Urban surfaces Non-irrigated arable land Irrigated arable land Vineyards Fruit trees and olive groves Heterogeneous agricultural areas Broad-leaved forests Coniferous forests Mixed forests Pastures Natural grasslands Transitional woodland-shrub and/or sclerophyllous vegetation Italy (Sardinia) + + (+) + + + Portugal (Coimbra) + + + + NP Switzerland (TI. From a statistical viewpoint. Results of the analysis on fire ignitions occurrence by land cover class in different European regions (Italy. all being characterized by low ignition probabilities. in Sardinia. This is also the case for two forest classes. On the other hand. The symbols + and – indicate.05. shrublands and coniferous forests. where. there is a clear positive association between the number of fires and urban and agricultural land cover types. all land cover types with high human pressure are characterized by very high fire ignition probabilities.18 Towards Integrated Fire Management Table 1. in grasslands. A similar result is obtained for Portugal. grasslands and shrublands. NP = the land cover class is not present in the study area. the number of fires is lower than expected by chance alone. this analysis is easy to conduct at any spatial scale and with any meaningful land-use classification scheme providing valuable information for the development of strategies for fire risk assessment and fire prevention. the number of fires is higher than expected by a random null model in urban areas and in all agricultural classes. the number of fires is lower than expected in forests. ‘broad-leaved forests’ and ‘mixed forests’ that are strongly affected by anthropogenic induced fires because of their proximity to settlements. More specifically. In Switzerland. According to the Monte-Carlo simulations. () = not significant at p = 0. the situation is slightly more complex. Forests. whether fire ignition occurrence probability is higher or lower than expected by a random null model on a given land cover class at p < 0. fire incidence in terms of number of fires by land cover class is not random (p < 0. though the results obtained slightly vary across the study sites. Portugal and Switzerland). Most artificial and agricultural land cover classes are characterized by high ignition probabilities. apart from ‘irrigated arable land’ and ‘vineyards’.01.

Fire Starts and Human Activities 19

2.1.5

Conclusions – Implications for management

The ignition risk (the chance of a fire starting as determined by the presence and activity of any causative agent, also defined as ‘fire risk’; FAO 1986; NWCG 2006) is an essential element in assessing fire danger (e.g. Finney 2005; Vasilakos et al. 2007). Nowadays, many fire management decisions are still exclusively based on the factors influencing the fire spread and suppression difficulty. However, as resources are limited, it is important to define priorities among areas. Under similar fuel, topographic or weather conditions, the areas with higher ignition risk should be given priority for surveillance (e.g. Vasconcelos et al. 2001; Chuvieco et al. 2003). The ability to understand and predict the patterns of fire ignitions is essential for managers. It will help improve the effectiveness of fire prevention, detection and allocation of fire fighting resources. Knowledge about where, when and why do fires start improve that ability. Fire incidence varies greatly among land cover types. Some types of land cover burn far more often than would be the case if fires occurred randomly. Further, fire ignition probability is usually connected to social, economic and cultural drivers. The disproportionately high fire occurrence in the urban and agricultural classes is largely due to the high human presence that represents the major source of fire ignition in most European regions. Population density, distance to roads and elevation were also found to be important predictors of the occurrence of fire ignitions. The present results may strongly contribute to solving the Fire Paradox by providing to the managers information on priority areas with high fire hazard and risk. In these areas, optimizing prescribed burning applications and the ecological benefits of the wise use of fire may be very effective in achieving the wildfire reduction goals. The temporal patterns of fire starts seem to be mostly related with the meteorological conditions, which affects the fire ignition and initial spread (e.g. affecting fuel moisture and flammability). The knowledge of these patterns can also have important consequences for fire management, namely in terms of vigilance and fire fighting preparedness. The knowledge on fire causes is also of crucial importance and should have a high priority in terms of investing resources for minimizing the wildfire problem. Nevertheless, there are important differences among European countries concerning the process of fire cause investigation. These differences result in different rates of investigated fires and especially in different classification systems, which preclude reliable, quantitative comparisons between countries. Given these discrepancies, the European Commission through its Joint Research Centre has recently launched a call for proposals concerning the “Determination of forest fire causes and harmonisation of methods for reporting them”. However it is necessary that results coming from this initiative are translated into policies aimed at harmonizing the different methods and criteria for fire investigation and fire cause classification. Although local actions targeted at well identified groups are an essential strategy for preventing fire ignitions, Europe should have common initiatives and support mechanisms on this topic. The existence of common criteria will help identify which countries and regions need the most attention and where resources should be preferably allocated to prevent fire ignitions. On the other hand, we also saw

20 Towards Integrated Fire Management

Box 1. Wildfires and pastoralism – the Portuguese case study
The relationships between wildfires and pastoralism were studied in Portugal using a national database of about 135 000 fires that occurred during a six-year period (2002– 2007). A sample of 7337 fires with a known cause was used to investigate the particular characteristics that differentiate the fires that were caused by shepherds (renovation of pastures for livestock) from all the other causes. On average, pastoral activity was responsible for 20% of all wildfires and for 11% of the area burned. Shepherd-caused fires burned mainly shrublands (78%) and forests (18%). In contrast the other fires burned more forests (56%) than shrublands (37%). Logistic regression methods were used to analyze the relative importance of 26 potential explanatory variables and to develop fire ignition predictive models. The factors having the greatest influence on the probability of pastoral fire occurrence were the presence of forest, active population, recurrence of fires in the past, altitude and slope. Presence of forests and active population had a negative influence and the remaining variables had a positive influence. The tests performed to evaluate the models showed good accuracy on predicting fire occurrence (82–85%, ROC). Concerning topography, results showed that fires associated with pastoral activity occurred at higher elevation and in more irregular terrain than the other fires. These fires also started more often in areas with lower human population density/activity, more distant to the major roads and where fires burned more frequently in the past. They were also more frequent in areas with more shrublands and less forests when compared to the other fires. Fires due to pastoral activities also showed a different temporal pattern, occurring mostly from mid-summer to mid-autumn (Figure 6). However large annual variations in the fire season were observed, possibly being related with the annual meteorological conditions and food availability for cattle.

Figure 6. Comparison between fires that were started for pastures renovation (6-year averages): main land cover classes affected (left); and annual distribution of fires (right). Reproduced from Catry and Rego (2008); (EU Forest Focus Programme: UTAD, AFN).

Fire Starts and Human Activities 21

that fires of different causes can have different spatial and temporal patterns. Thus having more information about fires investigated (with known causes) and analysing them separately by cause type (e.g. as we did for shepherd-caused fires), will also contribute to improving existing knowledge. Further important research topics should also include measures of temporal fire activity, such as predictions on number of fires per day in relation to natural and anthropogenic variables. Additionally, as only a low proportion of ignitions result in large burned areas, from a management perspective it would be important to further understand the characteristics of ignitions that result in large fires, as these should become a priority for initial attack and suppression. This was addressed in the case of Portugal, but more research is needed on this important topic.

References
APAS 2004. Estado del Conocimiento sobre las Causas de los Incendios Forestales en España. Asociación para Promoción de actividades Socioculturales, Madrid. 54 p. Badia-Perpinyà, A. and Pallares-Barbera, M. 2006. Spatial distribution of ignitions in Mediterranean periurban and rural areas: the case of Catalonia. International Journal of Wildland Fire 15: 187–196. Bassi, S. and Kettunen, M. 2008. Forest Fires: Causes and Contributing Factors in Europe. European Parliament, Brussels. 56 p. Cardille, J.A. and Ventura, S.J., Turner, M.G. 2001. Environmental and social factors influencing wildfires in the Upper Midwest, USA. Ecological Applications 11(1): 111–127. Catry, F.X., Damasceno, P., Silva, J.S., Galante, M. and Moreira, F. 2007. Spatial distribution patterns of wildfire ignitions in Portugal. Proceedings of the 4th International Wildland Fire Conference. CD. Seville, Spain. Catry, F.X., Rego, F.C., Bação, F.L. and Moreira, F. 2009. Modelling and mapping the occurrence of wildfire ignitions in Portugal. International Journal of Wildland Fire 18: 921–931. Catry, F.X., Rego, F.C., Moreira, F. and Bação, F. 2008. Characterizing and modelling the spatial patterns of wildfire ignitions in Portugal: fire initiation and resulting burned area. In: de las Heras J., Brebbia C.A., Viegas D. and Leone V. (eds.). International Conference on Modelling, Monitoring and Management of Forest Fires. WIT Press, Vol. 119: 213–221, Toledo, Spain. Catry, F.X. and Rego, F.C. 2008. Caracterização e análise dos padrões temporais e espaciais dos fogos relacionados com o pastoreio. In: Botelho, H.S., Bento, J.M., Manso, F.T. (Coords.) A relação entre o pastoreio e os incêndios florestais. UE Forest Focus Programme. UTAD, AFN. Chuvieco, E., Allgöwer, B. and Salas, J. 2003. Integration of physical and human factors in fire danger assessment. In ‘Wildland Fire Danger Estimation and Mapping. The Role of Remote Sensing Data’. Chuvieco, E. (ed.) Vol. 4: 197–218. EC 2008. Forest fires in Europe 2007. Report No 8. European Commission, Joint Research Centre, Institute for Environment and Sustainability. Ispra, Italy. 80 p. FAO 1986. Wildland fire management terminology. FAO Forestry Paper 70, Food and Agriculture Organization of the United Nations. Finney, M.A. 2005. The challenge of quantitative risk analysis for wildland fire. Forest Ecology and Management 211: 97–108.

22 Towards Integrated Fire Management

Goldammer, J.G. and Crutzen, P.J. 1993. Fire in the environment: Scientific rationale and summary of results of the Dahlem Workshop. In: Goldammer, J.G. and Crutzen, P.J. (eds.). Fire in the environment: The ecological, atmospheric and climatic importance of vegetation fire. John Wiley & Sons, Toronto. Pp. 1–14. Lloret, F., Calvo, E., Pons, X. and Dìaz-Delgado, R. 2002. Wildfires and landscape patterns in the Eastern Iberian Peninsula. Landscape Ecology 17: 745–759. Loboda, T.V. and Csiszar, I.A. 2007. Assessing the risk of ignition in the Russian Far East within a modelling framework of fire threat. Ecological Applications 17 (3): 791–805. Mazzoleni, S. 2007. Spatial and temporal statistics on wildfire regimes across Europe. Deliverable D4.1-1a of the Integrated Project “Fire Paradox”, Project no. FP6-018505. European Commission. 14 p. Mercer, D.E. and Prestemon, J.P. 2005. Comparing production function models for wildfire risk analysis in the wildland–urban interface. Forest Policy and Economics 7: 782–795. MMA 2007. Los incendios forestales en España. Decenio 1996–2005. Ministerio de Medio Ambiente, Centro de Coordinación de la Información Nacional sobre Incendios Forestales. Madrid, Spain. 106 p. NWCG 2006. Glossary of wildland fire terminology. National Wildfire Coordinating Group, PMS 205 (Boise, ID). Pausas, J. and Keeley, J. 2009. A Burning Story: The Role of Fire in the History of Life. BioScience 59: 593–601. Thirgood, J.V. 1981. Man and the Mediterranean Forest a History of Resource Depletion. Academic Press, San Diego 194 p. Turner, M.G. and Romme, W.H. 1994. Landscape dynamics in crown fire ecosystems. Landscape Ecology 9: 59–77. UN 2002. Forest fires in Europe 1961–1998. International Forest Fire News 27: 76–80. FAO/UNECE, United Nations (UN). Available at: www.fire.uni-freiburg.de/iffn/iffn_27/ content27.htm Vasconcelos, M.J.P., Silva, S., Tomé, M., Alvim, M. and Pereira, J.M.C. 2001. Spatial prediction of fire ignition probabilities: comparing logistic regression and neural networks. Photogrametric Engineering & Remote Sensing 67(1): 73–81. Vasilakos, C., Kalabokidis, K., Hatzopoulos, J., Kallos, G. and Matsinos, Y. 2007. Integrating new methods and tools in fire danger rating. International Journal of Wildland Fire 16(3): 306–316. Vazquez, A., Moreno, J.M. 1998. Patterns of lightning- and human-caused fires in peninsular Spain. International Journal of Wildland Fire 8(2): 103–115. Vega-Garcia, C., Woodard, P.M., Titus, S.J., Adamowicz, W.L. and Lee, B.S. 1995. A logit model for predicting the daily occurrence of human caused forest fires. International Journal of Wildland Fire 5 (2): 101–111. Vélez, R. 2009. The causing factors: a focus on economic and social driving forces. In: Birot, Y. (ed.). Living with wildfires: what science can tell us. EFI Discussion paper 15. European Forest Institute, Joensuu. Pp. 21–25. Weiner, S., Xu, Q.Q., Goldberg, P., Liu, J.Y. and Bar-Yosef, O. 1998. Evidence for the use of fire at Zhoukoudian, China. Science 281: 251–253. Yang, J., Healy, H.S., Shifley, S.R. and Gustafson, E.J. 2007. Spatial patterns of modern period human-caused fire occurrence in the Missouri Ozark Highlands. Forest Science 53 (1): 1–15.

whatever the problems and difficulties faced by the fire fighters to control and finally to extinguish it.2. absorbing energy ). step-by-step. The front shape will be modified by changes in the burning conditions (gusts of winds. by the combustion of the fuel particles located close to the ignition point. Fire ignition regroups fuel reaction to energy inputs (drying and thermal degradation as endothermic phases. (phase of pyrolysis) flammable gases associated with the oxygen of the air produce carbon-monoxide and carbon-dioxide and water. France 2. ignited at a given point or on a given small area. Fire initial extension is piloted.2 Flammability: Influence of Fuel on Fire Initiation Barbara Ubysz1 and Jean-Charles Valette2 1 Forest Research Institute. local relief) and local fuel heterogeneities. energy is input either by conduction (contact).2. releasing energy). charcoals and oxygen produce ashes and carbon-monoxide and carbon-dioxide. The phase of the thermal degradation of a fuel particle is endothermic (i. 2. water is vaporised from the fuel particle. if the burning conditions are stable and if the fuel is homogeneous. radiation and convection. During this initial phase. . Ignition is the consequence of the mix of fuel particles. which grows in accordance with its own dynamic during the initial phase and may continue to spread despite the existing prevention means and the fighting means used to reduce its intensity and its rate of spread.e.e. Avignon. oxygen and heat input at a given location and at a given moment. the centre of the circle is the ignition point and the radius of the circle is related to the spread rate of the fire. Poland 2 National Institute for Agricultural Research. and whatever the loss of human life and the associated damage. The basic phenomenon can be split into three sequential steps: 1.1 The fire initiation phase Fire initiation consists of two phases in a wildfire growing process: (i) fire ignition and (ii) fire initial extension (until the arrival of the fire fighters). Raszyn. 3. the pattern of the propagation of the fire front is circular. Each catastrophic wildfire is always a fire. The initial phase is common to all fires whatever the final size of the burned area. and combustion of flammable gases and charcoal as an exothermic phase). the combustion phase either of flammable gases or of charcoals is exothermic (i.

branches. Fuel material. heat content of wildland and forest fuel decreases with the richness of the tissues in cellulose and increases with their richness in lignin. particularly issued from Mediterranean tree and shrubs species. 2008). this characteristic largely contributes to the ignition and spread of fire (Alessio et al. Heat capacity (the amount of energy needed for heating the material) and heat content (the amount of energy released by the combustion of the material) are two important parameters for characterising wildland and forest fuel. Heat capacity varies between 0. thermal and physical characteristics Given their role in the basic mechanisms. and their physiological state. fire fighters and wildland managers cannot act directly on these combustion phases and must act on other parameters to limit the energy released and to reduce the fire extension. where the first step of the combustion phase consumes the available oxygen very rapidly and produces large amounts of black and toxic smoke. partially decomposed (litter and duff) whose characteristics are altered during the decomposition process. 2. Consequently. chemical. varies from 18 500 kJ/kg (oak leaves) to 24 000 kJ/kg (pine needles and Ericacea foliage) (Dimitrakopoulos and Panov 2001): these species are considered as highly energetic. consequently. Heat content depends on the richness in organic oils and in other organic volatile compounds. such smoke can also be produced during large and intense wildfires. nevertheless. Wildfires are open fires where oxygen is not the limiting factor. their stage of development. their environment.1 Chemical. and depending on the species. 2006). Szczygieł et al.4 and 1. 2007. Eufirelab teams promoted common methods to determine these characteristics (Allgöwer et al. Baker (1983) indicates a heat content from 18 600 kJ/kg to 23 200 kJ/kg for cellulose and hemicelluloses. 2008). twigs of various diameters. Fernandes et al. industrial and structural fires ignite in closed spaces. oxygen and hydrogen. is rich in organic oils and in organic volatile compounds and terpenes (Pinaceae and Cupressaceae) whose flash points are low.4 kJ/kg/K and heat content of most Mediterranean species. Wildland and forest fuel is basically constituted of living material (needles and leaves.24 Towards Integrated Fire Management In general. and parts of the root system) and of dead material.2. thermal and physical characteristics of fuel particles have been largely studied specifically to provide input data to the models predicting the behaviour of fires.2. (1992) established the relationships between heat and ash contents and flammability parameters and confirmed that ash content is fairly stable and does .2. either shrubs and trees. trunks. Fuel material is essentially composed of carbon. and 25 600 kJ/kg for lignin. In addition. nitrogen. and Fire Paradox partners have improved the cone calorimeter method for these types of fuel particles (Rein et al. this material also contained oligoelements which intervene in the photosynthesis and growth processes but which very rarely plays a role in the initiation processes of a fire.2 The role of fuel particles characteristics 2.

. In contrast. For more complicated shapes. its moisture content depends only on the water exchanges regulated by physical laws. 2007. Physical characteristics of the wildland and forest fuel contribute to the ignition and initial propagation processes. The subdivision of the finest class in two classes (more than 2 mm and less than 2 mm) allows the user to separate leaves. bark) or their diameter (twigs and branches). For live fuels. It increases the concentration of nonflammable gases in the mixture of gases of combustion. Fernandes and Rego (1998) proposed to evaluate the surface by water immersion and weighing the mass of water which constitutes a film around the fuel particle. medium (>6 mm to 25 mm).Flammability: Influence of Fuel on Fire Initiation 25 not significantly intervene in the combustion process as long as it remains lower than 5%. the easier the heat will be intercepted. dead material is not regulated by physiological mechanisms. or at least part of it. Surface and volume of leaves. The moisture content of a given fuel is obtained by drying the fuel and calculated with the commonly used formula FMC = 100 (Mi –Mf) / Mf where Mi is the initial mass or wet mass of fuel and Mf is the final mass or its dry mass. this heat is released by the initial source or by the combustion of fuel particles and transmitted either by radiation or by convection. Each species adopts specific strategies to resist drought.2. fine(>2 mm to 6 mm). Surface-to-volume and mass-to-volume ratios are two of the most important parameters for characterising the fuel particles. The larger the surface exposed to heat. needles and twigs can be determined through geometrical approximations detailed by Eufirelab members (Allgöwer et al. given that very fine. and very fine (≤2 mm). and reduces their ability to initiate and sustain the combustion process by increasing the endothermic part of the combustion reaction. needles and fine twigs from larger fuel particles and to determine more accurately the mass of the fuel consumed at the fire front of a prescribed burning. Fuel particles have been classified according to their thickness (leaves.2 Fuel moisture content Fuel moisture content (FMC) also plays a paramount role in the reaction of wildland and forest fuel to heat exposure. a given FMC is the balance between the water inputs throughout foliage and roots system and the water outputs by plant evapo-transpiration.2. Water vaporisation uses an important amount of energy due to the high heat capacity of water. Eufirelab members proposed to classify Mediterranean fuel particles in four classes: coarse (>25 mm). and to limit its own water reserves at a level which permits the survival of the plant. Fernandes et al. fine and medium fuels are generally consumed at the fire front of a wildfire. 2006). The limits of the three first classes are the same as the American standards. This water immersion method also allows the user to determine the volume of the fuel particle by adapting the pycnometer method to the specificities of the fuel particles. 2.

some end-users prefer to express the moisture content using a different formula: WFMC = 100 (Mi – Mf) / Mi FMC can easily be deduced from WFMC by this third formula FMC = 100 WFMC / (100 – WFMC) It is assumed that the mass loss (Mi-Mf) is exclusively the mass of water which was inside the material and vaporised during the drying period. Szczygieł et al. as an example. because at this high temperature proteins are destroyed and oils and other organic compounds are either decomposed or volatilised. Because these two methods cannot be easily applied outside. Through calibration tables. the mass of the sample. In general. but generally only during the driest and hottest period of the day in order to catch the relative minimum which corresponds to the peak of the daily fire danger. the Polish network of observation sites and the deduced map of fire danger (Figures 1 and 2). Eufirelab members proposed and adopted 60°C for live and dead fuels. teams adopt a lower drying temperature. and consequently the moisture content. many Mediterranean. Depending on the use of the parameter. This device provides quicker results but does not allow the users to carry out simultaneous replications or to compare simultaneously several fuel types. FMC can be available daily or every two or three days. Other teams prefer to use a moisture “analyser” or “determination balance” constituted by an infra-red dryer placed above the plate of the electronic balance: this analyser displays the initial current and final weights during the drying process and determines in real time the FMC. and more broadly European. Fire managers may adapt immediately their tactics and strategies as soon as the value is displayed. and data are displayed as frequently as required. Australian teams have designed and developed a device based on the measurement of the electric resistance of a pellet of a pellet made from ground-up fuel (Chatoo and Tolhurst 1997). rainfall and wind regime).26 Towards Integrated Fire Management Unfortunately. Eufirelab members recommended a duration of 24 hours. (2009) and Ubysz and Szczygieł (2009) present. Using this temperature level for drying wildland and forest fuel over-estimates the mass loss. and more commonly. the size of the containers: in order to be sure to reach the oven-dry mass and for simplifying the manipulation. for example the practitioners may stop the prescribed burning when the moisture content of the litter or of the duff falls below the prescription range. and of the density of the network of plots where the fuel is collected. . this temperature threshold respects the proteins and limits the vaporisation of the largest part of the organic compounds. or several species. the FMC is directly displayed on the screen of the device. The international standard drying temperature is 105°C. 10 hour time-lag fuels might be acquired at a “high” frequency. Fire Danger Systems are based on fuel characteristics and on topographical (slope and aspects) and meteorological data (local or regional temperature. Moisture content of 1 hour. the structure of the fuel. This duration depends on the type of oven. cement or plaster. They combine their monitoring with the monitoring of the moisture content of dead fuels (litter) and/or of living fuels (leaves of shrubs and trees). applied for inert materials like sand. Meteorological data can be continuously monitored. Therefore. This lower temperature induces a longer duration of exposure to reach the stabilised mass.

Flammability: Influence of Fuel on Fire Initiation 27 Figure 1. Polish network of fuel moisture content measurements. Polish deduced fire danger information map. . Figure 2.

Weise et al.2.28 Towards Integrated Fire Management Figure 3.3 Fuel flammability To update maps of fire hazard. Chuvieco et al. This is generally based on the knowledge acquired through previous events. Summer variations of Erica arborea FMC. During the summers 2002 and 2005. these thresholds were respectively reached after 8 to 17 days and 17 to 28 days. the maps need to be deduced from meteorological and fuel moisture content data.2. Moro 2006). and the first half after 40 days in 2005 and 47 days in 2002. Figure 3 synthesises the evolution of the FMC of Erica arborea samples collected during six dry summer periods (2001 to 2006) on one plot of the French network which covers the French Mediterranean region (Forest Focus project). In a recent synthesis paper. the drought was not too severe and Erica arborea was able to limit the decrease of its FMC: the first third of the variation was reached after 25 days in 2005 and 30 days in 2002. 2001–2006. (2009) demonstrate that live fuel moisture content can be used to predict the behaviour and the occurrences of wildfires in specific Mediterranean ecosystems. (2005) in preliminary tests indicate that the cone calorimeter can also be used for . During the other summer periods. 2. logical evolutions and also flammability data. Methods and main results of other studies on the flammability of wildland and forest fuel have been described (Valette and Moro 1990. This kind of monitoring of several species in a given plot and several plots at a given date improves the accuracy of the fire danger prediction by including biological data.

84 MIT = 6. either linear or exponential.org/specific. and of the resulting flame height and fuel consumption ratio. Guijarro et al. Even though the approaches and methods are different.Flammability: Influence of Fuel on Fire Initiation 29 Figure 4. the common species are classed at the same level of flammability. Figure 4 summarises these results and a general equation. Velez (1990) and Dimitrakopoulos (2001). linking FMC to the Mean Ignition Time (MIT): MIT = 0. Dibble et al.82 Measurements of flammability parameters permit the classification of species. Moro offers results on Erica arborea. (2007) indicate that the cone calorimeter provides data that enable comparison of combustion properties among many species.117 R² = 0. measurements have been carried out during the summers from 1989 to 2007. Fire Paradox partners improved the procedure in order to characterise wildland and forest fuel and to provide the fire behaviour modellers with accurate values of required input (Rein et al. 2009). Finally. 2008. relating time to ignition and moisture content. based on the flammability of the dead litter fuels.010FMC R² = 0. it implies a decrease of the risk of fire initiation. (2007) proposed a classification of wildland urban interface areas. which is of highest interest during the fires seasons.767*e 0.173*FMC + 2. . Madrigal et al. Mean Ignition Time of Erica arborea versus Fuel Moisture Content. Jappiot et al. The following table summarises the classification proposed by Valette and Moro (1990). It also implies a decrease of the rates of initial fire spread and of combustion.eufirelab. (2002) described an adapted method for fuel beds and concluded that an increase of moisture content and of bulk density of litters implies an increase of the time to ignition. On www. the graphs illustrate how fuel flammability is largely dependent on fuel moisture content.php.

concerning the fuel bed. Erica arborea Greece Pinus halepensis. Martin et al. However. It underlines the paramount role of the variations of fuel moisture content on the variations of their flammability and consequently on the fire hazard they present. fuel moisture content cannot be easily maintained at a low risk level over large areas. the recommendations towards wildland areas and forests managers. owners. Alexandrian (2008) analyses the wildfire causes following the new classification used by the French Prométhée system. this is the case for many European roads crossing wildland areas and forests. Many teams have studied in laboratory conditions the mechanisms of wildland initiation by incandescent objects like sparks and cigarettes. Quercus coccifera Cistus salvaefolius Highly flammable Arbutus unedo. Italy and Spain) and summarise the procedure followed by the Spanish Nature Protection Service SEPRONA for wildfire investigation and prosecution. Xanthopoulos et al. the existing means are dedicated for very local application. local or regional authorities that can be deduced. etc. at least during the driest and hottest seasons. end-users. Likewise. particularly when the car speed is low and the road is narrow. (2006) demonstrate that the probability that fire may be initiated by a cigarette thrown from a car window is very small but not null.2. factories. Pinus brutia. thermal and physical characteristics of the fuel particles of the wildland and forest fuel at a reasonable cost.) located in or close to wildland areas and forests (Vélez 2009). Quercus coccifera. analyse the content of three of them (from France. workshops. will concern only limited areas. many treatments can reduce the fuel load and the fuel porosity. Pistacia lentiscus Pistacia lentiscus Moderately flammable Cistus salvaefolius Phlomis fruticosa. 2. These recommendations will be related to the identified wildfire causes. (2005) describe the existing databases. Quercus coccifera. fire fighters. It is quite impossible to modify the chemical. Quercus ilex Spain Pinus halepensis Quercus ilex Arbutus unedo.30 Towards Integrated Fire Management Flammability Extremely flammable France Pinus halepensis. Arbutus unedo. Cistus salvaefolius. .2. Quercus ilex. So.4 Conclusions The above overview illustrates the role of the chemical. mainly wildland urban interfaces and vicinities of structures (houses. thermal and physical characteristics of the particles of wildland and forest fuel on fire initiation.

Many fire initiations are related to works carried out in wildland areas and forests during the periods of high fire danger (drought. the initial energy released is very weak and the initial phase is long. • associating the remote monitoring of lightning impacts to the radar monitoring of rain to predict the localisation of fire initiation. where an enormous amount of energy is released and the initiation phase is very dynamic. consequently. recommendations should be in the sense of: • equipping professional workers with fire fighting tools. the extinction is easy as soon as the workers have some tools and equipments to control and extinguish the fire. presence of specialised personnel). • using historical data to identify lightning-prone areas. In both cases. active information panels with clear and simple recommendations towards the visitors. this might be a consequence of the global warming. • including the days or periods of high fire danger in the “bad weather” clause of the working contracts and. To reduce the number of fire initiations related to “forest” activities. • improving information to common citizens. • integrating the prevention of fire initiations in the work organisation and schedule. if the weather conditions are not too severe. wind). • making aware the workers to the specific fire risks of the area. to exclude any penalty in case of absence.2. Generally. local or regional authorities have to make them aware of the risk and to encourage them to change their personal behaviour. • reinforcing fire monitoring on lightning-prone areas and when the weather conditions are favourable to lightning and dry storms. lightning wildfires occur during dry storms. This type of wildfire is typically ignited at a given point. . efficient ash trays. this fire can smoulder for a long time and be detectable later after the departure of the workers. Their frequency seems to increase (Alexandrian 2008). develop a lightning detection system and improve the accuracy of the predictive model of lightning fires.3 Coping with fuel flammability – practical recommendations Except the areas concerned by volcanism which are well known. • forbidding any activity in wildland areas and forests during days or periods of high fire danger.Flammability: Influence of Fuel on Fire Initiation 31 2. • managing recreation areas to minimise the fire initiations (no car parks in shrub or herbaceous strata. Recommendations should be in the sense of: • reinforcing information on fire danger. • excluding burnings during the summer period. But. except for fire prevention and fighting. monitored and managed. Recommendations should be in the sense of: • reinforcing research to identify the conditions under which such phenomenon occur. Other fire initiations are related to unintentional and imprudent acts committed either by local inhabitants or by tourists coming mainly from regions or countries where the fire danger is generally low. At that step. In general the initial energy is weak and the fire initiation is detected early.

recommendations should be in the sense of: • managing along the highways a physical barrier to maintain cigarettes discarded from vehicles. the last category of wildland initiation causes is related to the acts of pyromaniacs and arsonists. The first sub-category concerns persons who need psychiatric treatments. • encourage the reduction of fuels on the interface and in the vicinities of the structures. .32 Towards Integrated Fire Management • prohibiting any introduction of “hotspots” in these areas like barbecues or ovens. highways. • encouraging the inhabitants to clean the immediate border of the houses. • reinforcing the efforts to reduce the amount of fuels below the electric power lines which can not be buried. • promoting less flammable building materials for houses. • encouraging railway companies to implement brakes that do not emit sparks and windows that prevent passengers throwing cigarettes outside. • making the inhabitants aware of the specific risks they will face during the high danger period. To limit the risk of fire initiation. • encouraging the burying of electric power lines (high. railways. development or introduction of shrub and tree species presenting lower flammability. • promoting on both sides of the highway the planting of native shrubs and tree species with low flammability. • installing garbage bins and encouraging visitors to take the garbage back with them. all the above recommendations are inefficient. medium and low voltage) in order to reduce the risks of power discharges between the line and the ground vegetation. van or lorry crashes. Initial energy can either be very weak (a spark from an engine. and develop adapted refuse centres. Finally. • collecting rain water and store it in a tank for fire fighting purposes. and in a first step to banish any access to them. arsonists are completely responsible of their acts. • managing the borders of open roads and paths by reducing the existing live and dead fuels. Other wildfire initiations are related to accidents or events which occur along or at the vicinities of equipments like electric power lines. On the contrary. • closing and suppressing refuse areas. rain gutters and woody shutters. • managing the borders of the railways by reducing the existing live and dead fuels. such as: • promoting the selection. Some management measures can be recommended in wildland urban interfaces and in the vicinities of structures built in or close to wildland areas or forests. a cigarette discarded from a car) or larger in case of car. roads and opened paths. including roof. • making the highways users aware of the local fire danger through active luminous signs.

. A statistical classification of Mediterranean species based on their flammability components. Chatto.. O. Luois. Eufirelab: Methods for Wildland Fuel Description and Modelling: Final Version of the State of the Art. L. A. Peñuelas. I.. Curt. and Fonturbel. 2007. Centre for Forest Tree Technology. Fernandes. Eufirelab: Methods for wildland fuel description and modelling: Commonly used methods.. A New Method to Estimate Fuel Surface Area-toVolume Ratio Using Water Immersion. Blanc. Fernandes. F. L. P. A. 2008.P. B. Hernando. R. IV International Conference Forest Fire Research. P. The development and Testing of the Willtronics T-H Fine Fuel Moisture Meter. Pérez-Gorostiaga. J. Cuiñas. and Lebow. 2007. Pyric properties of some dominant Mediterranean vegetation species. Seville. Molina. Flammability of some fuel beds common in the South-European ecosystems. twenty or thirty years. Fire Management Branch. Vinet. M. Prediction of fire occurrence from live fuel moisture content measurements in a Mediterranean ecosystem.. R. Baker. Hernando. S. and Séro-Guillaume.. D. 44 p. C. Research Report No.. LM. Dimitrakopoulos. 2001. P. J. J. Deliverable D-02-06.. D. C. and De Lillis.. Lindberg. Morsdorf.. Fuel Wood Management and Utilization Seminar. 2002..Y.. Chandioux. and Yebra. R. 2009. E. A. Poster at the International Wildland Fire Conference. and Estève. A. Dibble.. References Alessio. Madrigal. K. M. regional and national levels.C. I. Deliverable D-02-05. and Tolhurst.C. M.. Michigan State Univ.C. International Journal of Wildland Fire 8: 121–128.A... Colin. Wood fuel properties and fuel products from woods. D.... 2007. When the arsonists’ acts are based on soil occupation.Flammability: Influence of Fuel on Fire Initiation 33 These acts may be based on neighbours’ disputes or related to hunting dissensions. M. this prohibition can be applied during the following ten. and Valette.. M. East Lansing. MI. 9–11. O. Ribeiro. In Mediterranean areas where the pressure of the landed property is important.. and Rego. P. 7p Guijarro. Francesetti... Ribeiro.K. International Journal of Wildland Fire 17: 274–286. Allgöwer. CFTT. Alexandrian. Chuvieco. White. Mendes-Lopes. 2006. Koetz. J. 1983. 1997... which have been partially or totally burned. Pp 14–25. . Martínez. Verdú. González. J. C. International Journal of Wildland Fire 16: 426–443. Lampin-Cabaret. Lampin. 13–17 May 2007. T. Forêt Méditerranéenne XXIX(4): 377–384...T. 2001.M. A. P. C. Combustion characteristics of northeastern USA vegetation tested in the cone calorimeter: invasive versus non-invasive plants. Morsdorf.. Hernando.. Guijarro. Calogine. 1982. Spain. Vega.. F.. Ogaya. N. International Journal of Wildland Fire 10: 113–118. International Journal of Wildland Fire 10: 23–27. Marzano. Camia. H. a dissuasive decision is to prohibit the building of houses or any other structures into wildland areas or forests. Dimitrakopoulos.. Characteristics and Flammability of French Mediterranean Dead Litter Fuels.I. Cuiñas. P. E. International Journal of Wildland Fire 18: 430–441. 2008. K.A. B.P and Panov. Allgöwer. Nov. 1998. F.. B. Borgniet.. such a regulation must be strongly supported by the population and the public authorities at local. M. P... mediator’s interventions may be efficient to solve problems... Estiarte. O. Llusià.. P.. Díez. M. Fernandes.. F. 57 p. Koutsias.. 46.. G. Influence of water and terpenes on flammability in some dominant Mediterranean species. C. In: Proc. R. Aguado. J. P. C. A..J.H. L.. Department of Natural Resources and Environment. Luso.M. E. Les statistiques “feux de forêt” de ces trente dernières années.. 18–23 November 2002 Jappiot. Rigolot.

Lampin. 2009. 14 p. Y. 1990. Hernando. J. Joensuu. D. Ghosn. 21–25. 2009. Piwnicki. FP6-018505. Inflammabilités des espèces forestières méditerranéenne. and Kwiatkowski. J. 2006. PIF1992-01.H. 54 p Moro. Project no. and Etlinger. International Journal of Wildland Fire 14: 321–338. Internal technical report. 15 p + annexes.R. A. 2008. Use of the cone calorimeter to detect seasonal differences in selected combustion characteristics of ornamental vegetation.C. In: Birot. Diez. Conese. C.. European Commission. Martinez. 2005.). B.. M. 94 p. Deliverable D-0502. Rein. Szczygieł. and de Castro. M. Teneur en cendres et pouvoir calorifique supérieur d'espèces forestières méditerranéennes : relations avec les paramètres d'inflammabilité. Molina. Guijarro. and Salas. R. European Commission. Martin.. Fire flammability of forest plant formations in Poland. Vélez. A. D. R. G. and Moro. Eufirelab: Towards methods for investigating on wildland fire causes. 2009. B. Méthode de mesure de l'inflammabilité du combustible forestier Méditerranéen. 34 p. Velez.C. R. Unasylva 41. 1990. Valette. L.34 Towards Integrated Fire Management Madrigal. .. and Kazakis. Pp. European Forest Institute. (ed..J. Valette. 2009. Project no. and De Castro.. Evaluation of forest fuel flammability and combustion properties with an adapted mass loss calorimeter device... Beall. Bonora. G. C.. C. 2005.C. Revue Forestière Française XLII numéro spécial 1990.. M. M. C.. J. Methods for the experimental study and recommendations for the modelling of pyrolysis and combustion of forest fuels. Fire Situation in Poland. International Journal of Wildland Fire 15: 567–576..2-3 of the Integrated Project “Fire Paradox”. 83 p. 2006. R.. 76–92 Weise. FP6-018505. Deliverable D2. Esko. Living with wildfires: what science can tell us. C.. and Moro. J. The causing factors: a focus on economic and social driving forces. 1992. D... White. J. F. Ubysz. and Szczygieł. M. EFI Discussion paper 15. Preventing forest fires through silviculture. C. Ubysz. Torero. Investigation of the wind speed threshold above which discarded cigarettes are likely to be moved by the wind. Journal of Fire Sciences 27:323–342. R. R. Pp. Szczygieł. Xanthopoulos. conséquences sur la combustibilité des formations forestières. Instytut Badawczy Lesnictwa. Marino. J. J.1-1 of the Integrated Project “Fire Paradox”... Internal Report. Deliverable D4. P. Guijarro. E. G. C.

especially in Southern Europe. which makes ecosystems more fire prone and increases the associated risk of large wildfires. European Commission – FAO. For example. Rome.g. wildfires are a subject of growing concern. in part because the abandonment of rural areas. This holistic and proactive approach to wildfire management (Morugera and Cirelli 2009) needs different formulations in Europe. However. In countries along the northern rim of the Mediterranean Basin. University Complutense of Madrid. In general.1 Introduction Recent socio-economic changes. as much as possible. the shift from suppression-oriented policies to preventive and integrative policies requires also an investigation of the human causes contributing to wildfires in Europe1. the prolonged protection of forest lands (and consequently fire exclusion from the ecosystem). Sabaudia. Corpo Forestale dello Stato. Established fire policies provide the framework for social prevention actions. wildfire issues receive different consideration within the national policy instruments depending on the risk severity of wildfires in the national contexts and the different political and administrative systems existing in each country. the process to induce changes in wildfire policies or the adoption of new political measures has been a fast ad hoc reaction to a past catastrophic situation. land-use change dynamics have aggravated fire hazards and disaster potential. environmental and nature protection policies) have contributed to change the structure of some ecosystems enabling the build-up of fuels.3. This situation requires a shift in paradigm from a short-term fire policy. Italy – Joint Research Centre. rather than proactive mitigation before an emergency arises (FAO 1999). Until now. based mainly on technological investments for emergency suppression measures. the forest sector is facing new realities that are causing changes in the European wildfire context.3 An Overview of Policies and Practices Related to Fire Ignitions at the European Union Level Cristina Montiel and Gema Herrero Research Group Forest Policy and Socioeconomics. some forest management actions and other policy measures outside the forest sector (e. Thus. in Northern and Central European countries wildfires have been historically far less serious. 13–15 May 2008 .2. Spain 2. Italy. In fact. Report. potential damage (Birot 2009).. including new governance methods and pre-extinction measures (Aguilar and Montiel 2009a). the wildfire problem has become worse during the second half of the 20th century. to a longer-term preventive policy of removing the structural causes of wildfires and reducing. and the expansion of wildland 1 Workshop on Forest Fires in the Mediterranean Region: Prevention and Regional Cooperation.

but also to socio-economic causes that affect fire ignitions (Vélez 2009). and should be tackled differently at the regional and national levels. the pattern and spatial distribution of the fire problem in Europe are different in the different regions and countries. causes. the LIFE+ instrument3 co-finance communication and awareness campaigns for fire prevention. while civil protection policies aim to protect human lives and goods. • Support for scientific research on various wildfire topics. The best instruments to introduce innovative approaches and recommendations for the necessary paradigm shift towards a solution to the current wildfire problem are a combination of legislation together with policy measures. EEC Nº 2158/92 on protection of the Community´s forests against fires. 2009. Vélez 2002. However. Regarding support to the national prevention policies. In Europe. Martínez 2004.36 Towards Integrated Fire Management urban interface areas (Galiana et al. The European Union (EU) has also considered the wildfires problem in their political agenda since the 1980s through various regulations2 that have encouraged the Member States (MS) to reinforce prevention of wildfires. 2.3. 3 The LIFE+ Regulation (EC) No 614/2007 is the new EU financial instrument for the environment for the period 2007–2013 and includes provisions for funding forest fire prevention projects. Information derived from the EFFIS database suggests that the pattern of wildfires is not related only to climatic conditions. Vilar del Hoyo et al. and efficient targeting of community funds to prevention.2 Description and characterization of policy instruments Depending on the decentralization level of European countries. Recently. risk levels) to create a common EU Fire Database. while the Rural Development Regulation constitutes the main source of investment in infrastructure for wildfire prevention. the main initiatives undertaken by the EC regarding fire prevention include: • The European Forest Fire Information System (EFFIS) that collects information on forest fires (e. taking into account wildfire incidence and socio-economic conditions. 2008). In agreement with national administrations. • Financial provisions to MS. developing knowledge. National and Regional Forest Plans usually include preventive and curative actions to mitigate wildfire hazard. Forest Focus Regulation EC Nº 2152/2003 concerning monitoring of forests and environmental interactions in the Community.g. . issues of wildfire management are handled principally within forest and civil protection policies at the national and regional level. about 95% of wildfires are directly or indirectly human caused. linking actors and policies. 2 Regulation EEC Nº 3529/86 on forest protection against fire. for example. the European Commission (EC) has adopted a Communication on the prevention of natural and man-made disasters (COM(2009)82) that focuses on subjects where a common approach is more effective than separate national approaches.

therefore. Wildfire legislation and zoning policy in Europe (by Montiel.An Overview of Policies and Practices Related to Fire Ignitions at the European Union Level 37 Figure 1. The Mediterranean Basin countries have by far the most and significant policies. Regulations are complemented and translated through planning instruments outlining specific actions. To prevent and reduce fire ignitions. Policy measures in each country are influenced by the perceived level of threat. national legal and policy frameworks should be adapted to existing ecological and socio-economic characteristics of each European region (see Figure 1). Fire occurrence and incidence are different from region to region. the key issues considered within the national and regional policy frameworks are the following: . In the rest of Europe (especially in Northern countries) forest production goals are more important than forest fires in forestry management. C. which varies with the intensity and scale of fire-related problems. UCM).. and implementation of regulations about fire prevention.

Likewise. In addition to forest and civil protection policies.g. special measures). it is common to prohibit the use of fire in agro-silvo-pastoral and leisure activities outside of the designated areas. However. it is a common and relevant issue in national fire plans. Portugal and some regions in Spain and in Italy). Mediterranean countries define longer highrisk fire seasons than Northern regions.g. like the wildland urban interface. • Detection systems: detection also plays an important role in reducing fire ignitions and fire propagation. agricultural and rural development policies. (ii) the recognition of agriculture’s territorial . etc. there are also other public policies – spatial planning. energy policy and environmental policy – that influence the structural causes affecting the wildfire problem (Lázaro et al. • Spatial planning has the potential to adequately coordinate all interventions needed to respond to natural hazards (floods. • Fire prevention campaigns: although public information and social awareness about fire prevention receives relatively little attention in national legislation. 2009).38 Towards Integrated Fire Management • Causes investigation: although the investigation of wildfire causes needs further development in many countries. • Placement of silvicultural treatments: (removal of brush. thinning or pruning) are carried out adjacent to hazardous infrastructures areas either to prevent fire ignitions in the area. France. • Fire risk zoning: this means the demarcation of areas that are potentially hazardous and where there is an increased fire risk. The duration of a high-risk fire risk season differs from country to country and is a function of the climate and the seasonality of fire causes. restrictions. wildfires. this policy may affect one of the most important causes increasing wildfire risk: the scattered settlements in forest areas. alongside urban policies. • Use of fire as a tool: when used for achieving prevention objectives.g. • The evolution of the Common Agricultural Policy towards a real Rural Development Policy brings about significant opportunities for intervention on wildfire management such as: (i) the linking of wildfire prevention objectives to local development processes. and/or to prohibit the use of fire at specific times of the year. or to prevent wildfires spreading from these areas. • Regulation of hazardous activities: in countries where these kinds of measures are established. These areas normally receive a special legal status (e. • Determination of a ‘high-risk’ season: the declaration of high-risk fire seasons is a decision taken by individual Member States. to avoid fire starts there are measures that regulate existing hazardous installations or their location. require special attention because specific preventive measures might be needed. Greece) to those that have developed regulations and basic criteria and conditions for the use of fire (e. it is an essential activity to reduce fire ignitions. In determining fire risk zones new fire prone areas. prescribed burning is directly related with the strategic reduction of fuel accumulation and the avoidance of fire starts because of uncontrolled rural burnings.) to achieve a more sensible use of the territory. and it is becoming a more important priority in the Mediterranean region. • Existence of a legal framework for burning: the legal frameworks in the Mediterranean Basin range from countries that do not contemplate the use of fire and even prohibit it (e.

The construction of new residences (first or second homes) near surrounding vegetation increases fire risk. Other factors influencing the temporal variation of fire regimes during recent decades. they can also be considered an indicator of the socioeconomic differences between areas and their respective levels of development. which are the result of 4 Self Assessment. especially in the Mediterranean area. are more than just a consequence of periods of drought. changes in landscape configuration. Southeast Europe. which address the multiple dimensions of forest ecosystems such as those related to the economy. natural resources and the environment (see Figure 2) 2. . and therefore in fire risk. • Shifts in forestry policy priorities from the production of wood and other raw materials to focusing on nature conservation. conclusions and recommendations of the Regional Session: Europe. increasing vulnerability to wildfires. changes in ignition causes and the success of predominant fire suppression policies should be also considered. such as climate change.3. an important segment of the causes contributing to the wildfire problem is left out of the traditional policy domains related to wildfire management (Forest and Civil Protection policies) and must be supported by other sectoral policies. and (iii) the possibility to take advantage of the new energy priorities to increase the use of forest biomass as a source of energy. favours its consideration as a priority for environmental management. The most relevant are4: • Rural depopulation. Spain. which may lead to increases in build-up of fuels. which increases the wildland urban interface.An Overview of Policies and Practices Related to Fire Ignitions at the European Union Level 39 role in wildfire prevention.3 Territorial policies and underlying wildfire causes Wildfires. 2.4 Risk practices in new fire scenarios: the need of social prevention adapted measures The main components of the wildfire problem in Europe – fire and territory – have substantially evolved in the last decades. A subset of this factor is the aging of rural population. Mediterranean North Africa and Caucasus. Some of the factors influencing fire regimes in recent decades are closely related to changing socioeconomic conditions. Fire behaviour is evolving because there are new fuel patterns. which also increases wildfire risk because of the loss of know-how in traditional burning practices. 2007). especially in the Mediterranean region. • The growing perception of wildfires as one of the main environmental problems for Europe. Therefore.3. landscape management and recreation. held in 4th International Wildland Fire Conference (Seville. • Concentration of population in urban areas. leading to the abandonment of rural areas and changes in landscape because the invasion by natural and pioneering vegetation.

40 Towards Integrated Fire Management Figure 2. The use of fire as a management tool in rural activities has been generally maintained. in which forests have reclaimed the abandoned lands. the existing fire culture can determine a type of ‘fire-dependent rural communities’. the use of fire by a neo-rural population or by users of wildland recreational areas becomes a new ignition risk factor because they are not able to guarantee fire control.. where fire use regulation and governance mechanisms for fire management are strongly needed. showing different fire ignition patterns and different vulnerability to wildfires. hence turning its use into a risk factor. Territorial policies in wildfire management (by Lázaro. The multiple combinations of fuel patterns. with no significant problems of rural abandonment. and without significant increase in new urban pressures. At the regional scale. suffering from rural abandonment. A. and therefore needing different fire management strategies: • Scenario 1: Disadvantaged rural areas. . The aging and declining population has also meant the loss of the fire culture. uncontrolled development and biomass accumulation. which require new approaches to fire management. These are rural areas in crisis. Spatial and human concerns are also defining new territorial models. These areas are characterized by a consolidated socio-economic context and productive forests. UCM). • Scenario 2: Dynamic rural areas. new trends in spatial-planning and new social realities. Today the use of fire in traditional rural activities is set in different spatial conditions in which continuity of woodlands and fuel load results in an increase in wildfire risk. have resulted in new fire scenarios. decline or abandonment of agrarian activities. Furthermore. it is possible to recognize three territorial scenarios (see Figures 3 and 4). which create emerging fire prone areas and new territories at risk (Vélez and Montiel 2003). However.

A.An Overview of Policies and Practices Related to Fire Ignitions at the European Union Level 41 Figure 3. ND: no data. PU: predominantly urban areas. IR: intermediate rural areas. UCM). Figure 4. Regional distribution of rural and urban areas in Europe (by Lázaro. Rural development index in European regions (by Lázaro... A. . PR: predominantly rural areas. UCM).

3. the negative effects of these national policies and the loss of traditional fire use knowledge has become evident by analyzing fire ignition frequency. social awareness and new governance mechanisms allowing the existing stakeholders in each scenario to participate in the integrated fire management scheme. in the dynamic rural areas scenario. on their adaptation to the spatial and socio-economic characteristics of the area under consideration. the development of new governance mechanisms and the improvement of decision support tools.5 New demands and challenges for wildfire management Fire has always been present in Europe. The three different types of fire territorial scenarios demand different policy and management measures regarding the biophysical and social components of the fire problem. Finally. Nevertheless the wildfire policies adopted by most European countries over the last century have been based on total fire exclusion. most importantly. and also adapting to the situation of low population levels and scarce human resources. 2. In the context of global change and new policy paradigms. the main political challenges to overcome are the adoption of long-term prevention actions. This strategy permits facing the main challenge of fuels management at lower costs. Alongside the need for empowerment of local groups. it is convenient to incentivize the culture of fire in rural communities by the enhancement of best practices of fire use and. Social preventive measures refers to public information. At present. In the disadvantaged rural areas territorial scenario. It has been widely used as a tool in rural practices. where agrarian activities have been abandoned. The effectiveness of interventions on structural causes affecting wildfires depends. measures should be taken with the intent to . and evaluation and proposal of policy changes. raising awareness of wildfire risk should be a priority in metropolitan rural areas and wildland urban interfaces. to promote regulations and new governance mechanisms to resolve conflicts that are at the origin of fire ignitions. the use of prescribed burning techniques by professional teams to carry out fuel management tasks is advisable. implementation and enforcement. to a large extent.42 Towards Integrated Fire Management • Scenario 3: Suburban rural areas and wildland urban interfaces. always guaranteeing the safety of fire fighters and the inhabitants of burned areas. elaboration and formulation. In addition. Nevertheless. learning processes related to the appropriate use of fire by rural populations and the reduction of human-caused wildfires can effectively contribute to reducing fire occurrence through the management of social conflicts and the development of new cooperation opportunities. These are areas with a predominance of urban developments. Adapting the national and regional programmes of social prevention measures to the fire territorial scenarios is very effective for reducing fire ignitions (Vélez 2010). The transitional areas between wildlands and urbanized spaces – the wildland urban interface – represents an increasing fire risk factor and are highly vulnerable. Governance in wildfire management needs to be participatory at all stages of the process including agenda setting and problem definition.

know-how transfer and training schemes are greatly needed. from information and planning to cooperation in pre-extinction measures. accountability when adopting preventive and remedial actions. spatial planning.. conflict resolution systems concerning fire use in the Mediterranean countries are greatly needed. as well as cooperation between stakeholders and public administrations. entails a social. 2. However. according to their knowledge. Central and Northern countries. interagency and multi-level coordination is essential for wildfire management. Finally. However. We need to move from a onedimensional perception of the negative impacts of fire to a more sophisticated one that also stresses its positive effects (Aguilar and Montiel 2009b). This approach would most likely result in wildfire management cost reductions. In fact. prescribed burning is an effective tool for fuels management. Further. rural development. and. namely the large and continuous areas with fuel models of high risk and the wildland urban interface territories (Galiana et al.6 Conclusions and recommendations The national and regional wildfire policies in Europe should consider all aspects of wildfire management as a whole and with a long-term view. Preventive measures should also be given more emphasis in fire policies as the only way to avoid emergency situations and costly suppression actions. which is the social acceptance and regularization of traditional fire use. secondly. although prevention is slowly gaining importance within national wildfire legislation . would contribute to effective fuel management strategies and to sustainable fuels load control. In addition. The success of the fire risk reduction actions adopted depends. for example. on this preliminary identification of causes. this technique. framed within their territorial contexts. where different physical and human factors are interacting at different socio-spatial and temporal scales. In fact. especially in decentralized countries. interests and concerns (Morguera and Cirelli 2009). policy and technical challenge. considerations must be given to the emergence of new fire prone areas in Eastern. considering the complexity of the fire problem in Europe. and to development of new decision support tools to evaluate new fire scenarios. All stakeholders in the different fire scenarios should have the opportunity to participate in the different stages of fire management. as well as the use of suppression fire. Recognizing the benefits of some traditional fire uses and accepting them instead of systematically suppressing them.g. Therefore.3. An integrated approach through the connection of the wildfires policies and the territorial policies (e. the starting point to set proactive and effective wildfire management policies is the identification of the factors driving fire ignitions. by reducing the need for some silviculture treatments and in fighting fires caused by rural activities. learning and lessons-drawing. at least in the defined territorial scenarios 1 and 2. multi-sectoral coordination and the multi-level implementation of wildfire management programs. first. to a large extent.An Overview of Policies and Practices Related to Fire Ignitions at the European Union Level 43 improve. Governments should define and implement the corresponding legal frameworks and platforms to make public participation in fire management possible and effective. participatory mechanisms. energy policy) is needed to address the structural causes of wildfire events. 2009).

9%). resulting in 55 100 hectares of new interface areas. Asturias (8.44 Towards Integrated Fire Management Box 1. Based on the Corine Land Cover data. where human presence and its related activities near forest fuels could ignite fires. Herrero Doctoral Thesis.5%) and Aragón (0. in 2000. WUI areas spread over more than 2% of the national surface and account for around 1 100 000 hectares.3%) and Madrid (6. . Spatial data on WUI extent. awareness and information campaigns to WUI residents. The general measures considered towards the prevention of ignitions and possible propagation of fires from urbanized areas to the forest masses include: fuel treatments around the structures. The WUI zone is considered in the Spanish forest and fire regulations. but its consideration in planning documents is still very scarce. location and their evolution over time provide key information to develop effective fire planning in order to both avoid fire initiation and prevent negative impacts for the population. Galicia (9. Distribution of the WUI in Spain.8 %. Canarias (7. In Spain. and the regulation of potential fire risk activities in urban areas. Figure 6.4%) have the highest proportion of WUI areas. there has been a growing concern about the WUI in the context of wildfire management due to the rapid expansion of houses near or inside forest areas and several catastrophic fire events. WUI spread in Madrid from 1987 to 2000. Source: G. Wildland urban interface (WUI) as new fire prone areas – the Spanish study case Fire risk is higher in the WUI zone. in progress. while the proportion is lowest in Castilla-La Mancha (0.1%). during the period 1987–2000. Figure 5. the WUI surface in Spain grew 6. During the last decade.6%).

new governance mechanisms. and must therefore be made available for integrated wildfire management. C. Prévenir les feux de forêt. 21–24 June 2009. Finally. 9–18. prevention strategies should evolve according to spatial. existing EU funds provide basic financial support for national. town councils. . and they should also adapt to different socio-economic and territorial contexts. Legal aspects should also be clarified and enhanced through the development of incentives and obligations concerning preventive actions.). Mediterranean. International Symposium. particularly in Mediterranean countries. Pre-conference Proceedings. Furthermore. C. G. subregional and regional prevention measures. etc. etc.).An Overview of Policies and Practices Related to Fire Ignitions at the European Union Level 45 and policies. Pays méditerranéens. deserve special attention and specific preventive measures should be considered as well. Politics and Ethics in Forestry. but carefully adapted to the different contexts and according to the existing territorial patterns (rural abandoned areas. Courrier de la Planète 88: 19. 2009a. S. fire risk analysis. On the other hand.) and concerns different stakeholders (forest land owners. New territories at risk. productive rural regions. or arson related to grazing and agricultural activities. In fact. (e) basic regulation on traditional and prescribed burnings. socioeconomic and natural changes. public awareness and education campaigns. prescribed burning could be an alternative or complementary technique for fuel management. Pp. 2009b. References Aguilar. Basic prevention regulations should address the following aspects: (a) forest management to avoid fires. (f) social prevention measures (public awareness. One of the reasons for this unbalanced situation is that preventive actions are a constant task over the year that does not have the same visibility as extinction investments and that receives less public recognition. and Montiel. and European level is also needed to face the new fire scenarios. (b) preventive silviculture measures in defensive and preventive infrastructures. More coordination at the national. The regulation of fire use practices for different land management purposes should be promoted through the national and regional legal systems. wildfire prevention covers various fields of action (forest fuel management. it is still undervalued when compared to fire extinction.). To do that. S. including non-EU states. New governance strategies for wildland fire management in Mediterranean countries. conciliation of interests. detection. Nancy. In: Buttoud. etc. in order to prevent and reduce fire ignitions as a consequence of negligence.): Change in Governance as Collective Learning Process: Management. should develop participatory and community-based tools for integrated fire management. considering influencing factors like spatial planning issues. wildland urban interface. etc. (d) establishment of risk periods in each country. adapted to the diverse regional and local conditions. governance mechanisms. based on a dialogue between the different stakeholders and the public authorities. (c) risk zoning to delimit areas regarding their fire risk and regulated land uses and activities according to fixed levels of risk. Aguilar. (ed. and Montiel. France. building enterprises. such as wildland urban interfaces.

Investigaciones Geográficas 31: 121–137. European Forest Institute. and Carnus. and Cirelli. 2009. A. Herrero. L. Vélez. Las interfaces urbano-forestales como nuevos territorios de riesgo: análisis y caracterización regional en España. Lázaro. 2003. Report of meeting on public policies affecting forest fires. S.. M.M. European Forest Institute. EFI Discussion Paper 15. Pp. Boletín de la AGE 47: 5–29. Doctorate Thesis.T. Análisis. Landscape Research Journal (under review). Project no. FP6-018505. G. Forest fires and the law. . in progress. E. L. Galiana. 28–30 October 1998. Prescribed burning and suppression fire programmes in selected case studies regions. Departament of Geography. 2010. 2008. Aguilar. G. Rome. Y. I. In: Montiel. La problemática del monte mediterráneo. Part 1. Vélez. 2004. 35–42 Vélez. 14th session. Madrid. Deliverable D7.. Birot. Rome. European Commission. Mª P. A. EFI Research Report 24. In: Birot.46 Towards Integrated Fire Management Birot. and Martínez. J. 21–25. Living with wildfires: what science can tell us. J.3-1of the Integrated Project “Fire Paradox”. FAO 1999.. and Montiel. In: Arbez. Y. C.. Y. 81 p. EFI Proceedings 42. R. Italy. European Forest Institute. J. 82 p. Vilar del Hoyo. R. Empleo de técnicas de regresión logística para la obtención de modelos de riesgo humano de incendio forestal a escala regional. Doctoral Thesis. and Sesbou.): Best Practices of fire use. EFI Discussion Paper 15. 2009. in European member states and in North African countries. Vélez. and Solana. Galiana. Prescribed burning for grazing improvement and fire social prevention: the Spanish EPRIF Programme. Morguera. (eds). Pp. J. (ed. R. 2009. Living with wildfires: what science can tell us. Martínez. (ed. M. Herrero. FAO Legislative Study 99. 2009. Martín. 2002. University of Alcalá. C. A guide for national drafters based on the Fire Management Voluntary Guidelines. European Forest Institute. A wildland-urban interface typology for forest fire risk management in Mediterranean areas.). L. Risk management and sustainable forestry. Causes of forest fires in the Mediterranean Basin. estimación y cartografía del riesgo humano en incendios forestales. Assessment document on the main strengths and weaknesses of the legislation and spatial policy instruments concerning wildfire integrated management in the EU.. R. 2009. In press. The causing factors: a focus on economic and social driving forces.

3. Wildfire Propagation: Knowledge and Search for Solutions .

.

Silva1 and Sandy P. in order to understand the mechanisms responsible for fire regimes it is important to look at historical data. Climate and Land Cover as Controls on European Fire Regimes Joaquim S. Gill proposed a description of a fire regime as including: the time between fires. ‘fire regime’ is used in the latter sense. It is not a straightforward exercise to assess whether one factor or another factor is the most important. 2009). and by actively suppressing fires.1 Humans. Baeta Neves”. including the traditional use of fire for agriculture and slash and burn practices. Both climate and fuel characteristics influence fire regimes. and the season of occurrence (Gill (1975). Agee 1997). without particular regard to the precise set of fire characteristics to be included. Humans influence fire regimes in three basic ways: by starting new ignitions.1 The driving factors of fire regimes The expression ‘fire regime’ has become a rather loosely used term since it was first introduced by A. and there has been considerable debate about which is the most important driving factor influencing the fire regime of a specific region (Minnich et al. since that depends on the type of fuels and climate (Bessie and Johnson 1995. University of Bristol. 2009). Technical University of Lisbon. Gill more than three decades ago. the type of fuel burned. this raises another unsettled issue of whether humans or natural factors are more important in determining fire patterns (Bowman et al. Carrion 2002. the intensity of the fire.g.1. Within the present text. Since then. 2008) has compiled nearly 800 such records from around the world. by changing fuel characteristics. Harrison2 1 Centre of Applied Ecology “Prof. Keeley and Fotheringham 2001. but more commonly the expression has been used to summarize the characteristics of the fires that typically occur at a site (Whelan 1995). and other authors have proposed modifications to the original definition (e. Fire regimes show a strong .g. Given its long history of human occupation. In many regions this assessment is even less straightforward because of the influence of humans (Pausas and Keeley 2009). Charcoal preserved in lake or bog sedimentary sequences provides a long-term record of changing fire regimes at a catchment scale. Andric and Willis 2003. 2005).M. Portugal 2 School of Geographical Sciences. Therefore. 1993. these aspects have assumed a particular importance in Europe as reported by different local studies (e. the time since last fire. Bond and Keeley 2005). United Kingdom 3. 2004. Some authors have used the expression to describe a particular fire. the expression has been widely used. The Global Palaeofire Working Group (Power et al.3. Archibald et al. and analyzed these to examine changes in fire regimes on decadal to millennial timescales. Gillett et al. Tinner et al.

Portugal. especially grazing pattern and land abandonment. and in particular the relationship between fire occurrence and land cover.1. 2009). In regions that have not experienced a recent expansion of large-scale agriculture and ranching. (2008) argue that the decrease of fire activity was an accidental consequence of human activities and resulted mainly from landscape fragmentation and/or the reduction of fuel loads in intensively managed landscapes. in press). 3. both at global and regional scales (Marlon et al. but corresponds with the expansion of large-scale agriculture and ranching (Bowman et al. (2008) also examined the potential role of humans on recent fire regimes.2 Factors determining the fire regime in different European regions In many areas of Europe. In this chapter we discuss the factors determining the fire regime in Europe. 2008). Macedonia (Greece). and Poland (Figure 1). Temperature changes may also be important through influencing the rate of curing of fuels. During the last glacial (ca 80 000–20 000 years ago). such as the northern hemisphere boreal zone or Europe. Because of the increased number of sites available. pre-dates the introduction of active fire fighting. though expressed less strongly. Temperature changes influence fire regimes primarily through vegetation productivity and hence fuel availability. In addition. Mouillot and Field 2005) and is incorporated in the representative concentration pathway (RCP) scenarios developed for the next International Panel on Climate Change assessment. this has resulted in asynchronous changes in fire regimes between the two hemispheres. Thus. The timing of the downturn in fire in several regions. and have argued that the major role of humans in global fire regimes during the past 2000 years has been to suppress fires during the twentieth century. most obviously in the Americas.50 Towards Integrated Fire Management response to temperature changes on centennial to millennial time scales. fire increased during the repeated abrupt warming events registered in the Greenland ice core and decreased during the following cool periods (Daniau et al. Ticino (Switzerland). Australia and Asia. This trend is also seen during the last 21 000 years (Power et al. Fire proneness assessment of land cover types has received recent attention from . the last decades have been characterized by drastic changes in land cover. The downturn in fire activity in Europe during the first part of the twentieth century has been seen.g. it has been possible to show that there have been opposing trends in temperature changes between the northern and southern hemisphere (with warming in the southern hemisphere corresponding to cooling in the northern hemisphere and vice versa). Records for the past 2000 years also show strong correspondence between temperature changes and fire. Marlon et al. by presenting recent studies developed in different European regions: island of Sardinia (Italy). Marlon et al. and particularly in the Mediterranean Region. in inventory data (see e. there is no downturn in fire during the last century. in order to understand which land cover types are ‘preferred’ by wildfires. and thus by reduced lightning and thus reduced ignitions. cold climates are characterized by a reduction in the vigour of the hydrological cycle. This recent development emphasized the need of addressing the problem of fire proneness. combined with reforestation. 2008).

The analysis of fire proneness within a certain region depends on the method used and on the available data sets.Humans. Adapted from San-Miguel and Camia (2009). Climate and Land Cover as Controls on European Fire Regimes 51 Figure 1. Besides the basic aspect of fire proneness. although it has to be taken into account that in the European context. but also to the approach chosen by each research team. are essentially climate and fuel-driven. researchers (e. The methodologies varied according to the data available. it is very important to have a sufficient number of geo-referenced fire perimeters in order to allocate the burned areas to the different land cover types. Nunes et al. whereas the mechanisms responsible for the distribution of burned areas across the landscape. 2009). 2005. the mechanisms influencing the start of a fire are strongly correlated with human activities leading to ignitions (Catry et al. On the other hand. Map of burned areas in the European Union by region (NUTS3) and location of the study regions. This is not always possible because not all countries and regions have gathered this information over the years. Moreira et al. Recent studies were developed aiming at assessing the influence of these factors on fire regime. . In this context we must have in mind that typically. data sets are short and may not necessarily capture a comprehensive range of land use types and patterns. The use of ignition points instead of burned areas may be a way to overcome this difficulty. Cumming 2001. 2009).g.

Wildfires occurred earlier than expected from a random null model. (2000) and to what extent these regions differed in the temporal properties of their wildfire time series. wildfires occurred later than expected. and shrublands. even in regions of high human pressure. . Therefore the study highlighted a close relationship between the timing of fire occurrence and land cover. the aim of the study was to analyze the temporal patterns of fire occurrence in Sardinia during the period 2000–2006. and Transitional Temperate) following the classification by Blasi et al. Fire frequency was higher than expected by chance alone in urban and agricultural areas whereas in forests. Bajocco et al. Sardinia (Italy) On the basis of the available fire history data for Sardinia for the period 2000– 2004. the earliest mean time of wildfire occurrence was usually related to the Mediterranean phytoclimatic region. but keeping both variables separate. grasslands. (2008) have evaluated the remotely-sensed phenological uniqueness of the Sardinian phytoclimatic regions (Mediterranean. The authors concluded that there was a strong climatic control on the temporal characteristics of wildfires. while the latest mean time of wildifre occurrence was associated with the Transitional Temperate region. forests and arable land did not appear to be particularly fire prone. governed by two complementary factors: climatic factors that determine the spatial distribution of land cover classes according to climatic regions. which was also the region that experienced the highest number of fires. Transitional Mediterranean. there was a significant resistance to fire spread. Using the same wildfire data set. In grasslands and shrublands mean fire size was significantly larger than expected from a random null model whereas in urban areas. in terms of both fire frequency and fire size. The results also emphasized a gradient in the mean starting Julian day of fires. The results obtained provided a clear and coherent segregation of the major phytoclimatic units in terms of phenological dynamics and temporal characteristics of wildfire regimes. aiming at assessing the seasonality of wildfires for each land cover class. Bajocco and Ricotta (2007) analyzed the fire proneness of given land cover classes both in terms of fire frequency and mean fire size. More specifically. The results obtained from the analysis of 13 377 fires showed that for most land cover classes. aimed at characterizing the fire regime according to different types of land cover. and the human population that directly influences fire ignition. grasslands. in terms of the mean fire size. fire behaved selectively. in order to identify land cover classes where wildfires occur earlier or later than expected from a random null model. with very high significance levels. permanent crops. (2009) was performed. and shrublands. fire frequency was lower than expected. in urban areas and in all agricultural classes. For instance. By contrast.52 Towards Integrated Fire Management other aspects like fire seasonality and fire size were addressed by the different studies. There was also a strong association between land cover classes and the Climatic Regions of Sardinia. and heterogeneous agricultural areas. in forests. The results showed that the timing of fire ignition was selective for all land cover classes analyzed. A complementary study by Bajocco et al. Additionally. since both land use classes burned in proportion to their availability.

fire fighting strategies. 1993. resource selection ratios were used to measure fire proneness for different land cover classes in 12 regions of the country. whereas annual crops. Ostryo-Carpinion. Moreira et al. fire-resistant. In terms of forest types. management. The resource selection ratio (Manly et al. allowed the identification of three main geographical areas with similar fire selection patterns. population density and the relative abundance of different land cover classes. Especially. human population density is higher in lowlands and in warmer. Fagion-Abietion cephalonicae. Silva et al. On the one hand fire frequency decreased in higher elevations where climate becomes cool and moist. Climate and Land Cover as Controls on European Fire Regimes 53 Macedonia (Greece) Bajocco et al. predominantly agricultural land cover classes) to less densely populated landscapes (flammable semi-natural and forest land cover classes). conifer plantations were more fire-prone than eucalyptus plantations. On the other hand. This ranking was further analyzed across vegetation zones. but also a gradual shift from intensively managed landscapes (relatively abundant. (2009) restricted the assessment of fire proneness to the main forest types in Portugal classified according . indicating that fires have a lower likelihood of starting in intensively managed agricultural landscapes. There were regional variations in susceptibility of land cover classes to fire. This zone was characterized by a gradual shift from warm and dry conditions to cool and humid conditions. drier zones where agriculture is the dominant land cover. 2001) was used to derive a ranking of land cover classes in terms of fire proneness. (2009). A cluster analysis of regional variations in selection ratios for all land cover classes. The study also showed a significant positive correlation between land cover availability and the resource selection ratio. Shrublands were the most fire-prone land cover. Pinetalia nigrae. which were discussed on the basis of differences in climate. Based on land cover information within 5591 burned patches (larger than 5 ha) and in the surrounding landscape. Portugal Patterns of wildfire occurrence at the landscape level were characterized by Moreira et al. semi-natural land cover classes were so scarce that. Quercion confertae. in the Ostryo-Carpinion vegetation zone. (2009) during the period 1990–1994 in Portugal. and the regional availability of each land cover class. Based on the study by Moreira et al. where fuel load is kept at low levels. permanent crops and agro-forestry systems were the least fire-prone. and broadleaved forests were the least fireprone. overall. It appeared that fire frequency was conditioned by two opposing trends. ignition patterns. Vaccinio Picetalia) across the elevation gradient. The results were discussed in relation to the moisture and temperature regime of vegetation zones (Quercion ilicis. Here climate conditions were more conducive to fire ignition but the landscape was characterized by land cover classes which are generally less fire prone. (2008) assessed fire proneness by using fire ignition locations across the region of Macedonia in northern Greece over a period of 13 years (1985–1997). this zone had fewer fires than expected by its area.Humans. The authors mention the existence of a dynamic ‘tension zone’ corresponding to a higher fire incidence.

stand composition was the most important for explaining fire probability.0 ha and the average burned area in some coniferous forests tended to be high. and the proportion of burned National Forest Inventory plots.54 Towards Integrated Fire Management to the main species.5% of surface area). and natural summer fires) and performed 1000 random Monte Carlo simulations on frequency and size.8% occurred . chestnut (Castanea sativa) forests. half of the fires in chestnut stands were significantly larger than 1. With this purpose. by Bajocco et al. Using the Corine land cover classification (Büttner et al.2% occurred in complex cultivation patterns (8. anthropogenic summer fires. unspecified broadleaf forests. the fire size tended to be small. using three different approaches: the use of resource selection ratios applied to burned patches. Anthropogenic winter and summer fires occurred mostly at low elevation in chestnut (Castanea sativa) stands. (2009) to analyze the relationship between fire frequency and land cover. by using univariate logistic models.4% of burned area). Stepwise regression procedures were used to build a multivariate logistic model which showed that stand structure plays a different role for different types of forest.3% of burned area).6% of burned area). the proportion of randomly located plots that were burned. 2004). In winter. cork oak (Quercus suber) forests. unspecified conifer forests.9% of fires occurred in non-irrigated arable land (37. Ticino (Switzerland) Fire proneness of given forest vegetation classes was analyzed both in terms of fire frequency and fire size (mean and median) for the period 1982–2005 in Canton Ticino (southern Switzerland). The results allowed ranking fire proneness according to the following decreasing order: maritime pine (Pinus pinaster) forests. (2009) also used a different data set (1998–2005 fire maps) to explore the influence of structural variables plus stand composition on fire probability. 11. 16. eucalyptus (Eucalyptus globulus) forests. Poland A database of the fires that occurred between 1996–2006 in Poland was used by Szczygieł et al. and to anthropogenic characteristics such as the closeness of roads or buildings.2% occurred in continuous urban fabric (8. The obtained fire occurrence pattern and especially the occurrence of anthropogenic fires in terms of fire frequency. in terms of fire probability (see text box). 9. From all variables. 8.5% occurred in land principally occupied by agriculture (17. Silva et al. anthropogenic winter fires. the authors investigated the data set in four fire groupings (all fires. holm oak (Quercus rotundifolia) forests and stone pine (Pinus pinea) forests. (2008). seemed to be also related with geographical parameters such as altitude and aspect. Lightning fires seemed to prefer spruce (Picea abies) stands and to avoid the summer-humid chestnut and beech (Fagus sylvatica) stands and the 50 to 100 m buffer area. the study concluded that 31. In beech forests. broadleaved forests and in the first 50 m from the forest edge. in mixed forests and in the spruce stands hit by natural fire in summer. This study revealed that the vast majority of fires occurred in non-forest lands.

8% of burned area). except for unspecified broadleaf stands (a decreasing trend) and for eucalyptus (a mixed trend). Case study of fire proneness: Portugal Fire proneness was studied in Portugal by using 1998–2005 fire maps. for five types of forest according to a logistic regression model.3% of burned area) and 3. In terms of dominant stand species. (2009). Climate and Land Cover as Controls on European Fire Regimes 55 Box 1. Adapted from Silva et al. including a total of 30 494 fires. whereas it is lower in holm oak and cork oak stands. Figure 2.6% occurred in mixed forests (2.6% occurred in broadleaved forests (8.9% of burned area). in order to relate forest characteristics (composition and vegetation density) with fire occurrence. With this data it was possible to build a fire probability model using logistic regression procedures (Figure 2). the dominant species and the stand age classes. The authors refined the results for the Polish State Forests. using the resource selection ratio.Humans. Probability of fire occurrence during years 1998–2005 as a function of a cover index. The analysis covered the relationship between the number of fires and the forest site types. The analysis confirmed the existence of different levels of fire proneness according to site and stand conditions. using data from 2002–2008. the highest . 6. maritime pine and unspecified broadleaf stands. in coniferous forests (5. starting from the driest sites to the moister ones. The lowest values (less than 1) corresponded to broadleaf forests and to mountain forests. The sites with higher fire proneness were conifer forests. Results suggest that fire probability is higher in eucalyptus. These maps were overlapped with the locations of the 1997–1998 Forest Inventory plots. Moreover it seems that fire probability increases for higher vegetation density (given by a Cover Index) for all species.

in terms of age classes.1. The individual case studies indicate that land cover. which showed higher values than class IV (60–80-years-old). corresponding to stands older than 80-years-old. However. if societies are really committed to minimize the problems caused by wildfires. is important in determining the availability of fuel and hence fire regimes. Societies should reflect on the models of land-use transformation. Finally. as well as a common data set on land cover occupation.) stands. taking into consideration the high costs of fire suppression and fire prevention. many of these fires are small. with resource selection ratio values greater than 1 only for stands younger than 40-years-old. 1990. which is partially influenced by climate but also a consequence of human management. The results on fire proneness at local scales clearly emphasize the importance of adequate land cover planning and proper fuel management for wildfire prevention. Vélez 2009). Although the importance of controlling human-caused ignitions should not be underestimated. the results showed that fire proneness generally decreased for older stands. Humans can play an important role in deliberately or inadvertently modifying the land cover and therefore the fuel characteristics of a given region. namely that which concerns the generalized abandonment of agricultural fields and the establishment of forest plantations (Nunes et al. fire managers need information concerning the spatial and temporal . In this latter case. new solutions for landuse planning and management should be envisaged. The exception to this trend was class V. 3.56 Towards Integrated Fire Management value was obtained by Scots pine (Pinus sylvestris) stands (more than 1). which have been followed in the last century. Individual case studies show that the number of fires is strongly influenced by human activities. Specifically. as seen during the extreme weather conditions occurred in 2003 in Europe (Trigo et al. The difficulties reported in some of the above described research points to a clear need for European countries to use common criteria for wildfire statistics.3 Conclusions Despite the wide range of methodologies used in the reported studies. climate factors and vegetation type appear to be more important drivers of large fires. prescribed burning as a fuel management technique has been suggested to be an economically and ecologically advisable solution (De Ronde et al. as measured by area burned. being higher in agricultural areas and wildland urban interfaces. The palaeo-record of biomass burning essentially confirms these conclusions. the lowest values being obtained by broadleaf species and fir (Abies spp. All other stands had values less than 1. some coherent conclusions emerge that have practical consequences for management policies. and the differences in spatial and temporal scales. Therefore. showing a strong response of burned area from decadal to millennial-scale climate changes over most of the last 21 000 years with significant increases in fire during warmer intervals. of fires. There is a fundamental distinction to be drawn between number of fire starts and the impact. Fernandes and Rigolot 2007). to improve fire prevention. 2005. 2006). These analyses suggest that policies on land-use planning should be driven to minimize fire hazard.

.. P. L. The severe weather wildfire – Too hot to handle? Northwest Science 71:153–156. Trends in Ecology & Evolution 20:387–394. The Corine Land Cover 2000 project. and others 2009. 2005.. 2007. What limits fire? An examination of drivers of burnt area in Southern Africa. Publication describing wildfire regimes distribution in selected European study areas..J. FP6-018505.. Project no. M.B. P. J.S. G. Evidence of selective burning in Sardinia (Italy): which land-cover classes do wildfires prefer? Landscape Ecology.M. J. Fire in the Earth System.D. would be the right agency to support and lead this task. Project no.P. The phytogeographical regions of Slovenia: a consequence of natural environmental variation or prehistoric human activity? Journal of Ecology 91:807–821. S.. Mari. S. Climate and Land Cover as Controls on European Fire Regimes 57 distribution of fires across the landscape. Balch. D.. M. Gitas. G.. J. G. L. W. S. S.B.J. Van Wilgen. and Soukup.. 2000. Ecology 76:747–762. G. Deliverable D4... Ecosystem classification and mapping: a proposal for Italian landscapes. Archibald. and Johnson.. Frondoni.. Artaxo. L. Jaffrain. B. C. P. F. Internal Report IR4... On the other hand.P.. ..J. 2009. Bajocco. and Rosati. Ricotta. M. F.A. and Ricotta.A. European Commission. Andric. Pp. R.W. M. D. Bond. D.. Ricotta. fuel and climate as controls on fire regimes. J. and Keeley. Bowman.M. Bajocco. S. Bond.. 557–567. Ferreira. W.. Global Change Biology 2008:1–18. Büttner. 1997. thus rendering fire fighting strategies more effective.. European Commission. 2009. D'Antonio.. Pezzatti. G. ecologically meaningful landscape classifications may provide geographical units with great potential for the development of strategies for fire risk assessment and fire prevention. J.J. This would provide valuable information which can be used to better understand the different role of humans. At the European level. FP6018505. 33 p. Vaz. Rosati. and Scholes. Torriani.. C.. Moreira.C. Feranec. 2003.L. C. Cochrane. Quantitative evaluation of fire proneness of different land use classes. Roy. R. the European Commission Joint Research Centre. References Agee.K. Angelis.Humans. E. K. Blasi.Z. Science 324:481–484. Catry. which manages the European Forest Fire Information System. 2004.. R. W.E. Doyle.S. Mouflis.A. ecologically suitable land cover maps may be helpful for the development of strategies for fire hazard assessment – for example. T. J.. DeFries. C. S. Therefore.D. 2008.G. it is of paramount importance that individual fire reports include both a map of the burned area and the corresponding ignition point location.. The relative importance of fuels and weather on fire behaviour in Subalpine forests. and Conedera. it also influences the seasonality of fires through its strong association with the bioclimatic characteristics of the landscape. A. and Willis. While land cover directly controls what is actually burned by fires in terms of fuel load and fuel continuity. Harrison. coherent and comprehensive fire database..1-1b of the Integrated Project “Fire Paradox”. Carranza. Bajocco. C.. Carlson. Maucha. In this framework. Pezzatti. G. Bessie.K.C. Fire as a global ‘herbivore’: the ecology and evolution of flammable ecosystems. Silva. 79 p. 1995.. J. EARSeL eProc 3:331–346..1-2a of the Integrated Project “Fire Paradox”. Applied Vegetation Science 3:233–242. aiming at achieving a sound. for optimizing the location of water points and fire brigades. I. and Mazzoleni.

1975. R. Marlon. F. Rego. Minnich.C...H.). and Chou.2-3 of the Integrated Project “Fire Paradox”. London.. A burning Sstory: the role of fire in the history of life. M. and Rego.V. P. J. and Keeley.C. and Moreira. Modeling and mapping wildfire ignition risk in Portugal. Resource selection by animals: statistical design and analysis for field studies. and Rigolot. Figures and Trends in the EU. FP6-018505. Deliverable D4.F. B. and others 2008.E. N.. J. M. A.C. J..-L.M. Gillett.. EFI Discussion Paper 15. In: Birot. and Bartlein.M. and Thomas. Carcaillet.. Moreira.J. European Forest Institute. 2001.J. Fire history and the global carbon budget: a 1º ×1º fire history reconstruction for the 20th century. M. Prescribed fire in industrial pine plantations. F.). A.. Fire and the Australian flora: a review. Alldredge. 2009. R. Changes in fire regimes since the Last Glacial Maximum: an assessment based on a global synthesis and analysis of charcoal data. Lightning detection rates and wildland fire in the mountains of Northern Baja-California.. Plant Biosystems 143:597–608. San-Miguel.J. B.M. Piwnicki. Living with Wildfires: what science can tell us. Bradshaw.. Vaz. Weaver. and Camia. Moreira.. Temporal (1958–1995) pattern of change in a cultural landscape of northwestern Portugal: implications for fire occurrence.. Catry. Gavin. Marlon. .J. 2008. C. S. F. Fernandes.W.. Cordova. S. P. Conedera. Fire ecology north and south of the Alps since the last ice age. 2005. History and management of crown-fire ecosystems: a summary and response.. S. Patterns and processes of Late Quaternary environmental change in a montane region of southwestern Europe. Pereira. F.. A. Bartlein. In press.. Dasgupta. F. Bartlein. C.. 2007. De Ronde.G. Moreira.E.. Vasconcelos. P.. Higuera. M. 91 p. J. 2009... E. Fire flammability of forest plant formations in Poland. BioScience 59:593–601. F. European Commission. R. 1993. Zwiers. Harrison. Ortiz.F. Power.. P. Y. 2009. Y. D.G. 2009.J. 11–18. S.. Keeley.J. Conservation Biology 15:1561–1567.. (ed. M..J. Holocene 15:1214–1226. Project no. C.L.J. A. Mouillot. Landscape Ecology 16:557–567. Pp.S. Global Change Biology 11:398–420. Vizcaino.L..P. R.D. A. F. F. Catry.J. F. 2001. 2004. Bação. 2005. Gill.. Goldammer. 2009. A. Forest Ecology and Management 241:1–13. I.. Szczygieł. N.D. L. Tinner. Quaternary Science Reviews. Carcaillet.. F. F. and Field.58 Towards Integrated Fire Management Carrion. Pausas. Ecological Studies Analysis and Synthesis 84:216–272. Forest type and wildfire in the Alberta boreal mixedwood: what do fires burn? Ecological Applications 11:97–110. Cumming. 1990. International Journal of Wildland Fire 18:563–574. and Silva.G. M.B. P. Nature Geoscience 1:697–702.C. Vaz. Harrison.. F. Ballouche..X. Joos. McDonald. Land cover type and fire in Portugal: do fires burn land cover selectively? Landscape Ecology 20:661–673. and Soares.. F. Atmosfera 6:235–253.S. E.L. Australian Forestry 38:4–25. Climate and human influences on global biomass burning over the past two millennia.. Quaternary Science Reviews 21:2047–2066. Mexico. J. and Ferreira. Catry. and Godinho-Ferreira. C..P. D.. and Kwiatkowski.. and Prentice. Manly.. Power. and Lotter. Silva.G. International Journal of Wildland Fire 18:921–931. 2005. Climate Dynamics 30:887–907.S. 1993. P. Mayle. J. 2002. P. J. J. R. Sosaramirez. Nunes. 2009. N. M. D. W.S. Fire regimes during the last glacial. 2001. Geophysical Research Letters 31:1–4. Regional variations in wildfire preference for land cover types in Portugal: implications for landscape management to minimize fire hazard. P.A. C. J. Harrison. Ubysz. The fire ecology and management of maritime pine (Pinus pinaster Ait. Detecting the effect of climate change on Canadian forest fires.. and Fotheringham. F.C.C. Forest Fires at a Glance: Facts. Rego. J. P.. and Flannigan. J.P. B.. J. F. Assessing the relative fire proneness of different fores types in Portugal. J. Chapman and Hall.P. C.R. Ammann. Wade. Daniau.

C. Calado. R. Atmospheric conditions associated with the exceptional fire season of 2003 in Portugal. Pereira.J.. Climate and Land Cover as Controls on European Fire Regimes 59 Trigo. Cambridge. 2009. Pp. R.). and Santo. M. International Journal of Climatology 26:1741–1757. Y.M. The causing factors: a focus on economic and social driving forces. EFI Discussion Paper 15.M. European Forest Institute. 2006. Dacamara. Mota. J. R. Vélez. Living with wildfires: what science can tell us.. Pereira. Whelan. Cambridge University Press. In: Birot.G. F. T.. The ecology of fire. B.. (ed. 21–25. C.C. ..E.J.Humans. 1995.

.

that a full representation of wildfire physics has been able to give predictions of fire spread. but usually on super-computers. Avignon. Canada. and scientists have recently developed similar models in Spain and Portugal.2 Fire Modelling and Simulation Tools Jean-Luc Dupuy1 and Daniel Alexandrian2 National Institute for Agricultural Research. and the United States. which limit their applicability is the accessibility of the large amounts of vegetation and atmospheric data that they require. because the results of the simulation have to be obtained in almost or faster than real time. Empirical fire behaviour models are established on the basis of a reasonable number of fire observations and predict the rate of spread or flame size of a fire (Sullivan 2009b). and slope). parametric laws fitted to the output of 3D physical models (see Box 1). Currently. France 2 Agence MTDA. fuels and weather. fire simulators can only use simplified models (empirical or semi-physical). However. Modelling approaches can be reduced to two: empirical and physically-based. GIS-based fire simulators can automatically compute fire growth under heterogeneous conditions of terrain. Two-dimensional (2D) fire simulators have been developed in parallel to provide decision-makers with powerful tools capable of predicting and mapping fire behaviour in real conditions. The so-called full physical models enable three-dimensional (3D) simulations of fire spread at the scale of a forest stand (<20 ha).1 Introduction The scientific observation of fires in natural conditions is a difficult task. fuel moisture. however. A way – and one of the objectives of the Fire Paradox project – to fill the gap between 3D physical models and 2D fire simulators is to use. Despite the fast development of computer resources. and predict many fire characteristics. fuel characteristics. There are well-known empirical models developed in the Australia.2. fire modelling has been a way to reduce the amount of observations necessary for understanding and predicting fire behaviour. it is not practical – so far and in the near future – to use this new generation of physically-based models to simulate fire at larger scales (10 km2) with resolutions on the order of meters or smaller. Aix-en-Provence. As for many environmental phenomena. These models solve the transport equations of physics over time on a spatial grid. Usually.3. The other aspects of these types of model. . Physically-based models are based on the principles of combustion and they attempt to quantify the basic fire mechanisms (Sullivan 2009a). as the “engine” of a fire simulator. France 1 3. simple equations give the fire rate of fire spread as a function of a small number of parameters (wind speed. It has only been in the last decade.

The gaseous fuels react with oxygen. this is the flaming combustion process releasing heat. In the particular case of woody fuels heat transferred to the vegetation makes the temperature of the fuel rise. Fire ignition. it is able to use various propagation models. • it includes a wind simulator to assess the local variations of wind direction and speed on the terrain. The rate of fire spread – i. The physical model used to establish parametric laws is Firetec. Basically. fine fuels such as grasses. and on the energy required to raise the temperature of the fuel to the ignition temperature. which is the initiation step of fuel combustion. Vesta Vesta is the Large Scale Wildfire Simulator developed in the frame of the Fire Paradox project. can occur only when a heat source. Beyond some threshold ‘ignition’ temperature (about 300°C). • it enables launching a series of simulations on a given area to calculate hazard maps. • it allows the user to simulate in an interactive way some human intervention as fuel breaks. the vegetation releases flammable gaseous fuel at a high rate: this is the pyrolysis process. • it may be run with or without spotting models (namely the probabilistic model developed in the frame of the Saltus project). fire spread is possible after fire ignition and when the amount of heat transferred from the preexisting fire to neighbouring fuel is high enough to cause its ignition. The basic idea of the software is to fill the gap between time consuming 3D physical models and existing semi-empirical 2D fire simulators.2.e. Besides the basic usual functions of a 2D-Fire simulator. etc. • as a ‘platform’. • it enables the user to compare fire simulated with real wildfires to validate the results obtained. twigs. heat transfer and combustion (Albini 1992). which increases their vulnerability . • maps and fires can be visualized in 2D and 3D. heat transfer and combustion The ignition and spread of fire are governed by the physico-chemical processes of pyrolysis. thanks to an agreement with the LANL (Los Alamos National Laboratory). fuel and an oxidizer are all present at the same time in the same physical space at high enough temperature to react. the speed of the fire – depends on the strength of the heat source. and foliage are largely responsible for fire spread due to their large surface area per unit volume.62 Towards Integrated Fire Management Box 1. • it allows fuel types description with the best accuracy available.2 Pyrolysis. aerial dropping. on the efficiency of the heat transfer processes. In wildfires. its main features are the following: • it is able to work on various vector and raster GIS file formats (import and export). 3.

combustion and heat transfer are affected by the different environmental factors and the nature of the complex coupling between them. A fire that burns only the surface fuels is called a surface fire and a fire burning also the tree canopy is called a crown fire. Moisture content also influences fire spread and how much of the existing fuel can burn. The dominant heat transfer processes between these fuels are radiation and convection. In natural fires. shrubs) to crown fuels (tree canopy foliage). their relationships. The amount of forest fuel and how it is distributed in the different vegetation layers is also a factor affecting fire spread. fire intensity (also known as fire power) is directly influenced by the amount of fuel available to burn. In addition. Because water content of the fuel must be evaporated before ignition can occur and because the boiling of water requires a huge amount of energy.2. but also the transition of fire from surface fuels (litter. even if only average fire behaviour is predictable. Empirical or physically-based models have also been developed for fire spotting. This is why so many efforts have been dedicated to measure or predict the moisture content of vegetation as influenced by meteorological conditions and biological cycles. Fire spread may be enhanced by the transport of firebrands some distance ahead of the fire. These firebrands may ignite the fuel at their landing point and a secondary fire – called a fire spot – develops. Knowledge of these effects. slope – and fire rate of spread. This is why empirical models succeed operationally in predicting fire spread within the range of .Fire Modelling and Simulation Tools 63 to ignition. This enhances heat transfer to the unburned fuel and it has been observed that fire rate of spread is more or less proportional to wind speed at least for low to moderate wind speeds and for surface fires. Crown fires are the most likely to produce fire brands causing spotting. fuel properties. Fire behaviour research has established significant correlations between the fire environment – wind. Forest fuel moisture and wind are probably the most obvious factors affecting fire spread. The terrain slope also affects fire spread and a steep slope is known to represent very dangerous situations for fire fighters. experiments and modelling. Fire spread is not only the propagation of fire on the terrain slope. 3. Spotting can occur up to kilometers from the main fire front. fuel moisture content is the most important parameter influencing fire start.4 3D-Fire modelling Fire modelling is a powerful technology for the simulation and prediction of fire spread.3 Physical processes and factors influencing fire spread Understanding how woodlands burn amounts to understanding how the basic processes of pyrolysis. and their quantification has been the purpose of many studies based on observations. 3. wind has mainly the effect to bend the flames towards the unburned vegetation and convect heat to its surface. grasses.2.

The natural heterogeneity of vegetation and mechanism of spotting are additional sources of fire behaviour variability. Global change is expected to cause important changes in the environmental factors that influence fire start and spread. provided that they use adequate computation and modelling approaches. the turbulence level. In contrast recent physically-based models are powerful tools to explore new environmental scenarios. Fire is. That means that one must expect great variations in fire behaviour from one moment to another and from one point to another. empirical models will be of small value since they are based on fire observations made in existing vegetation types. This physical situation is very unstable and explains why high-intensity fires are likely to exhibit very erratic and dangerous behaviour. cannot properly render the effects of turbulence. In order to assess the consequences of these changes on fire behaviour. which is significant mainly at local scales (<1 km). is increased by buoyancy effects. One can also expect that some species not yet adapted to heat and dryness will adapt to new climate conditions by reducing their plant moisture content or more likely by producing dead biomass.64 Towards Integrated Fire Management environmental conditions under which they have been developed. and in particular coupled atmosphere-fire models that solve the transport equations of physics to represent the fire. In addition. One can eventually expect that by migration new mixtures of plant species will take the place of species no longer adapted to their environment. The first consequences have been observed with the decline of some conifer forests in the south of Europe: an immediate consequence is the sudden reduction of moisture content of the vegetation. are powerful tools to explore the effects of turbulence (and more generally non-linear effects) on fire behaviour. which measures the amount of local wind fluctuations. The question of fire behaviour predictability however remains important. Other models. A simple physical criterion based on the relative force of wind and buoyancy has been proposed to identify such unstable situations. Fire observations by experts of wind flows and recent advances in coupled fireatmosphere modelling have revealed that wind turbulence plays a dominant role in erratic fire behaviour often observed in the field. Examples of use of full physical models in the field of research are given in Box 2. Hot gases in a fire plume are less dense than ambient air and rise vertically in the atmosphere due to the Archimedes’ principle. changes in vegetation characteristics are expected. Physical models. as some Mediterranean species already do. the driving force is called the buoyancy force and the upward vertical motion of hot gases is accompanied by downward vertical motion of fresh air. It is well known in the field of fluid mechanics that turbulence effects are only predictable on average. In particular. The fact that the fuel is spatially-distributed in these models also enables the variability caused by the non-uniform distribution of vegetation biomass to be accounted for. of course. but the prediction of the criterion itself remains delicate. . These facts mean that some aspects of fire are predictable at some scales and this is essential to decide strategies of fire prevention and mitigation. Tree species are known to have a large capacity for adaptation. by their nature. governed by the laws of physics and as such is not a completely random process. Fire practitioners or fire fighters operating on the ground must be aware of and account for this variability.

since such 3D-fire models have a limited use for operational purposes. • prevention and pre-suppression activities by assessing the effectiveness of fuel treatments. they produce output directly related to the input used by the user). Finally. economic assessment. 3. especially when a number of simulations are requested. moisture content. the 3D physically-based model is used to provide local fire behaviour characteristics like rate of spread or fire intensity. produce output that are GIS compatible.2. can automatically compute wildfire growth and behaviour for long time periods.e.e.g. As a first step. fuel and weather in each map pixel) change over the map. CFFDRS). The model serves to implement a database of fire characteristics for various fuel types. identifying opportunities for the use of prescribed burning. • making decisions by producing hazard maps and risk maps. The two North-American fire simulators Farsite (Finney 1998) and Prometheus – the most used worldwide – have the same functionalities.5 2D-Fire simulators A two-dimensional fire simulator predicts fire growth over a map. • operational purposes by supporting fire suppression activities and predicting wildfire behaviour. As a second step. simulation of aerial and ground fire suppression actions).g. So. line or polygon ignitions. the simulator thus requires the support of a geographic information system (GIS). the operational use of such models will be limited by the computational resources. by making simulations over distances up to several hundreds of meters in the direction of propagation. are based on existing fire behaviour models widely used (e. topography. and even in a long-term perspective. modification of fuel types. and so on.6 Linkage between 3D-fire models and 2D-fire simulators The scale of prediction of physically-based models resolving transport equations on a three-dimensional grid is still far from the landscape scale. Using spatial information along with weather and wind files. local conditions (i. This is the reason why such models cannot be a direct component of a fire simulator at a landscape scale. they: • • • • • can use point. slope. the library of simulated conditions is used to fit parametric laws that yield the rate of spread as a function of wind. . In a real situation. Fire simulators can potentially be used by land management agencies and fire fighting organisations for: • wildfire training and educational sessions by conducting all kinds of ‘what-if’ scenarios. Using deterministic models (i. these laws are applied to local input data to predict fire spread over the terrain map using a contagion process.2. enable some interactivity with the user (e. Behave. one way to overcome this limitation is to incorporate their results as the ‘engine’ of 2D-fire simulators. testing the location of potential access roads.Fire Modelling and Simulation Tools 65 3. topographies and weather conditions. fuel load.

a coupled atmosphere-fire model for fire behaviour simulation developed by the LANL (Los Alamos National Laboratory) and the INRA (Institut National de la Recherche Agronomique). but 50% is not efficient. about the size and location of fuel breaks in a forest landscape. and tree mortality models often also account for this effect. (ii) the size of the tree clumps does not affect significantly the fire behaviour. based on a simple formulation derived from the plume theory. Wildfires also have specific impacts in the wildland urban interface. Regarding firebrands. has been used to explore the relevance and implications of plume theory in the context of crown scorch height prediction: the assumptions of the plume theory were found to be inadequate. in particular on houses. and (iii) Higrad-Firetec. These simulations coupled with ignition models for building materials (also developed in Fire Paradox) should give an evaluation of the thermal risk in the future. but the effects of flow instability and unsteadiness on results should be further investigated before considering the operational use of such approach to fire impact measurement. (ii) WFDS (the FDS version for wildland fire simulation). Empirical models for the prediction of crown scorch height have been developed in the past. except if very dense trees (unrealistic for pines) are used to run the simulations.66 Towards Integrated Fire Management Box 2. A description of crown injury is often used as an input to statistical models predicting tree mortality after fire. FDS was used to simulate the thermal impact of a fire on a house and also the transport of firebrands around a house. showing an increase in flame residence time and on flame height in the leeward (near-wake) side of the trunk. Indeed firebrands are often the cause of house ignition at roof level. Three different models have been used: (i) FDS. as well as a two-dimensional CFD (Computational Fluid Dynamics) code. The results obtained by both approaches were. or about the effect of fuel removal or modification on fire intensity. In an attempt to examine flow and thermal conditions on a tree trunk. very similar. both computer simulations (WFDS) and laboratory experiments were conducted as complementary approaches to the same problem. The most representative of such models is the Van Wagner model for scorch height. 3D-modelling of fire behaviour and effects Three-dimensional physically-based modelling of fire has been extensively used in Fire Paradox in the frame of research studies on the understanding of fire behaviour and effects mechanisms. in the context of the evaluation of fuel break efficiency. Fire also impacts tree trunks. the Fire Dynamics Simulator developed by the NIST (mainly devoted to building or structural fires). Simulations of the effects of vegetation heterogeneity on fire behaviour have been run with HigradFiretec. when a surface fire spreads around it. . but that three-dimensional simulations are required for deeper investigations. These simulations revealed that: (i) a 25% cover of trees on the fuel break decreases significantly fire intensity as compared to the untreated pine stand (75% cover fraction). Higrad-Firetec. Similar simulations have been done in a pine (Pinus halepensis) stand using varied vegetation patterns on the fuel break. threedimensional simulations give indication of the risk they represent by monitoring of brands at different positions around the house. A simulation of fire spreading over a fuel-break is shown as an illustration in Figure 1. This conclusion could be extended to some models for crown fire initiation that also rely on plume theory. Regarding the thermal impact it was shown that twodimensional simulations can give useful results in terms order of magnitude. and so to understand the mechanisms of fire impact. in general terms. The impact of a surface fire on tree crowns is usually measured through the percentage of crown scorched in length or volume. Many questions arise from forest and fire managers about the efficiency of surface fuel reduction on tree survival.

Adapted from Pimont (2008). Continued Suppression fire is one possible use of fire against wildfires and as such could help solving the fire paradox.Fire Modelling and Simulation Tools 67 Box 2. As already known by fire managers. however. Figure 1. The fire spreads along the x-axis and the fuel-break extends from x=240 m to x=440 m. but this was observed only under weak wind conditions. Numerical simulation of a fire spreading over a fuel break in a Pinus halepensis stand. topography can help the development of such a “sucking” phenomenon. using the coupled fire-atmosphere model FIRETEC (INRA – LANL joint work). 3D-modelling using Higrad-Firetec was one approach followed to assess the feasibility and efficiency of suppression fire. Although it is difficult for the model to strictly reproduce the experiments. both approaches indicate that it is very difficult to find proper conditions for a safe and efficient use of backfires. Field experiments have also been conducted for the same purpose in Fire Paradox. Flame contours in red are deduced from computed isotherms. Cover proportion of trees on the fuel break is 25%. Green and yellow colour contours are used respectively for tree canopy and surface fuels (shrubs. . when a wildfire spreads over a flat terrain under the influence of even moderate winds: the backfire is expected to be “sucked” by the wildfire. grasses) representations (isocontours of computed biomass density).

J. rotation. UT: U. several tools are available to display information (descriptive statistics. E. Griffon.. European Commission. Export files describe the composition and the structure of the fuel complexes taking into account the physical properties of various components of the different vegetation layers (trees. Rocky Mountain Research Station. M. Forest Service. Rigolot. Ogden. References Albini. and climatic importance of vegetation fires. dedicated to hosting a wide range of models for forest dynamics and stand growth.fire. shrubs. to provide tools for managing the European database on fuels. 2009.g.68 Towards Integrated Fire Management Box 3. O. Report of the Dahlem Workshop held in Berlin. Pimont.int/fuel-manager. Rigaud.Fire in the environment: the ecological. Dynamics and modelling of vegetation fires: observations. 85 p. Pap. 400 p. and Vigy.. (eds.. updating) through a graphical user interface. A survey of available simulation platform technologies has led us to join the Capsis project. The application is connected through the Internet to the European database on fuels and manages the user’s rights. Several fire impacts on tree crown and trunk have been defined and can be visualized at the scene scale. Further details can be found at http://fireintuition. to visualize fire effects on trees and simulate post fire vegetation successions. etc) as well as on vegetation objects (selecting. Fire Paradox fuel manager: software and user’s manual. FARSITE: Fire Area Simulator-model development and evaluation. Department of Agriculture. de Coligny. Fire damage on vegetation objects have been mainly focused on fire-induced tree mortality. herbs and litter) composing the vegetation scene. A 3D vegetation scenes’ editor has been implemented allowing interactive manipulative functionalities on vegetation scenes (e. and Goldammer. A. RMRS-RP-4. 1998. FP6-018505.). P. F. zoom.J. F. An export module has been developed to prepare the set of files necessary to run the fire propagation model. 47 p..G. . Several creation modes of vegetation scenes are available including loading of a preexisting inventory file or the automatic generation of a new scene respecting a set of constraints on species distribution. E. atmospheric. The scientific objective was the representation of vegetation scenes and their transformation into fuel complexes including all the necessary parameters to run a fire behaviour model. S. Finney. Lecomte.A.1-6 of the Integrated Project “Fire Paradox”. Several renderers are available to display 3D vegetation objects. indicators) on the vegetation scene content or on the current selection. In: Crutzen. The technological objectives are to implement a user friendly platform to generate fuel complexes in 3D. adding. I. Product P6.efi. F. 1992.S. Moreover. Fuel editor The Fire Paradox Fuel Manager (Lecomte et al 2009) is a piece of computer software integrated in the data processing chain between the European data and knowledge base on fuels and the 3D physically-based fire propagation models. A new CAPSIS module – “FireParadox” – has been developed which implements data structure and functionalities of the Fire Paradox Fuel Manager. Project no. 15–20 March 1992. Res..

. Wildland surface fire spread modelling. International Journal of Wildland Fire 18: 369–386. 1990–2007.L. 323 p. A. F. Université de la Méditerranée. A. 2009a. 1990–2007. Doctoral Thesis. Sullivan. 1: Physical and quasi-physical models. 2009b.Fire Modelling and Simulation Tools 69 Pimont. Aix Marseille II. Wildland surface fire spread modelling. Sullivan.L. International Journal of Wildland Fire 18: 349–368. Modélisation physique de la propagation des feux de forêts: effets des caractéristiques physiques du combustible et de son hétérogénéité. 2: Empirical and quasi-empirical models. 2008.

.

2003). France and Spain in 2003. aiming at the identification and mapping of existing or potential WUI areas have been recorded in North America and to a lesser extent. Canada and Australia interest in study of WUIs increased after the huge fires of 1985 (Davis 1990). Essentially in USA.3 Wildland Urban Interfaces. Spain 4 Department of Geography. mainly during heat waves. 2005a). Luis Galiana3. destroying houses (LampinMaillet 2008). such as the destruction of homes by wildfires. Aix-en-Provence. Greece 3 Department of Geography. Mapping and Assessment Corinne Lampin-Maillet1. Lazaridou Thalia2. Camia et al. were influenced by human activities to a large extent. More crown fires involve WUI zones.1 Introduction Large areas initially consisted of contiguous forests. These areas that are characterized by increased human activities and land use conversion make up the wildland urban interfaces (WUIs). Autonomous University of Madrid. Marlène Long1. They are areas of human-environment conflicts.3. Spain in 2005. as landscape units. Radeloff et al. The aim was to assess fire risk in WUIs and particularly their vulnerability in order to improve the efficiency of prevention actions for the protection of WUIs from wildfire propagation. So large efforts. Spain 3. France 2 OMIKRON Ltd. University Complutense of Madrid. This influence contributed to the fragmentation of the rural landscape and forested areas were found surrounded by or intermixed with urban development. 2005a). 2004). Wildfire risk can be defined as the expected loss due to wildfires on a certain area and period of time. and Greece in 2007.3. Extreme fires affected Portugal. The significance of WUIs has grown in recent years mainly because WUIs. Oskar Karlsson4. Marielle Jappiot1. natural habitat fragmentation. Antonis Mantzavelas2. Gema Herrero4. Portugal and Spain in 2006. Urban and economic development in or near wildland vegetation poses a major threat to the environment (Johnson 2001. Many definitions of wildfire risk exist (Hardy 2005. have grown worldwide (Steward et al. introduction of exotic species. Since 2000 in Europe the WUI becomes involved in the forest fire environment. particularly in Europe. in Europe. The most common corresponds to the combination of hazard with vulnerability: risk = hazard*vulnerability. Residential areas are increasingly affected by wildfires with damage on goods and people. Fire Behaviour and Vulnerability: Characterization. Partozis Thanassis2 1 Cemagref. establishing relationships between . and biodiversity decline (Radeloff et al. Apostolopoulou Iossifina2.

As fire fighting resources and fire prevention infrastructures are not infinite. Fire hazard is generally computed as an index (San Miguel-Ayanz et al. (2002) proposed to define each component of fire risk as composed of different overlapping elements (Table 1). topography). then the outcome would be a structural (static or long-term) index (Jappiot et al. 2009. if the assessment is based upon factors that are likely to change frequently (daily or even hourly. in different and contradicting ways (Camia et al. or as a cause (Mantzavelas et al.g. On the other hand. 2009). fuel moisture. weather) then the outcome would be a dynamic (short-term) index. we consider fire hazard as the combined outcome of the probability of fire occurrence and the potential intensity of the resulting fire.g. Fire risk definition with detailed elements. We consider the probability of fire occurrence as the likelihood of a fire to happen in an area. Jappiot et al. Resources need to be allocated wisely. Blanchi et al. and (ii) the potential damage that the fire will cause once it occurs – vulnerability (Blanchi et al. • Concerning the other component vulnerability. But we also find other fire risk definitions where intensity is considered as part of vulnerability (Jappiot et al. Usually fire risk assessment consists of a combination of fire hazard assessment and vulnerability assessment. vegetation. 2003). the need for predicting a wildfire incident and its consequences becomes apparent. for the sake of operation and cost effectiveness. there are many different concepts of vulnerability across disciplines and topics (Gallopin 2006). as a state or characteristic. 2003). in space and time. in the case of a fire incident (Byram 1959). 2004). The assessment of forest fire risk is generally recognized as an indispensable component of fire prevention and suppression systems. 2002. similarly to the definition used by Blanchi et al. . 2004). If the assessment is based upon factors. In the present work. FIRE RISK Hazard Occurrence Ignition Probability Wildfire Probability Intensity Threatened area Wildfire Intensity Stake Reply Response Vulnerability theses three concepts (Chen et al. 2009). • Various authors in risk-related issues have used the term fire hazard. So fire risk encompasses two different components: (i) the probability that a wildfire affects the area during that period of time – fire hazard. 2005). 2008). (2002). e. fire intensity is considered as the potential energy release per unit length of fire front.72 Towards Integrated Fire Management Table 1. that change very slowly over time (e. Wilson et al. Assessment methodologies are categorized by their temporal scales (Marzano et al. According to natural risks. Finally. vulnerability can be considered in three ways: as a consequence.

The great challenge is to express vulnerability in measurable units or indices in order to be used for further estimation of the total fire risk. vulnerability may encompass the degree of fragility of men. In short. Mapping. where houses meet or intermingle with undeveloped wildland vegetation (USDA-USDI 2001). the WUI commonly described as the area where urban areas meet and interact with rural lands (Vince et al. The issue of the chapter is to map the location of WUIs on the territory and to determine which WUIs are at the greatest risk. from the principal hazards which have been identified (Coburn et al. According to Blaikie et al. According to Coburn et al. resulting from a given hazard at a given severity level. vulnerability is defined as the degree of loss to a given element (or set of elements). built environments and ecosystems to the negative consequences resulting from the exposure to hazardous events.. 2004). definitions present WUI as the line. to interact between two systems. For that several objectives are . Commonly. 73 In the first case vulnerability is the value that could be lost when hazard occurs. 1993). 2005. and islands of undeveloped lands within urban areas (Alavalapati et al. most of disaster mitigation work is focused on reducing vulnerability. 1994). interface is defined as the contact plan or contact line between two different systems (Brunet et al. The term of WUI community exists with the following definition “the urban-wildland interface community exists where humans and their development meet or intermix with wildland fuel”. vulnerability expresses the capacity of a person or group to anticipate. In geography. when a hazard occurs (D’Ercole 1998). (1994). Caballero et al. it is obvious that the definition of the vulnerability refers mainly to the impacts of a catastrophic event. and in order to do it. cope with. areas where homes and other structures are intermixed with forests and other land uses. Finally.Wildland Urban Interfaces.. Their spatial and temporal dynamic can produce situations which can be more or less dangerous for exposed society” (D’Ercole 1996). in the third case vulnerability corresponds to a “system which considers a lot of variables (natural and human). The concern of this chapter is the wildland urban interface and fire risk assessment. or zone where structures and other human developments meet or intermingle with undeveloped wildland or vegetative fuels. specifically human and wildland systems (Carroue et al. services and management (Macie and Hermansen 2002). organized societies. Furthermore. area. 2002). McGee 2005. In these WUI increased human influence and land use conversion are chancing natural resource goods. 2002). Caballero et al. Nowadays and more generally. (1994). 2005). It constitutes a privileged zone to exchange. (2004) stated that vulnerability can be expressed through the calculation of the potential damage when a single unit is exposed to a certain level of danger. there is a need to understand which elements or units are the most exposed at risk. economic structures. The vulnerability of an element is usually expressed as a percentage loss (or as a value between 0 and 1) for a given hazard severity level (Blanchi et al. resist to and recover from the impact of natural hazards. In the literature. In this case the aim is to identify the factors (variables) that are the source of vulnerability. Fire Behaviour and Vulnerability: Characterization. It supposes to define precisely this term. In the second case vulnerability can be the propensity of an element or a set of elements to suffer damage. includes the edges of large cities and small communities.

2002). developed. Location of the three European study cases. commercial and public buildings were not taken into consideration). industrial. a precise definition of the WUI has been proposed as follows (Lampin-Maillet et al. and at a landscape scale with a view to analyse their territorial development in the landscape. The third objective is to propose a method to assess vulnerability levels. In order to develop the previous objectives. The second objective is to propose a fire hazard calculation and mapping process taking into account structural and daily factors.3. temporarily or seasonally (agricultural. . The fourth and last objective is to present a specific approach developed with a view to assess and to map fire risk in WUIs through a total fire risk index. 3. Spain and Greece. which are inhabited permanently. 2010): • WUIs are composed of residential houses. The first objective is to present a method to characterize and to map WUIs at a local scale with a view to improve fire prevention. characterization and mapping of WUIs in European Mediterranean countries In order to map WUIs on the territory. three study cases have been defined in three European countries: France. The second one is located in southeastern Spain in Sierra Calderona (site 2 in Figure 1).2 Identification.74 Towards Integrated Fire Management Figure 1. Tools have been developed as the result of scientific research. The first one is located in southeastern France in the Metropolitan area between Aix-en-Provence and Marseille (site 1 in Figure 1). The third one is located in northeastern Greece in the western part of the prefecture of Thessaloniki (site 3 in Figure 1). • Houses are located at 200 m from forests or shrubland to consider area where brush-clearing is partially required or spot fires occur (Colin et al.

Then the combination of different types of dwellings and different classes of horizontal structures of vegetation produced a WUI typology. The second step is to characterize and map the structure of vegetation. Calculated on vegetation class. (2009) such as housing density. In other European countries. Mapping. We considered WUI as two intermixed elements: the first concerned the spatial organization of residential houses. each house was classified as belonging to one of the four configurations of houses: isolated. and the second concerned the structure of fuel vegetation. Concerning the structure of dwellings. the WUI area is located up to 300 m from forests.3. the effective fuel treatments are required only within a 10-30 m radius. Their distinction is based on quantitative criteria described in Lampin-Maillet et al. Spatial criteria had to be developed to specify the structure of dwellings in contact with the different vegetation structures. the most appropriate index to measure aggregation of spatial patterns is the aggregation index (AI) (Lampin-Maillet et al. 2002 which makes brush clearing obligatory within a 50 m perimeter around each house located at a distance of less than 200 m from forests or shrublands. and very dense clustered housings.. 2010). These classes take into account the distance between houses and the density of houses located within a 100 m radius around houses.Wildland Urban Interfaces. Among the different existing metrics in landscape ecology (McGarigal 2002). The method used to characterize and to map WUIs is based on three steps. Fire Behaviour and Vulnerability: Characterization. only the horizontal structure of vegetation that can be spatially recognized has been characterized (no vegetation. This aggregation index has a spatial representation. The structure of vegetation reveals its horizontal continuity which is designed for the measurement of aggregation levels of spatial patterns within the vegetation class in a land-cover map. scattered. The first step is to characterize and map the housing configuration. WUIs are the subject of the French Forest Orientation Law of July 9.2. scattered. dense (or very dense) clustered dwelling types. This distance takes into account the perimeter wherein fuel reduction operations can be imposed on home owners. we proposed a real and quantitative definition of terms corresponding to isolated.c). dense clustered. 75 • WUI are delineated by a radius of 100 m around the houses. (2009) and through the process of buffering and house counting described in the same paper. Therefore.b. Then according to the definition of the dwelling types established in Lampin-Maillet et al. 2010) Considering the above definition. These different distances are particularly adapted to the European context but they also could be changed according the specific local context of the country (vegetation clearance regulation or urban organization). The houses considered as located in WUIs are selected. sparse vegetation and continuous vegetation). Concerning the structure of vegetation. the aggregation .. after first approaches developed with housing density calculation (Lampin et al. 2007a. 3. in France. usually used by land managers and geographers. garrigues (200 m + 100 m) so it is also significantly exposed to firebrands from vegetation in the case of fire.1 At local level (Lampin-Maillet et al.

Radeloff et al. The WUI method has been applied in different areas in France and Spain. orchards/vineyards. 2010).76 Towards Integrated Fire Management Figure 2. and the two other classes were determined by sharing the numbers of value equally into two groups or by setting a threshold value equal to 95%: the first distribution of numbers were considered as low values of aggregation. 2005b. Lampin et al.g. . The third step is combining the two previous criteria through a geographical information system (GIS). and mixed forest). quarries. small grains. The calculation allowed mapping of the WUIs according to 12 types (Figure 2) by crossing four classes of dwelling types and three classes of vegetation aggregation indices in a raster format. scrublands. row crops) and pasture. bare rock/sand/clay. Aggregation metrics calculations were made within a moving window with a radius of 20 m and a map of aggregation index values was also drawn up including three classes of AI values.and high-intensity residential. 2006). and perennial ice/ snow (Steward et al.. open water. Typology of wildland urban interfaces (Lampin-Maillet et al. 2003. urban/ recreational grasses. the second one as high values. index enhances spatial organization of forests and scrublands. deciduous. The first class concerned values equal to zero. site 1 of Figure 1. transitional lands (mostly clear-cuts). commercial/industrial buildings. fallow. Figure 3 is an illustration of the WUI map carried out on French site. Excluded from vegetation are low. pasture/hay. arable land (e. Vegetation is defined as wildland forests (coniferous.

2007) The local scale of analysis must be complemented by a more general approach. . On one hand. multiscale analysis. i. and the interactions between them. 2004).Wildland Urban Interfaces. Map of wildland urban interfaces in study site 1 (Lampin-Maillet et al. The model is based on the analysis of structural (distribution of land uses) and functional elements. Fire Behaviour and Vulnerability: Characterization. The results from this method improve the landscape character assessment. fire behaviour is largely dependent on the landscape pattern. The WUI characterization developed results from the analysis at different scales (regional.2 At landscape level (Galiana et al. Mapping.e. These are two major reasons to apply the landscape character assessment method developed as an intermediate scale between local and regional levels (Bürgi et al. 77 Figure 3. WUI expansion processes are associated with counter-urbanization and development of second home dynamics which respond to spatial organizational models on an urban–regional scale (Antrop 2004). landscape.3. 3. and ultimately on the pattern of new residential development. Also. At the regional level. wildfires) and definition of its main spatial patterns.. 2010). local). On the other hand.. abandonment and transformation of the rural areas. the existing landscape pattern exerts an influence on the land cover and land use trajectory.2. the objective of analysis is the establishment of a descriptive model of the main territorial dynamics influencing the area (suburbanization.

Spontaneous XI. Spontaneous XII. Spontaneous urbanizations (not planned) on wildland terrain IV.3 Fire hazard assessment in WUIs (Lampin et al.3. the WUI characterization process consisted of identifying the different morphologies defined by urbanization processes. Agricultural foothill plains Towns Urbanizations Scattered rural settlements IX. Compact towns with extensions III. Scattered settlements on wildland terrain I. WUI characterization. the multiscale analysis consists in the Landscape Character Assessment (LCA).3. Compact towns II. The map can be updated on a daily basis.1 Steps for fire hazard mapping The work presented hereafter is aiming at the definition of a standard workflow for obtaining a fire hazard map that will take into account structural and dynamic factors within a GIS. Urbanizations planned cultivated slopes on wildland terrain VII.3. Figure 4 is an illustration of the WUI map at landscape level located in site 2 of Figure 1. Wildland mountain with lower bunter sandstones and cultivated gullies IV. 2009) 3. Western flat topped peaks III. Agroforestal slopes V. Landscape description established a hierarchical typology composed of two levels: landscape units and types (Countryside Agency and Scottish Natural Heritage 2002). Relation matrix from the combination between the morphology of the settlements and the type of landscape (Specific case: Sierra Calderona) Type of landscape/ Morphology of settlements I. At the local scale. Urbanizations planned - At the landscape level. The LCA provides the territorial context of the WUI studied in order to correlate urbanization processes with the type of landscape in which they occur and with the foreseeable risk evolution in these areas. 3. examining them in the context of the landscape type of the area in which they occur and characterizing them according to fire behaviour. . Scattered urbanizations (not planned) settlements on VI.78 Towards Integrated Fire Management Table 2. Scattered rural urbanizations (not planned) settlements VIII. based on the natural and cultural features present in the landscape and on evaluation of functional dynamics and uses. Scattered settlements on terraced slopes X. Urbanizations planned on wildland terrain - V. Small agricultural valleys II.

distance from settlements. Factors that are usually taken into account usually are: . 2009) and neural networks (Vasilakos et al. aspect. These variables should be regressed against a data set containing the perimeters of historic fire incidents. weights are assigned to each of the remaining variables . elevation. new techniques have been elaborated like principal component analysis (Xu et al. .. Further. Fire Behaviour and Vulnerability: Characterization.Topography (slope.g. biomass. distance from roads etc. Factors are integrated into equations. 2006). average annual rainfall. . Calculation of a structural index (Probability of occurrence) There are several approaches attempting to assess the combined impact of structural factors on fire hazard. in order to identify if any of the variables in consideration are not related to fire occurrence. logistic regression (Chou and Minnich1993.). Chi2 significance test) should be performed. Among those new techniques. site 2.Human presence (population. structure). In a second step. an exploratory analysis (e.. each one weighted according to the author’s opinion on its relative importance. 2007). The accuracy of the assessment can be easily estimated by the observed fire incidents. 79 Figure 4. Martinez et al. These techniques seek to establish a relationship between those factors and fire history. To overcome the subjectivity in weighting of the contributing factors. Mapping. WUI map at landscape level in Spanish study site. to produce hazard indices. aspect). elevation.).Wildland Urban Interfaces. etc. logistic regression has been extensively used as it provides a probability of occurrence as an output.Vegetation (type. In a first step. a group of potentially significant variables for the prediction of fire occurrence must be defined (for example: fuel types. distance from nearest road.

weather. when Zi approaches negative infinity. + bjXji + e (2) Where Xji is the value of the jth variable in consideration in the ith geographic unit (pixel or polygon). For instance. different strategies have been applied. The logistic model of fire occurrence probability can be expressed as: PFOi = (1) Where PFOi is the probability of a fire to occur in the ith geographic unit (pixel) and Zi can be calculated as: Zi = b0 + b1X1i + b2X2i + . relative humidity. 2004. 2007). Value range for the PFO: PFO. Alexander 2008). values from 17 to 32 are considered as high values. But. the Composite Index (CI) As discussed above. The Canadian Fire Weather Index (FWI) is generally considered as one of the most efficient in defining the daily risk. because it does not take into account the current status of fuel. Thus. 1994). Value ranges for the FWI: FWI is open-ended with ranges from 0 to 100+. A schematic presentation of the FWI structure is shown in Figure 5. For these reasons the FWI is the dynamic index proposed also here.80 Towards Integrated Fire Management according to its significance.. etc. indicating that the ith geographic unit is not going to burn. ranging between 0 and 1. Values from 0 to 8 are considered as low values. wind and rainfall) are usually derived from meteorological stations or sensors. the value of the index approaches 0. a heavy rain incident can reduce the chance of fire to zero. indicating that the ith geographic unit is definitely going to burn in the next period. Calculation of a daily (dynamic) index Considering the short-term component. the point source values of those variables have to be interpolated. values from 8 to 17 are considered as moderate values. before the final calculation of the FWI (Mantzavelas et al. the structural index PFO is calculated by considering all factors (human. represents the ignition risk that .. when Zi approaches positive infinity. the structural index is a long-term predictor of fire occurrence and can only be considered as an average probability. there is a need to obtain a surface of values for each one of the above-mentioned variables. and values greater than 32 are considered as extreme values (Raínha and Fernandes 2002. Most of them consist of trying to assess the effect of fuel water content on fuel flammability. as implied by the calculation method. topography. vegetation. even in a place where the (long-term) occurrence probability is thought to be high. the value of the index approaches 1. which is calculated by means of logistic regression. Similarly. Viegas et al. In order to calculate an FWI surface (map).) that contribute to the occurrence of the wildfire phenomenon. as it has been globally applied and evaluated in very diverse ecosystems besides Canada (Marzano et al. being a probability. On the other hand. The values of the required variables for the calculation of the FWI (temperature. As implied by equation 1. Combination of structural and daily indices. bj is the weight of the jth variable and e is an error term. which is thought to be a very significant factor in the occurrence of wildfires. a meteorological index like the FWI.

being the product of the PFO and the FWI has the same value range as the FWI (0 to 100+). PFOi is the value of the Structural Index for the ith geographic unit. The CI can be considered as the ‘valid’ portion of the FWI calculation. we have only dealt with the issue of fire occurrence. in order to better understand and define the wildfire hazard issue.g.Wildland Urban Interfaces. provided that the calculation of the probability of fire occurrence in PFO is correct. Fire Behaviour and Vulnerability: Characterization. such as: CIi = PFOi * FWIi (3) where: CIi is the value of the Composite Index for the ith geographic unit. Mapping. is contributed upon the daily fluctuations of the weather and their impact on fuel moisture status. We propose. it is reasonable to think that: (a) all ignitions cannot be explained by a low fuel moisture level. Simplified structure diagram for the Fire Weather Index (after Van Wagner 1987). in the case of a lack of fuel). then the CI is the part of the FWI value that remains meaningful under the combined influence of structural factors and fire history. That is. In other words. In order to have a broader view on the issue of fire hazard.. Thus. Value ranges for the CI: CI. whereas if the PFO value is 1 then the CI value is 100% of the FWI value). that an integrated approach is needed. and (b) not all ignitions develop into fires (e. for that reason. 81 Fire Weather Observations Temperature Relative Humidity Wind Rain Fine Fuel Moisture Code Wind Temperature Relative Humidity Rain Temperature Rain Fuel Moisture Codes Duff Moisture Code Drought Code Initial Spread Index Build Up Index Fire Behaviour Indexes Fire Weather Index Figure 5. which is the product of the Structural Index (PFO) and the Fire Weather Index (FWI). The Potential Fireline Intensity (PFI) Until now. Although the FWI is considered as quite efficient in defining the daily ignition risk. and FWIi is the value of the FWI for the ith geographic unit.. it becomes apparent. we have to explore the possible . the calculation of a Composite Index (CI). we may say that CI is the part of the FWI value that is validated by structural factors and fire history (if the PFO value is 0 then the CI value becomes 0.

3. usually taken to be equal to 18 000 kJ/kg (Byram 1959): PFIi = 18 000 * Wi * ROSi (4) Where: PFLi is the fireline intensity for the ith geographic unit (kW/m). A way to do so is to calculate the potential fireline intensity. 3. and ROSi is rate of fire spread in the ith geographic unit (m/sec). 2002).2 The Hazard Index (HI) building and mapping Finally the Hazard Index (HI) can be seen as a combination between the Composite Index and the Potential Fire Intensity. We propose the relationship defined in Table 2 for the calculation of HI.82 Towards Integrated Fire Management Table 3. The calculation of fireline intensity is significant in fire suppression and the study of the ecological effects of fire. the CI indicates the effects of current weather conditions and fire history upon fire occurrence. from 3500 to 7000 kW/m are considered as very high values. Wi is the available fuel mass in the ith geographic unit (kg/m2). the rate of spread and a ‘constant’ number. For example. and the potential fireline intensity is low (‘Low’ PFI value). if current weather conditions do not favor ignition or spread and this is also validated by the fire history (‘Low’ CI value). fire spread. from 350 to 1700 kW/m are considered as moderate values. The fireline intensity is calculated as a function of fuel mass. while the PFI indicates the severity of the fire should an ignition occur. Intensity values from 0 to 350 kW/m are considered as the low values. from 1700 to 3500 kW/m are considered as high values. On the one hand. and values greater than 7000 kW/m are considered as extreme values (Lampin et al.3. which is the potential energy release per unit length of the fire front in the case of a fire incident. mean values for Wi and ROSi are normally incorporated in the description of fuel types. Provided that fuel types are mapped according to the classification scheme introduced by Anderson (1982). The idea that stands behind the HI calculation is to integrate into one single index. CI\PFI Low Moderate High Extreme Low Low Low Moderate Moderate Moderate Low Moderate Moderate High High Moderate Moderate High Very high Very high Moderate High Very high Extreme Extreme High Very high Extreme Extreme consequences from a certain fire incident (Alexander 2008). the issues of fire ignition. and the potential to cause damage. The HI embodies the different aspects of the indices that this index is calculated from. then we should expect a low intensity fire. Value ranges for the PFI: PFI values range between 0 and 100 000 kW/m. or another classification scheme resulting in the description of fuel. or . Calculation of the Hazard Index.

similarly. ‘High’ and ‘Very High’) can be explained. Fire Behaviour and Vulnerability: Characterization.3. Vulnerability assessment comes from its threefold consideration (a consequence. Based on this approach. 3. related to the fire characteristics and by the ability developed by society to face the danger of wildfire.. no fire at all (‘Low’ HI value). Example of calculation of the HI in the Greek study site 3 for 14/04/2007. In consequence vulnerability is formed by an internal component..4 Vulnerability assessment in WUIs The main objective is to develop a set of processes for mapping a synthetic index of the WUIs vulnerability to forest fires. Figure 6 illustrates the map of this HI. 83 Figure 6.Wildland Urban Interfaces. related to the effects of the fire caused on the value of the affected goods and its capacity for recovery. as the kind of fire behaviour and intensity that is expected at a given place and time. and the potential fireline intensity is ‘Extreme’. and by an external component. a wide variety of factors influencing vulnerability of the territory to forest fires have been identified and methods to obtain parameters . Mapping. and a cause) as it is considered in the different definitions used above. a state or characteristic. If current weather and fire history indicate that fire occurrence is very likely (‘Extreme’ CI value). then we should expect a very intense fire that may cause extensive damage (‘Extreme’ HI value). Intermediate classes of the HI (‘Moderate’.

the vulnerability index can be determined and will show whether a particular area is susceptible to suffer damage from wildfire. These parameters are the basic units that have to be mapped and standardized into a common. Flow chart illustrating the elements involved in assessing vulnerability index. simple quantitative scale in order to allow them to be aggregated through a process based on multicriteria evaluation. the vulnerability index and it will show whether a particular area is susceptible to suffer damage from wildfire.84 Towards Integrated Fire Management Figure 7. The aggregation process can be repeated until all the hierarchical structure is aggregated into one single indicator. the modelling of potential high risk situations and the historical analysis of wildfires have played a very important role. and its cartography at local level for the calculation of these factors have been proposed (Figure 7). where each variable is given a particular weight. In the task of obtaining these parameters. also the conclusions obtained from the tasks of WUI characterization have been . Each factor will be further subcategorized into parameters. When all the components are calculated.

The method developed allows assessment and mapping of fire risk levels. it is important and efficient to focus risk assessment in the WUIs. are useful for the managers of territorial planning. a new perception of the territory is possible: WUI area. incorporated. Considering the WUI map.Wildland Urban Interfaces. Map of vulnerability levels on Spanish study site (site 2 in Figure 1). ignition probability and combustibility. . The final index which can be mapped (Figure 8) and the intermediate components. and outside WUI area. Because of their high vulnerability. Consultations with forest management and fire fighting experts of the experimental area (Sierra Calderona.5 A specific approach of fire risk assessment in WUIs through a total risk index (Lampin-Maillet 2009) The work presented hereafter is aiming at the definition of a process for obtaining a fire risk map through the calculation of a total fire risk index based on a WUI map. Mapping. España) by applying a DELPHI methodology have been fundamental for the determining the weights given to the factors that make up each of the indices.3. Fire Behaviour and Vulnerability: Characterization. Valencia. the factors have been grouped into four synthetic components (difficulty of extinction. demand for forest defense... 85 Figure 8. 3. demand of civil protection and territorial value) which are integrated into a global index. factors and parameters. The map can be updated on a basis depending on the territory WUI extension. emergencies and forest fire prevention and fighting. Finally.

In the case of the study area the equation is: Fire risk total index RI = 0. So a Fire risk total index RI has been built corresponding to a linear combination of the three indicators having the same weight but corrected by their explanation level (R2 value). housing density.86 Towards Integrated Fire Management A spatial analysis on the studied territory was performed in order to establish relationships between the distribution of fire ignition points and burned areas and different land cover data. environmental data. housing density. In reference to fire risk definition. WUIs corresponding to very dense clustered dwellings present also a high level of ignition density linked with human activities but a low burned area ratio (high urban component and low vegetation component) (Lampin-Maillet et al.Fire Ignition Density FID = Exponential function (territory type. 2010).Wildfire Density WD = Exponential function (territory type. A total index of wildfire risk was developed combining the three previous indicators (Lampin-Maillet 2009). road density. For that. exposure to very warm temperatures) with R2 = 57%. road density. Their combination can contribute to a pertinent and efficient assessment of fire risk. Results of this analysis are expressed through three main functions (or fire risk indicators) based on statistical multiple regressions with R2 more or less high: . The spatial analysis also allowed identification of a set of conditions that correspond with high fire risk in WUIs: housing density. it is possible to identify specific WUIs which present a high level of fire risk. . It comprised fire ignition points during the 1997–2007 period for which the fire area was more than one hectare. Around 565 fire ignition points were located in the study area. each of the three indicators includes information about hazard and/or vulnerability: Fire Ignition Density FID and Wildfire Density WD are particularly concerned by fire occurrence (ignition probability and wildfire probability) and Burned Area Ratio BAR is related to hazard and vulnerability through the intensity element. Figure 9 shows that WUIs corresponding to isolated dwellings present a high level of fire risk due to high levels of ignition density and burned area ratio. . vegetation more or less continuous. a digitalized database of fire ignition points created by the French National Forest Institute (ONF) was used. A digitalized database of burned area produced by the Administration of Agriculture and Forest of Bouches-du-Rhône was used. coniferous forests. It comprised 109 wildfires recorded study area during the 1990–2007 period. Provence-Alpes-Côte-d’Azur Region and a Spot Image from Spot 5 satellite imagery on the study area were used. At last. agricultural and natural components) at the 1:10 000 scale. land cover type. land cover type.63 BAR . WUI types. country road density. low aggregation of vegetation) with R2 = 36%. a thematic land cover layer obtained from the Spot Thema database elaborated in 2004 by CNES. housing density) with R2 = 51%. altitude. garrigues.89 FID + WD + 0.Burned Area Ratio BAR = Polynomial function (territory type. land cover type. The detailed level describes the territory (urban. As a result of the relationships established between WUIs and fire indicators calculated with past fire data (fire ignition density and burned area ratio).

87 Figure 9.. Figure 10. Map of wildfire risk global index in WUIs in the French study site (site 1 in Figure 1). . Mapping. Fire ignition density and burned area ratio according to WUI types in LampinMaillet (2009).. Fire Behaviour and Vulnerability: Characterization.Wildland Urban Interfaces.

63 corresponding to the ratio 36% / 57% for BAR. and particularly the vulnerability of the territory. combining fire hazard and vulnerability assessments. The WUI is often a location which is particularly appreciated as an area in which to live. and a correction of 0. These two processes contribute to assessment of fire risk. and also decreased from a high aggregation index to a zero aggregation index (Lampin-Maillet et al. Introducing the risk of wildfire with such maps. 2009). The method we developed for mapping . isolated WUIs with low and high aggregation indices of vegetation presented the highest values. The WUI map is key information to identify locations where vegetation has to be reduced in order to protect the houses and their inhabitants in case of wildfire and where careful behaviour are essential to avoid fire ignition. Regarding fire ignition density and burned area ratio. Certain types of WUIs represent a high level of fire risk in terms of fire ignition density.89 corresponding to the ratio 51% / 57% for FID.6 Conclusions In this chapter we proposed reproducible methods for characterizing and mapping WUIs at large scales and over large areas (Lampin-Maillet et al. 2009).88 Towards Integrated Fire Management With respectively a correction of 0. A map of this fire risk total index can be produced as illustrated at Figure 10 in the South of France. and on the other hand a method for vulnerability assessment and mapping (Galiana et al. a value of 1 (the best R2 value) corresponding to the ratio 57% / 57% for WD. However. wildfire density and burned area ratio. We also proposed on the one hand a method for fire hazard assessment and mapping (Lampin et al. Results also highlighted the fact that the burned area ratio generally decreased from isolated WUIs to dense and very dense clustered WUIs. These results could have interesting implications for fire prevention and land management. A specific approach has also been developed to calculate and to assess a total index of wildfire risk in WUIs (Lampin-Maillet 2009). 2010). 2007). Scattered WUIs with both low and high aggregation indices of vegetation also represented a high level of fire ignition density and burned area ratio even if these values were lower than those for isolated WUIs (Lampin-Maillet et al.3. is a way to make the inhabitants becoming aware of fire risk in WUIs. 3. 2010). This will globally decrease the risk of fire either by reducing fire propagation through biomass removal and/or by reducing fire ignition probability together with less carelessness. Accomplishing this goal is strictly related to the designation of suitable prevention messages and preventive actions which can be different according to WUI types. in our European Mediterranean context this lifestyle carries a certain risk: people should always be aware of the existing fire risk in such WUI and should respect and apply efficient recommendations to insure against risky situations. 2010) and at landscape scale (Galiana et al. WUIs have increased considerably all over the world in recent decades and this trend will certainly continue in the coming years due to the continuing trend of land abandonment combined with urbanization.

and Jacobson. Bovio. UNDP/DHA. R. California. Landscape change and the urbanization process in Europe. 2005. B. (ed). Byram.. J. Antrop.H. 2004. C. S.P. and Wisner. Mapping Probability of Fire Occurrence in San Jacinto Mountains. Karteris. D.. I.Wildland Urban Interfaces. Simon. Environmental Management 17(1): 129–140. Martinez. F. 2008. Martin. A. At risk: natural hazards. Davis. Ogden. 28 p. Carrega. Bürgi.. L. people vulnerability and disaster. in association with Scion. Proceedings of the IV International Conference on Forest Fire Research.. Molina. 2002. M. J.. .S. Camia. 89 WUIs is an appropriate tool for the assessment of WUI dynamics and associated fire risk dynamics in the context of ongoing changes in climate... L. C. M.. National Rural Fire Authority. EUFIRELAB/ EuroMediterranean Wildland Fire Laboratory. R. Luso. 2004. Sergiopoulos. D. A... Landscape and urban planning 67: 9–26. UT. 2004.. T. Conese. urbanization and vegetation. References Alexander. New York.H. R. D. 70 p. New York. Chou. M. N.. G. Disaster management training Program. and Pmomonis.. Rural Fire Research Group. Bonova.P.. A. Portugal. I.. and Schneeberger.Deliverable D-10-01Wildland-Urban Interface Management: A state of the art. Environmental Management 31 (4): 546–560. K. Martinez. Claval. Brunet.. P... Renard. 1993 (3rd Ed. Hersperger. Wellington. Dossiers des images économiques du monde. Limites et discontinuité en géographie.. Eufirelab: Euro-Mediterranean Wildland Fire Laboratory. E.E. and Alexandrian.M. 1982.E. M.A. Vulnerability and risk assessment. Forest fire risk assessment and cartography..R. H. Berolo. G. R.. J. Routledge. USDA Forest Service General Technical Report INT-122.. and Veyret. Paris.. G. Y. G.R. Forest Fire: Control and Use.. D.M. 1994. D. Aids to determining fuel models for estimating fire behaviour. a "wall-less" laboratory for wildland fire Sciences and technologies in the Euro-Mediterranean Region Deliverable D 08-02: Wildland fire danger and hazards: a state of the art.. I. Coburn. Blaikie. Giroud. 1993. Deshayes. Mapping. D. A... P. Chen. Intermountain Research Station. Combustion of forest fuels. Ferras. Guglielmet. 2003. J. Vayreda.. and Xanthopoulos. Proposed revision of fire danger class criteria for forest and rural areas in New Zealand. A. Edition SEDES. 1959. and Théry. Forest policy and Economics 7: 705–708.. Caballero. 518 p. a "wall-less" laboratory for wildland fire Sciences and technologies in the Euro-Mediterranean Region . Di Meo. Salas. USA. 1994. (ed. Blond. 22 p. Carroue. M. P. McGraw-Hill. D. Montpellier. Fire Behaviour and Vulnerability: Characterization. 2002.). Alavalapati.. Picard. Lampin. C... Les mots de la géographie.. Landscape Ecology 19: 857–868.. M. 280 p. P. 62 p. and Newman. Wildland-urban interface: Challenges and Opportunities. 2004. Miossec. J. C. 2nd Ed. 57 p. Spence. Jappiot.R. London. Christchurch. and Minnich. Y. Fox. Sabate. Driving forces of landscape change – current and new directions.J. G. W. L. Anderson. Cannon. Towards an integrated Approach to natural Hazards risk assessment using GIS: with reference to bushfires. 159 p. La documentation française.). Blanchi. A methodological approach.. and Gitas. R. In: Viegas. Carter. J. In: Davis K. H.W. Mallinis. R.

. E. Loureiro. Gorostiaga. J.. Long.) EFI Discussion Paper 15. A. Application à la région du volcan Cotopaxi (Equateur). Long M. FP6-018505. J.. Book of Abstracts of the IV International Wildland Fire Conference. Landscape Character Assessment: guidance for England and Scotland (Wetherby. Environment and urban systems 33(2009): 224–232. Caracterización y clasificación de Interfaces Urbano-Forestales mediante análisis paisajístico. Madrigal. D. C. H. G.. Delboube. Galiana. European Commission. Vega. Marcillat. Jappiot.. G. España). Loddo.Sustainable Development. SALTUS . Characterization and mapping of dwelling types for forest fire prevention. L.Y. D’Ercole.. Lazaridou. J. Mantzavellas. M. Deliverable D5. Environmental impacts of urban sprawl: a survey of the literature and proposed research agenda. and Borgniet. R. C. The wildland-urban interface: paradise or battleground? Journal of Forestry 88(1): 26–31.. G. Loddo. Herrero. Bouillon. and Dimitrakopoulos. 2009. C. Book of Abstracts of the IV International Wildland Fire Conference. Bulletin de l’IFEA 25(3): 479–507. Assessing Wildfire Risk in Time and Space. A. Morge.. A. 2010. Loddo. R. Ditana. European Institute of the Mediterranean.. M. Diez. 2007a.Botelho. Alexandrian. Countryside Agency and Scottish Natural Heritage). Spain. C. L.. C. Pereira. T. Jappiot.. L.. Environment and Planning A 33: 717–735. Herrero.. D.. Organismo Autónomo de Parques Nacionales. Binggeli. I. Morge. 2002. 321 p. Journal of Environmental Management 91 (2010):732–741.Spot fires : mechanisms analysis and modelling Final report 1998. Lampin-Maillet.. Mediterranean yearbook.. M. L. 2007.. and Delogu. Lampin.. Global Environmental Change 16: 293–303.. G. Coste. C. Morge. D’Ercole. Gonzalez-Olabarria. Organismo Autónomo de Parques Nacionales. Galiana. Summer Fires in the European Mediterranean – The Cases of Greece. Solana. resilience and adaptative capacity. 2001... Seville. Pp 243–247.. 289 p. 2002. C. and Ferrier. El ejemplo de Sierra Calderona (Comunidad Valenciana.. Lampin-Maillet. M.. Johnson. and context. Forest Ecology and Management 211: 73–82. Wildland fire hazard and risk: Problems. Method to characterize and to map wildland urban interfaces. E. C.A. C.. Lampin-Maillet... Lampin-Maillet.. G. Gaulier. Mantzavilas. Italy and Spain... 2006.C. Jappiot.. J..90 Towards Integrated Fire Management Colin. Caractérisation de la relation entre organisation spatiale d’un territoire et risque d’incendie : le cas des interfaces habitat-forêt du sud de la France. Représentation cartographiques des facteurs de vulnérabilité des populations exposées à une menace volcanique.C. and Solana. Apostolopoulou.. Lampin. C.. C. Dettori. J. Jappiot. Thèse de doctorat de l’Université de Provence. E.P. 2008. Yves Birot (ed. Ministerio de Medio Ambiente. D. G. M.. B. Countryside Agency and Scottish Natural Heritage.. N. Guijarro. 104 p... Long. Partozis. . T. 41–47. Pp. .. 2008... Seville.. C. 2009. Approches de la vulnérabilité et perspectives pour une meilleure logique de réduction du risque. Computers.. Morge. Economy and Territory. Ministerio de Medio Ambiente. Linkages between vulnerabilty. Galiana. Davis. 2009. ENV98-CT98-0701. 2007b. Topaloudis. J. Gallopin. Aix-Marseille I. G. 1990. J. J. Lampin-Maillet. M. 2005.. P. Pangea 29(39): 20-28.C. Med. Spain. P. Brigaglia. Bouillon. J. 1998. Hernando. Jappiot. 2007. Project no. F. M. C. G. and Ferrier. P. M. Martinez. C. definitions. Partie de Living with Wildfires: What science can tell us... 1996. Joensuu. D.P. M. and Delogu G.. A. T. Lampin. G. M. Mapping wildland-urban interfaces at large scales integrating housing density and vegetation aggregation for fire prevention in the South of France.. European Forest Institute. S. Long. Herrero.. J. Characterization and mapping of wildland-urban interfaces: a methodology applied in the study case in Sardinia. J.R. Hardy..2-1 of the Integrated Project “Fire Paradox”.. 285 p. Solana. D.

C. NE Portugal: calibration and application to fire management. J.A (eds). Canada.. 89 p.. C. Proceedings of the IV International Conference on Forest Fire Research. W. European journal of GIS and Spatial analysis.. Apostolopoulou... I. C. and Dudley. Ganteaume.. .. Sneeuwjagt. J. Conserv. K. Hammer R. C. I. 2005a.. Human-caused wildfire risk rating for prevention planning in Spain. Martinez. The Role of Remote Sensing Data. T. Proceedings of the IV International Conference on Forest Fire Research. Mapping. Mantzavelas. M. 21–61. C. M. L.. J. T. Gomez. M. Brewer. Mantzavelas. Morgan. Jappiot. 2008... Lampin. Vega-Garcia. Alibert. Deliverable D5. I. Carlson. and Jappiot. Lampin.. E. Long..S. Marzano. In: Chuvieco. Alexander. Method to assess with good spatial accuracy the meteorological and fuel moisture components of the fire risk. Karteris. Bonora. T. Gómez Fernán. T. P. Holcomb. I. C.. Morge. 2002. D.. N. Mantzavelas. England.. European Commission. and Giannakopoulos. C. T. Martinez. European Commission. Molina. L.C. Rural and suburban sprawl in the U.. Sussex. Topaloudis. S. I. G. C. The wildland urban interface in the United States. Curt. Project no.).2-2 of the Integrated Project “Fire Paradox”. Adapted method to assess and map fire hazard in wildland urban interfaces. Martin. Berolo.. Apostolopoulou.. A. Partozis.. M. FP6018505. D. Guide for end-users to characterize and to map wildland-urban interfaces. M. 2005b. Martinez. L. T.. K. T.. Portugal. M. T. European Commission. Deliverable D5. J. T. Lampin. European Commission. 18 p. M... D. San Miguel-Ayanz. J. D. Cartographie des interfaces habitat-forêt. M. 2009.. J. V. Bouillon.. and Guillande. Fox.1-1 of the Integrated Project “Fire Paradox”. John Wiley & Sons. Jappiot. 2004.... 91 Lampin. J. Deliverable D-08-05. Project no. FP6-018505. Lampin.). Conese.. Prototype of an Intensity Scale for the Natural Hazards: Forest Fire. Human influences on Forest Ecosystems: The Southern Wildland-Urban lnterface Assessment. Long. M.. P.1-4 of the Integrated Project “Fire Paradox”.I.. E. M.. F. Tolhurst. Defossé. and Fernandes.. T. Completion of recommended WUI fire mitigation measures within urban households in Edmonton. Landscape Pattern Metrics.. Galiana. Project no. G. Internal Report IR5.. R. E. Deshayes. 2003. T. 2005. M. and Solana. A. Project no. Aguado. Macie. 2002.. and Topaloudis. Jappiot. Long. McGee.I. Midwest from 1940 to 2000 and its relations to forest fragmentation. P. Radeloff. J.F. I. (ed. Manlay.. 19 p.L... Volume 2. E. L. A. Lazaridou. Mallinis. M.. Une approche spatiale pour estimer le risque d’incendie de forêt. Environmental Hazards 6: 147–157. M. Luso. Ecological applications 15(3): 799–805. Deliverable D5. Lampin. G. V. Gitas. E... D. 2009..S.. Lampin. FP6-018505.. Jappiot.. FP6-018505. Singapore. T.. Fried.. Mantzavelas. Topaloudis. 13p.. C. In Viegas.. Biol. Common methods for mapping the wildland fire danger. Rainha. 2002.Wildland Urban Interfaces. Stewart.B. G. Pp.. 19(3): 793–805. and Chuvieco. R. Herrero. 2006.A. 2009. Bovio. and Lencinas Daniel. K. J. S. R.. G. (ed. Journal of Environmental Management 90(2): 1241–1252.M.D. Salas. and Stewart...B. and Borgniet.. World Scientific Publishing.. and Hermansen.. Portugal. Lazaridou.. 2002. 2007c. Apostolopoulou. Dauriac. Guglielmet.. A. Partozis. Apostolopoulou. L. M. EUFIRELAB.)..2-5 of the Integrated Project “Fire Paradox”.. C. Lazaridou.. Using the Canadian Fire Weather Index (FWI) in the Natural Park of Montesinho. Current Methods to Assess Fire Hazard Potential. Wildland Fire Hazard Estimation and Mapping. C. Information géographique et gestion des risques l16(3–4): 320–340.. T. Literature review and synopsis of the methodologies to estimate vulnerability.. Lazaridou. (ed. Borgniet.C. V..F. Chapitre du livre Encyclopedia of Environmentrics. Carrega. Bouillon. and McKeefry. Hammer R. Revue internationale de Géomatique.. S. Partozis. Partozis. R. M. I. In Viegas. Fire Behaviour and Vulnerability: Characterization... Luso.. S. A. Radeloff.. Jappiot.. McGarigal.

D. Comparative study of various methods of fire danger evaluation in Southern Europe. Development and Structure of the Canadian Forest Fire Weather Index System. and Hammer. C. Measuring and incorporating vulnerability into conservation planning. H. and Ferreira.. Nosenzo. A. Proceedings of the 2nd International Wildland Fire Ecology and Fire Management Workshop.... International Journal of Wildland Fire 16: 306–316. M. Sol. Bovio.. Burgman.W.. Pp. S. L. Integrating new methods and tools in fire danger rating. Coimbra. Z.. Florida. November 16–20. A. Wilson.. Newton. Science in China: Series E Technological Sciences 49(I): 140–149. 37 p. H. Xu. D. Tech. Macie.92 Towards Integrated Fire Management Stewart. Orlando.) 2005.S. Hao. Proceedings of II International Conference on Forest fire Research. L. M. V. S. 2006.. G. Possingham. Kallos. Serv. Pressey. Vasilakos..). Duryea. Portugal. Guofan. G.. Hatzopoulos. Dai. (ed.. Characteristics and location of the Wildland-Urban-Interface in the United States. Viegas. Mapping forest fires risk zones with spatial data and principal components analysis. (eds.L. . Forests at the wildland-urban interface: conservation and management.C. and Hermansen. Urban wildland interface communities within vicinity of Federal lands that are at high risk from wildfire. For. 6. A. B. U. 2001. Department of Interior. Environmental Management 35(5): 527–543. 2005. In: Viegas.A. D. For. 66: 751–777. 35. p. R. E.L.. Rep. Boca Raton.B. Kalabokidis. L. 571–590. Vince... 1994..A. Fed. E. 1987.X. K. and Weston. CRC Press.. Radeloff.. 2003.I. USDA-USDI.Reg. Van Wagner. 2007. and Wang. C. Tang. Ottawa. S. K. R. Y. Department of Agriculture and U. C. Can. J. 2003.S. and Matsinos.

and the civil society. Portugal and Spain). These impacts can be reflected in many ways – for example. The growing importance of wildfire issues at EU level is also reflected in the increasing number of research projects funded to better understand and address this problem (e. Eufirelab. Italy. the loss of human lives (McCaffrey 2008). Fire Paradox). this change was galvanized by the large-scale wildfires and their consequences in the Mediterranean region during the 2000s. an average 500 000 ha of forests are burned annually (San-Miguel and Camia 2009). 120 000 ha affected. The last decades have seen increased social awareness and growing concerns for wildfires’ negative environmental and economic consequences. These included. In Europe.e. wildfires continue to be a major environmental threat (Requardt et al. and causing significant human casualties (González-Cabán 2007). Even worse. infrastructure networks (i. Australia 2007. In Europe. but have also a significant social dimension. particularly. For example. Wildfires affect the forests and other wooded land. decreased well-being of the population (local and wider). where 8 million ha burned. national and global scales. Greece. Armando González Cabán2 and Verónica Farreras1 1 Forest Technology Centre of Catalonia. and temporary or permanent loss of employment possibilities and economic activities. For a worldwide perspective on the effects of fire on the earth systems see Bowman et al. Pacific Southwest Research Station. and neighbouring systems such as urban areas. affects millions of people worldwide.g. 2007) where. USA 3. during 2003 in Portugal about 400 000 ha of forests lands were burned. the large fires affecting Greece during the summer of 2007 caused 64 casualties. and during 2007 in Greece around 270 000 ha of forests and other wooded lands were destroyed (JRC 2007). the wildfires affecting vast forest areas of Portugal in 2005 caused economic damage worth almost €800 million and caused 13 fatalities. loss of human life or health.4.3. 2009. and. agriculture lands.1 Introduction Wildfires are a societal problem that threatens many ecosystems. and causes major ecosystem and economic impacts at local regional. Riverside (CA). Events in other parts of the world also helped to increase the concern on this topic in Europe.4 The Importance of Economics in Fire Management Programmes Analysis Robert Mavsar1. USA 2007. Indonesia 1997/1998. These major wildfire events clearly showed that they are not only an environmental problem. power-lines and transportation corridors). and California. For example. Spain 2 USDA Forest Service. more than 1 million ha. and especially in the Mediterranean countries (France. affecting millions of people. during 2005 in Spain some 190 000 ha of forests lands were damaged. Barcelona. having major economic impacts. and according to the Greek authorities the economic .

To answer this question. we look in detail at the problems and a possible solution for the adequate estimation of costs related to wildfires. economic models can help to estimate whether investments in wildfire related measures (e. the economic analysis of the efficiency of fire management is generally evaluated by the cost plus net value change concept (C+NVC) (Althaus and Mills 1982. These include prevention (e. Petsini Arlapanou and Petsini Arlapanou 2007). equipment. Donovan and Rideout 2003. However. One of the thorniest questions in fire management is to determine how the limited financial. The understanding and assessment of socio-economic impacts of wildfires should be considered an essential part of the fire risk assessment. For the implementation of these decision support systems reliable knowledge and data are needed on the physical and economic impacts (negative and positive) of wildfires and the economic efficiency of fire management measures. These events increase the public interest and concern about the wildfires and prompted the development and implementation of improved policies and management measures at different levels.g. 3. Decision support systems based on economic models can help to decide what would be the optimal use of the resources. development of wildfire related policies. suppression. fire fighting and amount of wildfire acceptable). publicity campaigns). which must be justifiable and efficient.2 Estimating the economic efficiency of and social preferences for fire management programmes 3. González-Cabán . fuel treatment (e.94 Towards Integrated Fire Management damage was estimated at €2–5 billion (Papachristou 2007. the implementation of such measures requires substantial investment of financial.g. and human resources should be most efficiently spent and distributed among alternative fire management options. presuppression and suppression. education.g.4. mechanical fuel removal). fuel management) are financially justified.4. This chapter will first introduce the basic model used for the estimation of the economic efficiency of fire management programmes and revise some of the problems related with its application. as well as planning and implementation of management practices (Morton et al. For example. In the subsequent section we discuss why the social preferences should be considered when planning fire management programmes and present a study dealing with this topic that was conducted in Spain. economic and social impacts of wildfires (EU 2005). the combination of investments in fire prevention. 2003). and restoration measures. The main objective of these improved policies and management measures is to minimize the negative environmental. human and organizational resources. prevention. In section three of this chapter.1 Economic efficiency of fire management programmes To combat the wildfires problem forest managers can apply different management measures. The last part of the chapter presents the main conclusions.2. or to choose the most efficient amongst several alternatives (i.e. prescribed burning. thinning.

wages for fire fighting crews. fuel for aerial resources). Nevertheless. The model is derived by adding pre-suppression costs to the costs of fire suppression and the net value of the change in resources outputs that result from the fires (difference between benefits and damage to the resource). Thus. the objective of the C+NVC model is to find the most efficient management levels (González-Cabán et al. or to maximize the social benefits (Donovan and Rideout 2003. In Figure 1 the slope of the level curve is equal to the negative of the ratio of the marginal contributions of suppression and pre-suppression in reducing damage. both inputs are allowed to vary independently. does not determine how frequently this equipment will be used during the fire season. where we intend either to minimize the costs and the net value change. Selecting the most efficient management level is an optimization problem. NVC level curve. Suppression costs are fire fighting expenditures during a fire season (e. that pre-suppression can be substituted by suppression. It should be noted. equipment for fire fighting crews). 2007). 1986. This approach estimates all costs and benefits related to fire management measures.g. The independence of pre-suppression and suppression expenditures means that one input does not determine the level of the other. Mercer et al. 2007). As mentioned at the beginning of this chapter. purchase of fire fighting resources such as fire engines. but remain related through the NVC function (Donovan and Rideout 2003). the pre-suppression may affect the optimal level of suppression through the NVC function (Figure 1). . In the C+NVC model pre-suppression and suppression expenditures are considered as independent inputs. For example. Lankoande 2005).The Importance of Economics in Fire Management Programmes Analysis 95 Figure 1. but only in the case that NVC is held constant (Donovan and Rideout 2003). related through the Net Value Change (NVC) function (Donovan and Rideout 2003).g. the purchase of some fire fighting equipment before the fire season. The pre-suppression costs represent the expenditure on fire management prior to the wildfire season (e.

2008). In the model. their most efficient and cost-effective allocation must be determined (Wei et al. while the net value change (NVC) sums all the changes in the quantity/quality of resources outputs (goods and services) that result from the fire multiplied by the unit value of the output.g. To be able to apply the C+NVC model and to include it into a decision support system (DSS) for planning and evaluation of the economic efficiency of management programmes it requires adequate information. fire detection facilities. but also to decide how to distribute the resources (spatial component). However. and assessments of the risk associated with these values. given that the pre-suppression expenses are at an optimal level (Figure 2). the most efficient level of fire programme is the point where the summation of costs and the net value change is minimized (i.96 Towards Integrated Fire Management Figure 2. For example. It requires information about the direct and indirect effects of fire on spatial and temporal provision of goods and services. suppression resources. once agencies have decided how many resources to use (e. the cost part (C) is obtained by the estimation of the programme costs (pre-suppression and suppression activities). The objective of the evaluation of economic efficiency of fire management measures is not only to help identify the most efficient programme level. and information about how fire-induced marginal changes in the quality and quantity of goods and services will affect social welfare. It is apparent that information is . One of the difficulties related to the application of the C+NVC model is the estimation of the NVC component. quantitative estimates of effects on resource outputs. Illustration of the C+NVC model (Donovan and Rideout 2003). there are still considerable shortcomings in the knowledge and data availability. for each evaluated fire management program alternative it would at least need information on: monetary estimates of economic efficiency.e. For example. P* in Figure 2). Namely. fuel treatments).

improve accessibility. improve forage and grazing. manage invasive species. meaning that present actions can influence decisions on applications of fire management measures in future fire seasons. and available only for some countries or regions within countries. and maybe the most fundamental. According to Mercer et al. unreliable. wildfires significantly reduce the amount of fuels (similar to the effect of fuel management). forest lands). improve wildlife habitat. ‘extended’. Knowing the social preferences for different management actions and identifying when they might differ from the manager’s viewpoint can help agencies to understand and predict how different audiences may react to these decisions. Another important issue to consider when estimating costs and benefits of fire management is that multiple objectives can be accomplished by these measures. In addition. enhance aesthetics and recreation activities. prepare sites for seeding and planting. and therefore the available data is most often incomplete. 2008). For example. For instance. reduced fire risk) also in subsequent seasons. (Mavsar et al. 2007. At the European level. it is important to consider other factors. More on this issue will be presented in section 3. Lagged benefits are not only a characteristic of management measures. dispose of logging debris. 2007). there is no common methodology to collect and report such data. and (ii) help to develop 1 For an interesting discussion on how public perceptions affect the implmentation of fire management programmes see Laband et al. and manage endangered species (Wade and Lunsford 1988). they notably affect the fire risk and intensity in the following seasons. but also of wildfires. as well as the costs and expenses of fire management and suppression activities. .2. (2007) showed that fuel management treatments applied in the current season can produce benefits (e.4. However. These. (2008).3 of this chapter. but the social value of different goods and services. thus.g. is the general lack of reliable data not only on the effects of fires on the quantity/quality of goods and services produced by the natural resources (in our case. (2006) and Laband et al. In many cases. (2007) this reduced risk of wildfires can last up to 11 years. Mercer et al. benefits should also be taken into account when estimating the costs and benefits of fire management measures. control insects and diseases. Namely. it should not be the only attribute considered when deciding about fire policies or management measures. For example. this knowledge may (i) help the administration recognize when policies or actions may be supported or opposed by the public1. prescribed burning can reduce fuels. Martin et al. The estimation of the costs and the NVC is further complicated by the possible temporal interrelation of fire events or management measures. for example social values and preferences for a certain programme or management measure (Daniel et al.2 Social preferences for fire management programmes Economic efficiency is an important criterion in deciding which fire management programme would be the most efficient.4. 3.The Importance of Economics in Fire Management Programmes Analysis 97 lacking in one or both of these areas for many of the types of goods and services that can be affected by wildfire (Venn and Calkin 2007). another important problem.

Therefore. there are goods and services (e. However.3 Estimating the costs of wildfires A precondition for the estimation of the socio-economic damages caused by wildfires is to understand what the important factors are and how they influence the quantity and quality of goods and services provided by a natural resource (e. like prevention. 2009). when existing. It can serve as a support tool in the design of policies and/or specific fire management programmes. However. managing endangered species and fire dependent species (Wade and Lunsford 1988). In contrast. Dale 2009. and mushrooms) that reflects their value. Dunn et al.4. it is important to understand that in contrast to forest goods and services that have a market price (e. it should be acknowledged that wildfires can have positive effects (González-Cabán 2007). fruits. or.g. Thus. water purification) that are not . Direct costs are those which are directly related to a wildfire event. These positive effects (benefits). and incurred as a result of the fire and/or exposure to the fire. land managers may want to design a fire prevention measure that mirrors society’s preferences – for example. and other costs associated to the individuals who suffer a loss of benefits as a consequence of a wildfire. social preferences may allow policy makers to better identify priority-attention areas in fire management. for fire behaviour with a lower impact on forest resources. nowadays estimating the costs of wildfires must also include these goods and services (Butry et al. Thus. and direct suppression costs. recreation activities. In general. A wide range of costs can be related to a wildfire. mainly the decreased quantity/quality of timber. improving the wildlife habitat. biodiversity.98 Towards Integrated Fire Management information or education campaigns. For example. improving understory forage and grazing. 3. Zybach et al. monitoring and pre-suppression costs. An important concern in evaluation of fire management efficiency and estimation of wildfire costs is the proper consideration of environmental (forest) goods and services that might have been damaged (or enhanced) as a consequence of a wildfire. such as losses or damages of environmental (forest) goods and services. timber. 2000). 2003).g. which might help to get public support for a certain policy or action. restoration costs. must also be included into the estimation methodology. 2001. only a few of them have traditionally been considered. in the last decades it has become obvious that forests are also important as providers of environmental and social goods and services (Farrell et al. indirect costs are related to the risk of wildfire occurrence or as a response to the occurrence of a wildfire. Further.g. Only a complete picture of the damages enables a reliable estimation of the socio-economic impacts of wildfires. erosion prevention. This knowledge is vital for the identification of goods and services that can be or are damaged by a fire. In estimating the economic impacts of wildfires. learning about the social preferences could be useful in different ways. This is seldom the case in Europe and there are only a couple of case studies where social preferences for fire management measures have been elicited (see Box 1). if faced with a fixed budget for wildfire mitigation. and there are different ways to categorize them (Ashe et al. we can divide costs into direct and indirect. 2008. property. forest).

Bennett and Blamey 2001). 2008). The label ‘choice experiment’ refers to a survey-based valuation method that simulates actual market behaviour (Hanemann and Kanninen 1999. In a choice experiment.The Importance of Economics in Fire Management Programmes Analysis 99 Box 1: Case study on social preferences regarding fire management measures2 Burned forest area or dead trees? A choice for Catalan citizens The objective of the study. results show that the attribute of highest concern is fire propagation (area burned). The final part was designed to collect some socio-economic data about the respondents. social preferences should be considered in the decision making process. population size. We applied the choice experiment methodology. For example. However. Gerona and Lérida – in June 2007.g. A plausible explanation for this may be the publics’ familiarity with information on the quantity of area burned. Thus. can be described in terms of its attributes. In our empirical application. accessibility). we have no information about their value (González-Cabán 1998. Even though these methods have improved considerably in the last decades. the design of fire prevention and protection programmes also depends on other factors such as vegetation type and characteristics. conducted in Catalonia (NE Spain). The result implies that from a social perspective reducing the forest area burned is one of the most relevant elements when designing fire prevention and protection programmes. 207 interviews were conducted in three Catalan provinces – Barcelona. The second part contained the choice experiment exercise and a number of debriefing questions. This technique is based on the idea that any alternative. they still have limitations. available funds. 2001. Bateman 2002). This limitation is especially important when developing a system for the assessment 2 For detailed information see Mavsar and Farreras (2008). respondents are presented with a series of choice sets comprising at least two alternatives and are asked to choose which alternative they prefer (Hanley et al. complexity of the fire problem. One of the most important limitations is that the estimated values cannot be easily extrapolated. but might also be influenced by other factors (e. as well as the payment mechanism and the consequences of each choice. Further. income in the region. as this is often used by the media in Spain to quantify the consequences and the severity of a wildfire. available experience and expertise. which should be considered when used. or characteristics. . The study results suggest that additional fire prevention measures increase the welfare of the Catalan population (similar results were reported in Riera and Mogas (2004). Mavsar et al. or good. The first part of the interview presented the attributes to be valued. recreation values in different forests are probably different and this difference is not necessarily only the result of different site characteristics. but are not the only factor influencing the design of fire management programmes. was to elicit the social preferences regarding fire prevention measures in terms of their impact on fire behaviour – fire propagation and intensity – and to estimate the value of these measures for the society. Non-market valuation methods can be applied to value these goods and services. traded in traditional markets (hereafter non-market goods and services) and thus.

Once this is accomplished. Thus. the obtained values are site specific and transferable to other sites only under some specific conditions (e. assessing the costs of wildfires) and to find adequate solutions (i. it is important for decision makers to start considering economic analysis as an integral part of a proactive fire management. 3. 2008).100 Towards Integrated Fire Management of socio-economic impacts of wildfires.g. given the complexity and high cost in measuring them. first. development of a decision support system (DSS) requires.4 Conclusions The extreme wildfire events of the past decade have shown that fires are not only an ecological. evaluate efficiency of fire management measures and provide information on public preferences). similar forest site and population characteristics). Therefore. the next step is to develop models that would adequately simulate the behaviour and impact of wildfires. In this case.e. • the impact of fire management measures on risk. benefit transfer method) or employing proxies (e. C+NVC model). historical fire data. a tool that provides information that can help managers make better informed decisions that can lead to selection of the most efficient alternatives in a given situation. Nevertheless. availability of fire . important problems still remain to be solved in the future. in addition.g. 2009). answers to these three concerns. In this respect. because the fire may spread over areas where no values for non-market goods and services have been estimated. in most cases only fire suppression costs and loss of timber production are considered. the main issues are the inadequate understanding of: • the impacts of wildfires on the spatial and temporal provision of goods and services (e.g. This also implies that the role of economics in the fire management process should change. what is the value of the losses). and the potential economic impacts (positive and negative). Furthermore.g. with the use of the right data. Economics has an important role in helping to quantify the magnitude of the problem (i. fuel models. quantify the effects of different management measures).g.g.4. geographic data.g.g. while damage to forest non-market goods and services are often omitted or very limited (Pettenella et al. non-market valuation methods) and models (e. how the quality and quantity of a good or service is affected and for how long). • the potential effect of the changes caused by wildfires on society’s wellbeing (e. restoration costs) this problem could potentially be overcome (for example see Rideout et al. simply applying values from another site may generate significant estimation errors. as mentioned earlier. Nevertheless. The absence of non-market values is somehow expected. weather parameters. but even more a socio-economic problem. development and implementation of a DSS necessitates that a significant amount of information be collected and processed (e. extent and severity of wildfires (e. the effects of fire management measures. adequate techniques (e.g. Although there has been considerable development of economic methods (e.e.

Butry.E.. Edward Elgar. Doyle.. Dale.. Bowman. A decision maker must still weigh and assimilate a large number of relevant factors. 1982. Total Cost of Fire in Australia.J. S. their costs must always be considered (Mills and Bratten 1982). Cochrane.S.. and while many of those constraints are valid. L. 2002.J.A. A DSS could show some of the costs of imposing institutional constraints on the programme. standardized procedures must be established on how to collect and process the needed data.. USDA Forest Service. What is the price of catastrophic wildfire? Journal of Forestry 99: 9–17.T. 2008. which are relevant to fire management program planning and helps trace the interaction of relationships that are too complex and numerous for a person to easily follow. Harrison. C. and Holmes. DeFries. fire management policies). McAneney.P.A.W.M. Science 324: 481–484. there is a need not only to develop a relevant DSS. P. M. D..g. Lakewood. Prestemon. values of market and non-market goods and services). and the costs of fire management activities. J. In the case of a fire economic efficiency DSS. 16 p. Economic valuation with stated preference techniques : a manual. and Blamey. Finally. Proceedings of the III International Symposium on Fire Economics. the final decision on what kind. J. W. R.M.P. Considering these limitations. CO. Bateman. D. B. Pye. Bond. I. J. E. but furthermore to ensure that adequate data is collected in a standardized form.J. T.. and others 2009. References Althaus. Balch. Western Forestry Leadership Coalition. R.. The choice modelling approach to environmental evaluation. D'Antonio.S. D. and Mills. Resource values in analyzing fire management programs for economic efficiency. UK. After careful consideration of all the information provided by the DSS. . many of which cannot be measured in common units or even be measured quantitatively at all. Pp. this would provide the decision maker only with information on how to possibly design the most economically efficient fire management programme. it is important to recognize that no decision support system or computer simulation model is a substitute for a decision maker. T...J.S.M. Mercer. A. the social value of affected goods and services. Fire in the Earth System. 9 p. A DSS simply quantifies some of the factors.. 2001.The Importance of Economics in Fire Management Programmes Analysis 101 management practices and resources. 2009... and to what extent the fire programme will be implemented rest squarely on the decision maker’s shoulders. Bennett. this will change for the best in the near future. K. Cheltenham. The True Cost of Wildfire in the Western U.S. J. Hopefully. the results depend on the model assumptions and restrictions (e. and Pitman. Cheltenham.M. and must place fire programs within the context of other management programs and institutional constraints (Mills and Bratten 1982). 20. with the potential to deal with territorial differences. Ashe. In addition. There are still many obstacles to overcome on the path to the development and application of a model for the evaluation of economic efficiency of fire management and protection programs at the European level. J.K. and other sources. 2001. Carolina. However.J. I. One of the main problems is the lack of reliable and comparable data on fire effects. J.P. Elgar Pub. Planning and Policy: Common Problems and Approaches. Artaxo.C. Carlson. PR.

Shinkle.H. Department of Agriculture.). I. Seville.G. Spain. Council Regulation (EC) No 1698/2005 of 20 September 2005 on support for rural development by the European Agricultural Fund for Rural Development (EAFRD). and Padua Wildfires: How much did these fires really cost? USDA Forest Service.M. Corvallis.. European Commission. 2007. EC Joint Research Centre. 137 p. D. E. and developing countries. and Piwnicki. Dunn.. M. Study Contract No: 30-CE-0162979/00-21..S. DC.S. Project no. 302–442. González-Cabán. Andersson. Mourato. K. Choice Modelling Approaches: A Superior Alternative for Environmental Valuatioin? Journal of Economic Surveys 15: 435–462. Ryan. Tykkä. fire and forests : a synthesis of wildfire social science. USDA Forest Service.. Huttl.. and P. Laband. C.J. Vreke. Deliverable D4. M. D. The statistical analysis of discrete-response CV data. 18 p.M. and Solari.N.. Carroll. B. D. N. D. (eds. Washington. and Hussain. A. EC DG AGRI. Laband.. C. Three essays on wildfire economics and policy [Thesis (Ph D )].. Valuing environmental preferences: theory and practice of the contingent valuation method in the US. Understanding Public Perspectives of Wildfire Risk.E. 2003. R. 2007. G. 2008. Washington State University. 2008.E.. FP6-018505.. (eds. 2008. Washington. P. Forest Science 49: 331–335. Raish. Wildfire risk: human perceptions and management implications. 2007. Rametsteiner. Project no. Pacific Southwest Research Station. C. Japelj.. Stn.J. T.. and Rideout. Forest Service. E. JRC. A. Hanley. Oxford [England]. R. Martin.C. W. 2006. M. B. Forest Policy and Economics 10: 444–449. Mavsar. B.4-1-2 of the Integrated Project “Fire Paradox”. S. The Old. 2008. Kovac. McCaffrey.P. Mavsar. Pacific Southwest Forest and Range Exp. I. T. A. Study on the Development and Marketing of Non-Market Forest Products and Services. J. Lankoande. Oxford University Press. EU 2005.. Prokofieva. 2001. Database of market values and on non-market values. FP6-018505.. Wildland Fire Management Policy and Fire Management Economic Efficiency in the USDA Forest Service. 2003.N. Forest Science 52: 477–488. Raish. In: Martin. and Kent.. Deliverable D4. V. 1986. and Farreras. A.. EU.. Developing fire management mixes for fire program planning. 2008. 48 p. Hanemann..E. 1999. Weiss. Donovan. van Apeldoorn. R. The impact of Forest Service litigation success on administrative appeals of proposed fuels reduction actions. González-Cabán. European Commission. D. Aspectos Económicos de la Evaluación del Danõ de Incendios.B. M. and Raish. Forest Fires in Europe 2006. and González-Cabán. Factors that Influence Administrative Appeals of Proposed USDA Forest Service Fuels Reduction Actions. Database on perception indicators. U. Moseley. and Mills. DC.102 Towards Integrated Fire Management Daniel. S. Palahí. Farrell. and Kent.D.. Wildfire Risk: human perceptions and management implications Resources for the Future. and Kanninen. González Cabán.. 2000. Resources for the Future. Pp. R. and Willis.). Institute for Environment and Sustainability. Hussain. Forest Ecology and Management 132: 5–20. and Wright. S. Fuhrer. 2005. Farreras. Pacific Southwest Research Station. People.. An Integer Programming Model to Optimize Resource Allocation for Wildfire Containment. Proceedings of the Wildfire 2007-4th International Wildland Fire Conference. G. W. González-Cabán. A. August 2005. W. V. R. Mavsar. and others. 2007. Grand Prix. J. K. . González-Cabán.. E. Pp. C. A.. Offical Journal L: 1–40. A.. R. European forest ecosystems: building the future on the legacy of the past. In: Bateman. F.B. P.4-3 of the Integrated Project "Fire Paradox". E. Kraus. A.. 11-22. 56 p. S. A.M... New York. A. Oregon State University Press. 1998. Ramčilović. M.. 8 p. van Wijk.

B. Jeffrey. 2008. Requardt. Forest Service. Estimating rates of substitution for protecting values at risk for initial attack planning and budgeting. J. http: //www. Southern Region. Rideout. Camp. Portland.. Barcelona. Social.E. 2009. Dubrasich. Forest Policy and Economics 6: 521–528. Brenner. and Kirsch.P. David. Zybach. I. Feasibility Study on means of combating forest dieback in the European Union.E. Forest Fires at a Glance: Facts. S. A. School of Forestry & Environmental Studies . Evaluation of a risk reduction in forest fires in a Mediterranean region..J. 2004.W. 2003.. Janse. and Camia. T.P. and Calkin. R. B.E. Challenges of accommodating non-market values in evaluation of wildfire suppression in the United States. USDA Forest Service. Yale University. Forest Policy and Economics 10: 205–219. European Forest Institute. Y. D.C. P. 2007. . Figures and Trends in the EU. Minister Says.D. In: Birot.. A. and Petsini Arlapanou. FEES: designg of a Fire Economics Evaluation System. Ziesler. A. and González-Cabán. Pacific Southwest Forest and Range Experiment Station. EC DG Environment. American Journal of Agricultural Economics 89: 63–77.S..B.. 1982.. Review of the methodologies for the assessment of economic impacts of forest fires and economic efficiency of fire management. Mavsar. Wildfire Cost-Plus-Loss Economics Project: The “One-Pager” Checklist. M. 2009. T. San-Miguel.E.. Pp. D.. J. M. Y. D. R. Köhl. R.. 108 p. An optimization model for locating fuel treatments across a landscape to reduce expected fire losses. and Marker. M. and Economic Impacts of Wildfire. D. EFIMED. Schuck. 2008. D. Assessing the Environmental. Poker. Greek August Wildfire Damage Seen at €2 Billion. Kling.cy/nqcontent..Global Institute of Sustainable Forestry. L..com. A. J. Riera.. Mavsar. Wei. A. P. 2009.S. 2007.bloomberg. and Botti. 5 billion Euro damages. 2007. Ressing. J. and Lunsford. 79 p. Morton. M. and Tyrrel. Evaluating Alternative Prescribed Burning Policies to Reduce Net Economic Damages from Wildfire. G.L. Rideout.. Loomis. U. 26 p. OR. G. 1988.. and John.. J. and Päivinen. and Bratten.com/apps/news?pid=20601085&sid=ax1IYpsKfGmQ &refer=europe Petsini Arlapanou. Department of Agriculture. T. Venn. D. D. J. Canadia Journal of Forest Research 38: 868–877. (ed. 2007. http: //www. 20 p. 2007. Living with Wildfires: What Science Can Tell Us? EFI Discussion Paper 15. A.D. H. Mills.J. U. Bloomberg.cfm?a_name=news_view&ann_id=74961 Pettenella. R. D.. Papachristou. Goio.The Importance of Economics in Fire Management Programmes Analysis 103 Mercer. and Mogas. F. A Guide for Prescribed Fire in Southern Forests. Proceedings of the American Agricultural Economics Association Annual Meeting. A. John Wildland Fire Lessons Learned Center. M. Florian. stockwatch. P.B.. Wade. Research Paper.S. 56 p.J.). 11–20.

.

Daniel García-Marco3. 1. Spain 4 GAUF. What should be the required training (and education)? 3. Spain 3 UNAP. Firing Operations Leader (FOL). and António Salgueiro4 1 Unit of Forest Fires. 3. Guadalajara. the last decades have been characterized by drastic changes in land use. Understanding and anticipating fire behaviour is also a critical part in managing fire as a tool. Three aspects of the WFA position should be addressed. with fast changes in fire behaviour. Fire analysis to plan fire fighting tactics in advance (pro-active way) instead of simply reacting to fire (re-acting way). Specifically. a Wildland Fire Propagation Expert. in Europe dramatic fire events are characterized by being often beyond the capacity of suppression in large areas. The wise use of fire as a tool to fight high intensity fires. there is the need of three new Fire Specialists. Lisbon. University of Lleida. the task of managing a wildfire is more complex than before and requires well trained managers who have a good understanding of fire propagation. or in more detail. Portugal 3. National Forest Authority. Firing Specialist (FS). 2. 2.5. How relevant is this job task in each region for this job? 2. offering very few and localized opportunities of extinction. thus taking advantage of the existing opportunities. In these large fires. Which kind of experiences can be transferred from some countries (or regions) to others? . two critical tools are required to improve the efficiency in the complex task of managing a wildfire. combined with forest extension. including the wildland urban interface (WUI). Barcelona.INFOCAM. Marc Castellnou2.3. requiring a higher level of knowledge in wildfire propagation: 1. and principally in the Mediterranean countries. 1. Department of Internal Affairs.1 Introduction In numerous areas of Europe. Wildland Fire Analyst (WFA). especially grazing pattern and land abandonment. Spain 2 GRAF. Here. Therefore.5 Improving Fire Management Success through Fire Behaviour Specialists Domingo Molina1. These changes have caused a more dramatic wildfire spread than in previous decades.

To forecast fire behaviour changes in both time and space and to determine critical points or lines where a greater alignment of factors (i. FS) include: 1. It should be noted that the approach toward the WFA position is different in Europe than in the USA.e. slope. Additionally.2. Castilla-La-Mancha. 2009). This forecast has to be supported by tools like the Campbell Prediction System Language (CPSL. FOL.106 Towards Integrated Fire Management In relation to the second aspect above. the tasks of the Fire Specialist group (WFA. Aragon) and beyond (i. 2. fuel model. the propagation patterns of large wildfires. These large wildfire propagation patterns have worked well in Catalonia (Spain) and have been used later with success in other regions of Spain (i. Then. 2008. To help plan the suppression action to be carried out. it could also be appropriately applied for the densely populated Mediterranean Coast in Europe. in Europe. Canary Islands. .5.1 Tasks As experts on fire propagation. and sharing of existing ones has been promoted. aspect) might put the fire behaviour out of suppression capabilities. they assign priorities on which sectors of the fire perimeter should be contained first and which should be addressed later to diminish the potential fire damage.e. Campbell 1995) and the large wildfires propagation patterns (Castellnou et al.2 The Fire Specialist Group 3. most often the main fire runs occur in a matter of a few hours to a few days and often affect housing areas.e. but also to identify opportunities in the anticipated fire spread. helps to identity best management action in pre-attack or fuel management. CPSL has been developed in southern California. Grillo et al. and to employ tactics that take advantage of these opportunities. Portugal. significant new training materials (i. were recently produced in Europe. specifically aimed for such training. Therefore. This is implemented by means of identifying which fire perimeter sectors are above the fire suppression capabilities (by which tool. Therefore. 2009). This chapter is about the need for a comprehensive knowledge on fire propagation in support of fire management actions and the possibility to introduce some new job positions of fire behaviour specialists. they set up a hierarchical classification of opportunities to contain the fire perimeter. where there are fast spreading large wildfires that affect housing developments in a very short time. 3.5. Molina et al. On the other hand.e. at any given position in time and space). In Europe. Italy). A major reason is that large wildfires in the USA last longer than in Europe and maybe far from towns. wind. These training documents are an innovative contribution to introducing changes into regional and national forest and fire management services across Europe. they spot safe and efficient opportunities in time and space. suppression agencies not only need to forecast wildfire behaviour changes. developed after detailed study of past wildfires. CPSL provides for a qualitative analysis that includes a language to communicate estimated fire behaviour changes.

but also in tactics and in strategy. Forest Service. Also.e. Finally.Improving Fire Management Success through Fire Behaviour Specialists 107 3. an important role has to be played by a careful study of past fire propagation patterns and suppression actions (Molina et al. To plan and perform backfires and burn out operations. recommending tactics and strategies. . hand tools. the FS participates and often leads operations merging diverse suppression tools (i.e. aspect) might put a given fire front out of suppression capabilities. A FOL may also train other fire fighters in advanced fire analysis and control and supervision of operations to ensure that both safety and efficiency would be accomplished. Among the duties assigned are: 1.e. To use fire as a tool. To train workers in safety and efficient work.2 What to expect from a Wildland Fire Analyst? A WFA is an expert in both fire spread patterns and fire behaviour forecast that is employed (ideally all year round) by the agency in charge of wildfire control (i. time. and therefore to make an assessment on the best strategy that will take advantage of every opportunity to weaken the fire head. The FS is the one who decides how a fire operation should be performed in order to fulfill objectives. Personnel safety must always be a priority. fire sectors). 5. fuel model. The FOL can be in charge of any operation involving the wise use of fire.5.e. Second in line. and in wildfires. to incorporate lessons learned specially in fire behaviour. Therefore. In this training.2. this firing specialist should forecast fire behaviour changes in time and space and determine critical points or lines where a greater alignment of factors (i. technical fire. To lead operations combining different tools (i. A firing specialist may train other fire fighters in basic safety procedures. the FOL helps in selecting opportunities associated to changes of fire behaviour. Safety of personnel must be always a priority. Emergency & Fire Fighting Service). To elaborate an advanced forecast of fire behaviour potential in active fires. The FOL may also lead operations merging different suppression tools working in a given fire sector. Each of the Fire Specialist group tasks is performed by different specialists at different scales (i. slope. and to confine or extinguish the fire. The WFA is a key specialist who facilitates the wise use of fire as a tool (both in fuel management when there is no wildfire and in actual fire suppression actions). to the training and response of the whole organization. the FS has to identify and select opportunities to contain the fire perimeter and propose tactics to implement the control. fire control efficiency has to be re-assessed often. aerial means. 2. Also. basic fire analysis and merging fire suppression tools for optimal results. 2009) to allow for a secure and efficient fire control. 3. Later. wind. the fire behaviour specialists must train workers in basic fire analysis and in job hazard abatement actions all year round. in prescribed burning in rangelands or understory burning.e. To assess or be in charge of prescribed burning actions. hoses). To elaborate full reviews of major past and present fires. hand tools and fire engines). 4.

4. we have the LACES protocol (Lookout. there is a need to plan for any eventual misinterpretation or failed forecast (or failure of adequate accomplishment of the tactic by fire workers). 5. the assessment and supervision of any proposed suppression action (in terms of efficiency and personnel safety). To assist the Incident Commander to set up the best fire control strategy for a given fire (based on specific wildfire propagation patterns or fire typologies (Castellnou et al. For this. In this training. The WFA position fills a gap that it is too often present when setting up a Local Emergency Command Group. the prescribed burning plan and the infrastructure enhancement plan. to allow that new fires could be managed as incidents with known patterns of spread (and not as random in their behaviour). 2009). it is assessed. Specific duties of the WFA when not involved in active fire fighting operations are: • To document current and past wildfires and to report them in detail. • Where and how to implement slow down operations. the WFA participates with the ‘fire management specialist’ in the tracking. 2009) or LCES protocol (USDA Forest Service 2004).108 Towards Integrated Fire Management 3. both in suppression and pre-suppression operations. Escape routes. At the same time. and if selected. This is especially true for backfiring or burn out operations and operations combining multiple tools (Figure 1). and therefore. In this respect the contribution of the WFA is like an additional check on those actions. • To assess or be in charge of the implementation of any given fuel management action or any given infrastructure enhancement action. Based on this assumption. • To set up the fuel management plan. In this way. When a potential opportunity is found. Elsewhere are specific details on LACES safety protocol (Molina et al. an appropriated tactic is proposed. both resources and efforts are aimed to victory and there is not room for an ‘unpredictable fire blow-up’. Identify areas where forecast fire behaviour allows opportunities and track opportunities with a light vehicle. reviewing and improving a high-quality standard fire safety education and training program. and Safety zones). applying. Anchor lines. • It should be highlighted that the WFA is the expert that assesses actual and forecast fire behaviour to report to the Regional Fire Resources Boss about: • The distribution of resources covering risk and about the potential outcome of different on-going fires to ensure that proportional action is taken and that not all available resources go to the first fire to break out. However. and also how to arrange a hierarchical classification of opportunities to contain the fire perimeter. • To work on developing. an important role has to be played by detailed case studies of past fire propagation patterns and suppression actions (Molina et al. To characterize the different wildfire propagation patterns from past fires in the region. suppression action has a ‘winning label’ from the very beginning. . 2009) to make future fire control actions more secure and efficient. Communications. It is very common to see a stressed out Incident Commander (fire boss) not receiving the best help possible to address how to match resources and effort with actual and future wildfire behaviour.

and maybe not by a WFA. but also in advance to support both planning and training. (Photo by Daniel García-Marco). 3. it can be concluded that the role of a WFA in Europe should go beyond just planning and should include proposing suppression actions. and the FS does this within a fire line. Backfiring operations supervision in Gran Canaria 2007 (Canary Islands.3 Wildfire Behaviour Forecast The main task of the WFA and FOL is to have an educated. The WFA does this not only during fire events to support suppression. a tactic and an execution window. Therefore.Improving Fire Management Success through Fire Behaviour Specialists 109 Figure 1. taken from standard protocols. The tactic and execution window would be addressed most likely by a FOL and a FS. Spain). a proposed methodology (and thus a tactic and explicit limitations). This plan would have an objective (and thus a strategy). . as a result of this. The WFA forecasts behaviour of one or several fires and. The FOL does the same within a fire sector. and an execution window as a result of the objective and methodology. this plan would have a strategy. In short.5. develops a fire control plan for the fire boss (Incident Commander) to approve. detailed look at the wildfire to assess potential changes in spatial and temporal fire behaviour. This is different from what is the standard role of the equivalent specialist in the USA.

and the FS may help the Crew Boss. It may be a written document or an oral communication (most agencies record all radio telecommunications today). In that case the head fire should be addressed later in time and space (see text box).5. The second objective has clear magnitudes (referenced to a map) and could be tested later to see if it was accomplished or not. The left flank should be stopped before this creek while the right flank should be anchored before this other creek on next morning. it has to be simple. this requires simple language and clear maps).4 Suppression planning (objective and strategy) This is the process of establishing the best plan to accomplish secure and efficient fire control when suppression forces arrive to the fire scene. A bad objective could be “We want to control the fire as soon as possible”. see below). This is not the case with the first objective. Additionally. easy to recognize. It is important to recognize when there is not any possible safe and efficient tactic to use. maps and drawings might help to set up an alternative plan. This suppression plan has to be communicated and we must check that it has been properly understood (mandatory checking action). The FOL may help the Division Supervisor. Security issues must always be a major part of the plan.1 Strategy to control fire propagation The strategy for fire propagation control is about how to establish the fire control objectives.4. To do so. sometimes the chosen strategy may be not to address the head fire because it is out of suppression capability even with indirect attack.110 Towards Integrated Fire Management 3. The suppression plan is a set of sequential actions to be taken within a time frame. The WFA helps the Incident Commander to plan for a secure and efficient fire control with the best strategy possible. The objectives have to be communicated to all personnel involved in this emergency. Lastly. The WFA helps the Incident Commander to set up this plan in one or several wildfires. It is not required that this plan is something too intricate. So in this moment. slowing down the head in these fields in the bottom of the valley. It is up to the Incident Commander to decide the strategy after listening to the WFA and the leaders of the different agencies involved in the Incident Command System Group. Objectives have to be easy to communicate. In case of failure to control the fire through this suppression plan. A better objective could be “we want to control the fire. it is necessary to forecast fire behaviour changes in time and space and determine critical point or lines where a greater alignment of factors (Campbell 1995) might put the fire out of our suppression capabilities. Failure could be due to any miscommunication among fire workers. as well as the current fire scenario and its potential runs (dead man zone concept. On the contrary. and with clear magnitudes. It is the duty of the WFA to re-asses (or validate) that the proposed fire control lines are both secure and efficient to establish.5. a drawing in a map may be provided to allow a better understanding of tasks and duties. 3. the priorities are the head and then the left flank with larger potential over the right one”. easy to understand by those in charge of executing it (i. But the priorities should be re-assessed according to changes in fire behaviour. .e.

The objective is to hold the flank where fire runs will begin driven by wind in the night when the northeast wind goes down to surface. Fire on area 7. where the fire is making head runs. Area 2 is a key sector in the spread of the fire. reduce the wide of right flank that goes to area 7. is anchoring the back and flanking towards the head. Night extended attack strategy: Set up by Incident Commander helped by the Operation Section Chief and the WFA. avoiding backing fire spreading to area 2. • stopping the left descending flank. 4. and nor to area 6). is contained out of the sector with few hoses and fighting multiple spotting. hold the back of the fire out of area 2. Therefore.Improving Fire Management Success through Fire Behaviour Specialists 111 Box 1. slow the head spread to reduce the fire front that arrives at area 6 and. present in the field one hour after the fire began. with it. The objective is to contain fire in area 1. The fire goes slowly through area 3 and the execution window is very wide. containing downhill fire approaching area 3. Tactic. 1. Spain) Initial strategy: Set up by Incident Commander (the commandment of the first arriving crews) helped by the FOL. reduce the amount of suppression resources in a front with no potential of spread but with populated areas. 3. The key factor is the lack of pre-attack infrastructure to help suppression efforts. 2. 3. where the northeast expected winds can spread the fire to area 8 and then to area 5. set by this same Incident Commander. stop the fire that goes downslope in area 6 and the left flank. avoiding right flank spreading to area 4. Extended attack strategy: Set up by Incident Commander helped by the Operation Section Chief and the WFA. The objective is to avoid the fire reaching area 8. • avoiding that fire spreading to area 2 and to the populated areas of the village Torre . Therefore. 1. For safety reasons the hose lines on the right flank cannot hold all the line but is narrow. Torre de Fontaubella 2007 (Catalonia. The fire goes from area 1 to area 4 by spotting driven by strong northwest wind (not surface topographic wind). The fire is in areas 1 and 4. The front that jumps to area 7 is narrow and this slows the rate of spread. 2. Therefore. Part of left flank is under control. de la Fontaubella. • avoiding fire runs following the crests in area 7 and spreading to area 8 where there are no anchor points to begin the fire fighting work. The head and the back of the fire are under control (the fire does not spread to area 2.

e. For a given strategy. fire fighters are igniting fire to consume fuels in advance of the spreading wildfire. or land/sea breeze) is required to allow a safe and efficient tactic to be performed. always looking for opportunities that (for the given resources) will maximize fire fighting effectiveness without compromising safety. The execution window is not only useful to ensure efficiency. Let us look at an example: the chosen strategy in a fire sector is to control that fire line at the road.4. As large fires may change so quickly.112 Towards Integrated Fire Management 3. • to rely on both engines and air tankers. The Incident Commander has the leadership. Another example could be when a wind change (i.5. and what they should be communicating. • to backfire from the road. i. Strategy is up to the Incident Commander. Setting an appropriate execution window helps the look-outs understand what they should be looking for. In Figure 1. Different tactics are available. Then. 2009) or LCES protocol (USDA Forest Service 2004). the planning responsibility. It refers to the limits in time and space of the tactic that was decided upon.e. the Division Supervisor may choose: • to control it with engines and hoses. All of them are different possible tactics. Therefore. the training. . Choosing one tactic among others is based on specific training. Sometimes. As we mentioned above.5. while tactics are up to the Division Supervisors and Crew Bosses. this approach can be very effective in limiting the spread of the main fire.4. we cannot light the fire line until suction (from the main fire) is noticeable. any possible tactic has to be validated by checking whether there is an appropriate “time execution window” and “space execution window”. there is a starting restriction in the “execution window”. and the management and co-ordination of resources. decisions about tactics in a particular fire sector can be taken by the Division Supervisor and in any fire perimeter site by the crew boss. and the forecasted fire behaviour changes in time and space and trying to determine critical points or lines where a greater alignment of factors might put the fire out of suppression capabilities. and forest regulations. timed with day/night topographic changes. resources available.2 Tactics The term tactics refers to the planning and execution of actions for accomplishing the strategy. several tactics could be possible in order to implement the control goals. 3. Somehow adding new dimensions the execution window in a given fire sector can be described like this: “our execution tactic would be effective (and therefore.3 Execution window for the tactic It is the third part of the wildfire control plan. but it is also part of the LACES safety protocol (Molina et al. suppression protocols. Safety of personnel must always be a priority. a backfiring operation may require steady wind suction in the fire line to conduct an efficient suppression fire. but also the “weather window” and the “fire behaviour window”. The best tactics (to allow for a secure and efficient fire control) should be taken considering the resources available. It is the way to pursue the goals set by the strategy.

Starting the ignition before the fire begins spotting. 3. This is especially true in fast large wildfires involving wildland urban interface areas. Set up by the WFA and the Incident Commander.Improving Fire Management Success through Fire Behaviour Specialists 113 Box 2. and executed by Strike Team Leader (the boss of a three fire engine team). Execution window for backfire on the head fire • Meteorology. applied by Division Supervisor and FOL. Starting the ignition after successful evacuation of the people remaining on the road. • Safety. Spain) • Backfire at head of the fire front. The suppression fire is set up by the WFA under the Incident Commander supervision. • Aircrafts help the backfire operation and later on right flank. Starting the backfire when the suction of the fire holds the ignition inside the fire line. made successfully in the left flank and stopped for safety reasons in right flank. • Fire Behaviour. • Hose line in the back of the fire. . it will require a common technical language to describe fire behaviour changes. Upslope hose lines in the flanks set up by the Division Supervisor.5. before the daily topographic wind arrives. and while the fire perimeter sends firebrands less than 20 m away”. There are three keys to success: • To have a fast forceful response: increasing distributed vigilance and suppression resources. Tactics in an extended attack at Torre de Fontaubella 2007 (Catalonia. burned area as safety zone. Firing from left flank to right flank for a better way out for firing specialists. Set up by WFA and Incident Commander. Lastly.5 Wildfire Analysis: a needed tool Fire spread analysis is the needed tool to accomplish a safe and efficient wildfire propagation control. Executed by the WFA. before wind changes direction due to storms. • To have a suppression agency flexible enough to respond to rapid changes of fire behaviour: It will be necessary to plan in advance through fire behaviour anticipation (pro-active approach) and not to react to fire (re-acting approach). Burn out to make wider the safety zone. directing prevention efforts to increase suppression efficiency and widening and combining the techniques used to contain fire spread. • Safety. it would be still in shape) before the fire reaches the counter-wind area. • Downslope hose line in both flanks helped by aircrafts as lookout. It will require lowering the decision levels to ensure a faster response to changes.

It will require slowing down the fire head (high intensity fire behaviour). and an additional specific postgraduate or master degree in Forest Fire Management. wildfires). Specifically. FS) does differ from the perspective of an Incident Commander who is a manager of resources. The existence of Fire Specialists fosters these three key aspects. fire effects. Lifelong learning (not necessary at universities) might permit the acquisition of formal qualifications and competencies. the vision of an expert in fire analysis (i. some people may qualify for these positions with no formal academic degrees. This is a needed job position and requires a specific education and training. Therefore. and has the time and focus to apply this knowledge to fire analysis. while allowing the standard fire-line construction to contain the fire perimeter. sectors. and fire safety.2 Required competences training and experience The required level of education of a WFA should include a university degree in Forestry (or related degree). However. In Table 1. the WFA has specialized knowledge on fire behaviour.5. for all these key aspects. FOL. In summary. “to have a fire analyst as an assistant to the chief of the suppression forces (Incident Commander) is like having the peaceful feeling that our efforts will end up in success or that we are not just playing around the fire spread to justify our salary”. 3. In the words of some Incident Commanders we have interviewed.114 Towards Integrated Fire Management • To be able to use fire as a tool to contain high intensity fire spread situations. Clearly this progress needs to continue and to multiply. It is perceived that the investment in training and positions results in improved fire management effectiveness. Going into the details. the usual way to reach the qualifications is through attendance at a course run by a university. As we have seen. 3. applying fire analysis in every job position of a fire fighting organization is of paramount importance. WFA.1 Agencies with Fire Specialist positions on duty in 2009 The current situation is rapidly improving over recent years in many regions in Europe. Several years in fire fighting operations are certainly important to qualify as a WFA. the WFA can properly identify fire suppression opportunities but can also assess in time when some efforts are either meaningless (inefficient to control the fire perimeter) or unsafe for the workers.e. and so improving knowledge and understanding all through the different levels in the chain of command. However. it should be highlighted that it should be possible to transfer experience through high quality case studies.5. These first steps in Europe are promising. some specific data is provided regarding Fire Specialists and their training. safe and efficient wildfire suppression is harder to be accomplished without a proper WFA on duty. promoting the use of fire analysis in the three spatial scales of the fire (fronts. Therefore. but the efforts need to expand and accelerate – positions need to be recognized and become a part of the regular fire operations during fires and year-round to support planning and training. This specific degree in Forest Fire Management must include credits in prescribed .5.5.

Aerial means operators. Fire Analysis training Region Wildland Fire Analyst Catalonia (northeast Spain) GRAF. Methods and protocols are tested in 20093 Tenerife (Spain) There are moves towards employment of WFA.4 Sardinia (Italy). Aragón (Spain) 5 forest fire managers in training Andalucía (Spain). GRAFF - Gran Canaria (Spain). follow active fires Basic training in wildfire analysis Continuous training through past wildfire analysis. Understory prescribed burning. and FOL In progress No 4 – There is not yet an official function L´Aude (France). FS / FOL Several FS. but GAUF personnel are often requested to help in this task. GAUF units training includes 5 days training sessions of fire analysis and prescribed burnings. 4 additional assistants. there is not yet an official function of fire analyst. 2008 No Northumberland (UK) - Improving Fire Management Success through Fire Behaviour Specialists 115 1 Fire suppression support. Continuous training: Past wildfire analysis and wildfire reports (in progress). formed by full time WFA as well as part time Wildland Fire Operative Technicians. Look-out. Training fire Incident Commanders in fire analysis. Fire analysis for most levels (in progress). Since 2002. GAUF2. 1 additional assistant in summer Castilla-La Mancha (Spain). Regions (and countries) and their Fire Specialist positions on duty in 2009. Fire analysis for most levels. 4 FOL (in training) FS in training No 4 forest managers analyze past fires. Expected 2010. and FOL FS Several FS. Control operators. Since 2009. Several FS. In the civil protection system. 150 h for FOL. in Portugal. Division commander for fire suppression operations in Mediterranean Ecosystems. Since 2006. Portugal (all country). Continuous training: Past wildfire analysis and actual wildfire reports. Basic Fire suppression. This could eventually allow for a more specific WFA position. 3 An analysis unit. Mastros de Fogu. 4 There is a move towards this new job position but there is a complex protocol (in large wildfires) that accommodates several different tasks for up to four forest fire managers on duty. 14 full time (some in training phase). and FOL Fire analysis for all levels1. Continuous training: Past wildfire analysis and wildfire reports (in progress). Since 1999 8 full time. Every year since 2006 they organize a 5 days training sessions concerning the wildfire analysis. Fire commanders are being trained in fire analysis. including prescribed burning and analysis exercises for GRAF.Table 1. is expected to be operative by summer 2010. Incident Commandment. . UNAP since 2009. Training high and intermediate fire commanders in Front and Sector fire analysis. 2 full time. 2 All GAUF members are forestry engineers as well as certified as prescribed fire specialist. Fire Brigade unit commander for Fire suppression operations. GRAF units training includes 200 h of training (FS).

fire analysis. 2009) and capable to forecast and assess fire behaviour changes under changing topography and weather (Campbell. Braun 1995. climate. the social and cultural role of fire in Europe. they will cover the state-ofthe-art of scientific knowledge in fire ecology in the European biota. Molina et al. strategies and tactics. Country ‘Annexes’ providing specific information on particularities of wildland fuels. 1997. In terms of suitable education. • A close-up participation in recreation and analysis of fire spread in numerous past fires as shown by Molina et al. the new training systems and practices will strengthen European co-operation. The required extensive training may include: • A manager level position in Forest Fire Management during at least two fire seasons. and Geller 1996). (2008). fire behaviour. Additionally. training. The WFA should also be able to use fire simulators (Finney 1998. Finney et al. • Familiar with job hazards abatement actions and aptitudes (Beaver 2001. This is something not required for the Incident Commander but it is a must for a WFA. By addressing European-wide training needs. the WFA should be: • Familiar with Wildfire Typologies in the region (Castellnou et al. the impacts of fire on atmospheric chemistry. Castellnou et al. and with conditions leading to accident identification (Putnam 1995. (2006) and Grillo et al. (2001). Hart 1995.116 Towards Integrated Fire Management burning techniques. Mangun 1995. Castellnou et al. Expertise (both education and extensive training) in the wise use of fire (prescribed burning as well as backfiring operations in suppression actions). experience and certification of a WFA. and Cheney et al. In those works. as a result of this pattern either being properly perceived or not being properly perceived. fire ecology. 1995. This helps to understand minor changes in fire behaviour and how to address them. in regard to how fire effects are linked to fire behaviour. 2006). and the use of prescribed fire in ecosystem management. • A deep knowledge of fire ecology. fire prevention and suppression methods and technologies and the use of prescribed fire in ecosystem management. several case studies are shown. • Familiar with prescribed burning techniques. in which the fire control actions either did match or did not match the fire spread pattern. fire weather. and will amplify support at the local service level for the development of qualifications and competencies as WFA. USDA Forest Service 2001. Additionally. fire effects. 2009). • Knowledgeable in fire ecology and fire effects (this may be obtained partially in prescribed burning actions). These training materials will include fire science. (2009). Fire analysts may enhance their performance by becoming knowledgeable about the body of works by Campbell (1995). management options and country specific legislature and protocols may be included. human health and security. . these should be similar to what it is required for an Incident Commander. Mangun 1999) in order to carry-out fire analyses that help achieve efficient and safe fire suppression. will develop common principles for informal learning.

as listed above. Therefore. If these hours are spent under adequate supervision. . Spain) have set up excellent working documents to ensure a smooth transition to a system that relies heavily on the WFA tasks. These diverse aspects (that indeed provide background knowledge to Fire Specialists) merge wildfire propagation with the wise use of fire. Having paved the way. in relation to fire suppression. These assignments may be carried out by any fire manager.5. For example. 3. some recent developments (i. large wildfire analysis. prescribed burning plans are their duty. it has also contributed in enhancing the credit of those already doing this highly needed job. Additionally. the trainees should be encouraged to anticipate fire behaviour changes (a pro-active look at fire spread).5. in the present volume).6 Discussion Molina et al. see Miralles et al. It should be highlighted that having an expert wildfire analyst on duty in those cases could have made a significant difference in the final fire perimeter and in human safety (including job hazard abatement actions). There are already different attempts towards this accreditation throughout Europe (see Miralles et al. The certified WFA may perform several tasks in both prevention and suppression. learning and training in prescribed burning provides an outstanding understanding of wildfire propagation because they allow for many hours to observe changes in fire behaviour and spread. It is rewarding that there are new positions as WFA in different agencies.Improving Fire Management Success through Fire Behaviour Specialists 117 In summary. WFAs are able to write down fuel management plans or integrated fire management plans. and arguments stated in this chapter strongly suggest that it is clearly beneficial to have a WFA on duty. it is obvious that the WFA position is very important and deserves recognition. demonstration sites and training actions in prescribed burning and detailed documentation of fire analysis of recent medium and large wildfires.7 Conclusions and recommendations Field experience. They can also carry out a host of other tasks.e. it integrates diverse aspects of wildfire propagation using the knowledge obtained. This chapter has clearly been based on Fire Paradox activities including monitoring of active wildfires. (2009) have shown several case studies in which the fire control actions did not match the fire spread pattern because this last one was not perceived properly. CastillaLa-Mancha. It is also clear that a proper certification system is needed to ensure that the most qualified candidates are hired for the job. but they can be accomplished much better by a certified WFA. As to the required competences needed for the Firing Specialist and the Firing Operations Leader positions. In fact. 3. in the present volume. it is certain that there will be even more WFAs in more agencies in the coming years.

1995. fire engines or logistic experts. MT. D. U. Missoula. Fire specialists would have the knowledge to propose adequate strategies and organize tactical operations. Department of Agriculture.S.O. Collins. L.P. Symposium on Fire in California Ecosystems: Integrating Ecology. MT. Additionally. 47 p. Rocky Mountain Research Station. Ft. Therefore. Proceedings of the findings from the wildland fire-fighters human factors workshop. Pp.B.A. Patrones de Propagación de Incendios Forestales: In: Vélez. R. Campbell. M.. Research Paper RMRS-RP-4. MT.e. as the top experts. Missoula. Proceedings of the 5th International Fire Safety Summit. 9951-2855-MTDC. CA. knowledge on fire propagation should be common to all people involved in fighting fires. Use of FARSITE for Simulating Fire Suppression and Analyzing Fuel Treatment Economics. D. Miralles. Chilton Book Company. E. . 1997. M. high-speed teams. The psychology of safety. Addressing the common behavioural element in accidents and incidents. (ed..118 Towards Integrated Fire Management Fire specialists (i. Cheney. 2001.).and large-size wildfires. Cultural attitudes and change in high-stress. 1998.K. T. Hart. Pp.S. 2009. Molina. inflection points) where fire suppression actions would have an increased probability of success. Forest Service. The Campbell Prediction System: A Wild Land Fire Prediction System & Language. 28–30. and Martinez. D. 34–39.S. USDA Forest Service. It is envisioned that the Fire Specialists should be standard job positions in those wildland suppression actions that exceed the limits of initial dispatch. 1995. critical points) and opportunity sites (i. N. Campbell ed.). and FS) would efficiently identify critical points (or lines) where fire intensity and spread could increase seriously (i. M. Geller. 9951-2855-MTDC.R. and Bahro. CO. Radnor. FOL.M. Gould. Missoula. Prevention. WFA. Dead Man Zone – a neglected area of firefighters safety. C. (ed. Incendios Forestales: Fundamentos y Aplicaciones. Finney. In: Putnam.. Forest Service. In: Proceedings of the findings from the wildland fire-fighters human factors workshop. fire behaviour specialists would help to solve the fire paradox by means of merging a comprehensive knowledge on fire propagation with mastery of the wise use of fire. In short. It is strongly recommended that Fire Specialists should be viewed as being equally indispensable as helicopters. Sapsis.S. B. the Fire Specialists should train teams in job hazard abatement actions and in fire fighting exercises all year round. Putnam. and Management. FARSITE: Fire Area Simulator-model development and evaluation. However. 129 p.C. T. M. J. 1995.e. Australian Forestry 64(1): 45–50. Braun. 2001. (ed. Pp. Castellnou. 274–282. E. D. 1996.e.). U. PA. McGraw-Hill. Department of Agriculture. Finney. Evaluating risk and reward relationships in wildland fire-fighter safety.. airplanes.A. References Beaver. San Diego. International Association of Wildland Fire. D. and McCaw. it is necessary to address how to certify the required qualifications to ensure that those job positions will be filled with the best workers available. A. Therefore the message to policy makers is to encourage agencies to have WFAs on duty for mid.

. Incendios Forestales: Fundamentos y Aplicaciones.D. Gran Canaria. Granada. 2009. R. Thirtymile fire investigation: Accident investigation factual report and management evaluation report. 9951-2855-MTDC. T. and García-Marco. Molina. Molina. Department of Agriculture.. USDA Forest Service 2001. Planes de quemas. Editorial AIFEMA. Castellnou. Invest Agrar: Sist Recur For 15(3): 271–276.J. and García.M. Forest Service. Lecciones Aprendidas y Retos de Futuro.S. USDA Forest Service. Pous.J. Incendios Forestales: Fundamentos. U. Missoula. F. J. Blanco. D.).M.. D. Granada. R. In: Vélez. D. 9951-2808-MTDC. Pacific Southwest Region.S. USDA Forest Service. Putnam. 1995. García. Prescripciones. Fire Siege of 2003. F. Molina. 2008. 144 p. 256 p. (ed. Mangun.. Findings from the wildland fire-fighters human factors workshop. 274–282. D. MT. Análisis del Incendio Forestal: planificación de la extinción.Improving Fire Management Success through Fire Behaviour Specialists 119 Grillo. MT. 9551-2845-MTDC. E. Grillo-Delgado. MT. Forest Service. Molina.B. Editorial AIFEMA. 1999. Spain. Galan. McGraw-Hill.). Safety Protocol Review. USDA Forest Service 2004. 2006. D. 1995. Uso del fuego prescrito para la creación de rodales cortafuegos: estudio del caso “Las Mesas de Ana López”. J.M. 2009. Spain. U. Martínez. U. Southern California Geographic Area. E. Mangun.R. Winthrop. Wildland fire fatalities in the United States 1990 to 1998. M.. Missoula.. . 34 p. D. (ed.M. Pp. 106 p. D. Investigating wildland fire entrapments. Missoula..F. M.S. Vega de San Mateo.. Department of Agriculture. R. and Fababú. España. Department of Agriculture. Forest Service.

.

3. But under certain meteorological. the suppression system cannot carry its operations efficiently as in the case under normal conditions.6 Wildfire Scenarios: Learning from Experience Marc Castellnou1. Portugal in 2003 and 2005. Spain 2 Unit of Forest Fires. so in an instant fire can advance by 200 m. southeast France in 2003 and 2009. • With such extreme fire intensities. a negative feedback process has developed. fire spread is faster than the fire suppression rate. For example. For example. In other words. The applied strategy for most wildfire suppression cases was based on a fast and aggressive attack of all ignition points in the landscape. In the wildland urban interface situation. This demonstrated that strong suppression resources capabilities are still inadequate when faced by extreme fire behaviour. • Fire perimeter growth. and Greece in 2000. Spain 3. and no aerial or terrestrial resource can stop it. Oriol Vilalta1 and Domingo Molina2 1 GRAF. fuel and topographical conditions. The efficiency of this strategy is outstanding and it succeeds in limiting most wildfires to very small burned areas. Asier Larrañaga1. • The distribution of the resources along the perimeter is slower than the growth of the wildfire perimeter. Department of Internal Affairs. Suppression of small fires can be effective under virtually all conditions however it is reasonable to say that once a fire develops under certain extreme weather conditions it is impossible to limit fire spread until the weather moderates. massive spotting at 200 m is very usual in large fires. at every place and under all meteorological conditions. increases the probability of damage to houses and people and the necessity of actions and resources to protect them. Extreme fire intensity thus overruns suppression limits. 2007 and 2009.1 Introduction The occurrence of episodes with large wildfires has affected different regions of Europe. Barcelona. During all these large fires suppression crews faced extreme fireline intensities and fast fire spread which made it impossible to quickly suppress them by the known attack techniques (direct attack). where the relative amount of resources used to fight the wildfire is lower in comparison than the resources which are used to protect the houses and people. Every house can be regarded as an emergency (Rifà and Castellnou 2007). University of Lleida. fire behaviour becomes extreme. When fire behaviour is clearly superior to the power of the suppression resources.6. which is faster than the suppression capacity. crown fires with large walls of flame greater than 25 m length are not unusual in large fires. Marta Miralles1. northwest Spain in 2003 and 2009. But this strategy that only focuses on total suppression has shown not to be successful . the system becomes overwhelmed and turns out not to be efficient for various reasons.

during the next 30 years. 2002. So we need to gain suppression effectiveness to address these fires.e. Processes like the interaction between different fire fronts or the changes in oxygen supply following changes in the convection column (by means of instability of the atmosphere) are very difficult to simulate but are essential for predicting fire behaviour (Dupuy 2009. Different research initiatives have attempted to make progress in this direction: • Wildfire spread patterns. Furthermore. Mazzoleni et al. For these reasons. However wildfires are very complex events. So. to anticipate as much as . especially those with extreme fireline intensities. Based on this. • CPSL (Campbell Prediction System Language) analysis. to mitigate wildfire effects it is necessary to focus on changing the landscape. • Fire scenarios: large wildfire spread patterns that occur in a given forest massif. Bahro et al. towards a more tolerant and resistant land use. it is important not only to create and to improve simulation models (Vega et al. helps to predict where the fire behaviour will change (Campbell 1995) based on factor alignments (slope. Finney 1998). Finney 2007). The anticipation of the attack opportunities is far from being a standard in suppression systems. 2009. included in prevention programs: locating the places where adequate and proactive forest management could reduce fire intensity. wind. based on main paths and predicted fire spread schedules. and the investments in models and simulations are impressive. mainly because of the complexity of the wind-relief interactions. The dominant spreading factor can be determined. and in this way it is possible to pre-determine which type of changes can be expected and will generate opportunities (developed in the first half of the 20th century). A change in strategy is needed. from strategies based on direct attack to strategies based on slowing and reducing fire behaviour and defending urban assets. the available data is scarce. we will not have the desired landscape yet. Mazzoleni et al. Fire fighters cannot stop them. A great opportunity to begin this task is before the fire occurs. aspect – fuel temperature. in a “fire shed” and homogeneous fire behaviour. Anderson 1983. Andrews 1986. Anticipating wildfire spread pattern provides the opportunity to increase the success-rate of the suppression attacks by allowing fire fighters to pre-plan decisions and timely application of suppression methods. There are several forecasting tools for fire behaviour (i.). creating opportunities for the likelihood of successful suppression. Bahro et al. we recommend a methodology to anticipate the local spread paths followed by large fires moving in a given fire scenario (Castellnou 2000. 2008). 2008). Rothermel 1991. (2007) proposed a methodology for planning based on a “problem-fire”. 2007). • Simulations based on minimum travel time which permit identification of fire opportunities (Finney et al. • Surface-fire prediction by static simulations (Behave) (Andrews 1986) or spatial simulators of fire spread (Finney 1998) can be used with advantage to anticipate sites where lower fire intensities can be expected. Dupuy 2009. In the same way. and we will have large fires that no suppression effort can stop. and ‘provinces of fire’ (with homogeneous fire behaviour) (Castellnou 1996).122 Towards Integrated Fire Management under large wildfires. This clearly requires an important investment in knowing and anticipating fire behaviour. However. but to put emphasis on the investigation of historical large wildfires.

This process requires a structured standard (fire types) which allows the relevant information to be identified and highlighted. Every wildfire can be related to a certain meteorological condition and a synoptic situation of the parameters which define the fire (wind intensity.Identify the schedules of fire spread which are repeated in every zone with a homogeneous fire regime. This cannot be seen as a genuine opportunity. type of wind. 3. The implication is that the fire spread potential increases notably by not identifying such sites. 2009). Vegetation structure and drought 1 Trigger point: site in which fire spread and fireline intensity will be beyond the holding capacity of the suppression forces in direct attack. areas associated with large fire potential. and fire prevention should locate the needed infrastructures (Rigolot et al.2 Research of former wildfires and related synoptic situations Predicting fire behaviour and fire spread patterns requires going back to the records of wildfires that have occurred in the past. but forest management should be directed to reduce the potential fireline intensity. • Areas to reduce fuel combustibility: not directly related to the main local spread path. which are based on repeated. • Determine the fire scenarios in zones with homogeneous fire regime: . observations of paths of spread following the topography (Castellnou et al. The objective of these reports should include capturing the lessons learned from these fires. and to anticipate extreme weather conditions associated with extreme fire behaviour in each scenario. and relative humidity during day and night). Those are classified for given fire types. 2009). The fire reports should describe fire behaviour.Study the probability that the determined scenario of opportunities would be useful. Campbell 1995) to determine the main paths followed by the fire and its trigger points1. and which characterizes the wildfires. these determine fire spread pattern and fire behaviour. By understanding the schedules of fire spread we can now determine which areas are inaccessible for suppression and which ‘zones’ present opportunities for successful fire fighting.6. . To be safe and useful for suppression attack it should be prepared before. These areas are in the main axe of the local spread path. The methodology to determine strategic zones is based on the development of the following: • Research and study of historical large wildfires in an area. These strategic zones are: • Critical areas to be managed: where the suppression resources can have an opportunity to reduce fire intensity during the day of a large wildfire.Wildfire Scenarios: Learning from Experience 123 possible the suppression opportunities. based on the analyses of past fire spread (propagation) and CPSL (Campbell Prediction System Language. . this information usually becomes part of the fire fighters’ knowledge.

The transfer of useful information and the lessons which are learned from the fire are the main objectives of these reports. The contrast between wildfires with large perimeters (large areas) and very complex wildfires with crown fires. 2009). in combination with the socio-economic changes during the last century. and roads and observation posts for fast suppression. have influenced the landscape and the load and distribution of fuels.from prevention based on infrastructures and water points to facilitate the creation of anchoring points. the problem of huge wildfires is changing. general behaviour and existing resources. operate in extended attack and most of the personal is involved full time. (Castellnou et al. . This includes patterns of spread and fire type. For every fire scenario. like prescribed burning and new regulations with respect to wildfires (Herrero et al. also only have a small number of real opportunities.124 Towards Integrated Fire Management affect the quantity and distribution of fuel. 3. but also in anticipating fire management opportunities). requires a shift from few resources exploiting the existing opportunities. at local scale. wildland urban interface problem areas and simultaneous fires. • Used strategy and discussion of the performed tactics and operations. in operational actions and in politics. or lessons related to safety are some likely products. 1998. the most relevant historical fires are analyzed. 2002). and the influence of topography by itself on fire behaviour. The development of the territorial scenario has caused changes in the scientific world. • General information about the fire: Date. focusing on the following aspects: • Main conclusions. and therefore also fire behaviour. These crews have a solid training background (not only in using suppression tools. unfortunately.3 Fire scenarios In Europe. we do not only transfer data from fire perimeter simulators like Farsite (Finney 1998) to trajectory simulators which draw spread path schedules and identify points where fire fronts multiply (Finney et al. hour. Millan et al. Montserrat 1998) and. For analyzing the local winds. Nowadays there is also a change at the operational site. the implementation of selected tactics. to make it feasible to determine possible return-period/recurrence-interval. Hence. into numerous suppression troops which. It is important to link the former wildfires to the drought level at that time. This has caused a change: . The developments of different land use. more and more regions and countries integrate new tools of fire prevention in their systems. • Location of the fire with respect to the topography. the interaction between wind and topography.6. if available. 2009). from temporary workers during the 1960s to highly specialized suppression crews like GRAF (Catalonia) or GAUF (Portugal). In addition. Lessons with respect to the organization of the suppression operations applied. size. • Meteorological conditions at a synoptic scale (Castellnou 1996.

2009 (Andalucía. . Another way to explain the lessons learned from past fires is by the use of video and documentaries. These crews have a large capacity to analyze fire behaviour and to identify local operations and possibilities. 1979). and the wind increase on the 3rd day (Sagarzazu Ansaldo and Defossé 2009). duration and social importance since the formation of the specialized crews for wildfires in Catalonia (GRAF).g. Figure 1. Spain) the main fire potential was limited by the presence of agricultural areas. Spain). When working with reports of large wildfires. Fire behaviour in Sant Llorenç Savall 2003 (Catalonia. This was the first significant wildfire in terms of size. which act as a fuel break making suppression actions easier (Ruíz and Senra 2009). Figure 2. Binggeli 2009). Picture from the Epuyén-report (Argentina. and those which faced major problems in fire suppression An example of the extracted information from former wildfire reports can be seen here: • In Casarabonela. Furthermore they assisted in coordinating the various different volunteers following the strategy of the fire fighters (GRAF 2009). These are very educational for civilian people to help them understanding how a wildfire behaves (e. When the fire crossed the area of the main river.Wildfire Scenarios: Learning from Experience 125 Box 1. Examples of large wildfire reports and their use in simulation. The videos can also be used as for a professional public as an education tool about the use of fire. the direction and speed of fire spread changed drastically due to the slope. it is necessary to identify those in which the most important lessons were learned.

There is a significant change: from the objective of eliminating fire from ecosystems. This is based on the repetitive observations of fire spread under similar topography and meteorological conditions. the pre-suppression actions and prevention needs are able to be effectively planned. 2009). Rigolot et al. For a given scenario the first step is to identify areas with a common fire type. 2003. The macro-topography (with synoptic situations associated to large wildfires) is the main factor which controls fire propagation in complex terrains (Taylor and Skinner 2003. there may be more than one type of wildfire. Common ‘fire types’ are proposed by Castellnou et al. and especially how large wildfires develop. Therefore well documented and relevant wildfire history for the region should be studied. but each fire type and its associated opportunities should be applied to a specific territory. where subsequently the underlying pattern can be seen (Castellnou 1996. 2007 and 2009. Based on these fire spread patterns and the associated fire potential. Fire behaviour and spread pattern of . wind-driven or plume-dominated fires) several specific and repeated aspects are determined for every pattern. With this purpose it is necessary to understand very well how fire behaves. the south of France and Greece in 1994. and propagate in almost every type of fuel. Bahro et al. Expósito and Cordero 2004. adapting the land use and forest management to wildfires by reducing the vulnerability of forest stands. The history of wildfires is essential. The existence of repeated fire spread patterns linked to a specific territory provides an opportunity to identify a ‘problem fire’ and to set up the proper planning (Castellnou 2000. From the former wildfires. Portugal. Expósito and Cordero 2004). This fire types are a useful way to extract information from the main observers of fire behaviour. Wildfire type characterisation for homogeneous areas of risk From the three basic patterns of fire spread (topographical fires. 2008). depending on the available fuel. These fires do not respond to small details. including agricultural areas or decorative vegetation (Bajocco and Ricotta 2007. the fire fighters. We have also seen examples of this type of destructive behaviour in Spain. from the characterization of coincidences of synoptic and topographic conditions with repeated fire path schedules.126 Towards Integrated Fire Management . to the advantage or disadvantage for the suppression crews (Campbell 1995) and predicting how fire front will subdivide into several directions (Finney et al. but to larger landscape-scale features. Large wildfires are not very selective concerning fuels. Iniguez et al. 2009). based on the study of the wildfire history of Catalonia. The planning of the use of terrain features is not done based on some general and abstract fire types. 2002) will assist in understanding fire spread of the large wildfires (Castellnou et al. connecting the world of suppression to the world of prevention. only differing in intensity and therefore the distance which the fire reaches. Identifying fire-design and strategic zones Identifying where the fire behaviour will change. (2009). Nevertheless on the same site. using the opportunities provided by fire.to prevention based on the strategic opportunities of large wildfires. a series of fire scenarios in different synoptic and topographical situations are determined. an adaptation to fire. 2007).

typical development of the synoptic episodes. which will be one of the criteria to select the best sites where to invest resources. as well as real opportunities for suppression. by identifying the type of each particular wildfire. • determine sites with different vegetation treatments. • select the best sites where fuel reduction methods should be applied to reduce fire spread potential. This will lead to a development of common schedules. to establish safety zones from where suppression operations can be applied. It is a way to provide information and provides criteria to locate the infrastructures which will support fire suppression actions. is the anticipation of fire trajectories (fire spread schedule in time and space) in the near future. This will then serve to make sound decisions in fire fighting.4 Conclusions and recommendations The most useful tool to stop wildfires – which escape from the conventional suppression methods (mainly because of high flame lengths and fast rate of spread).6. but on information extracted from past fires. related to the topography. and will result in safer and more efficient suppression actions. This type of understanding (based on the underlying mechanisms) is obtained from detailed analyses of former wildfires. allows the identification of strategic sites for each environment. • recognize when resource allocation will not result in a positive outcome. and . We can then consider if these opportunities could be used or not in this terrain. By gathering this information. i. particular properties of the fire spread are defined for the specific study area. Taking the information collected of the fire-design as a reference point. The fire-design represents a ‘best-reference fire’ or a fire with the ability to become a large wildfire in a specific environment (Castellnou et al. With this information. similar scenarios and repeated windows of opportunity for each wildfire type. and to predict how it spreads. we endup with a concept of ‘wildfire-design’. Each of these strategic sites is associated with a fire potential area. and gathering the experiences of the validated opportunities of historical wildfires. 3. we create databases of wildfires which will describe the shape of fire spread patterns in a particular synoptic (fire weather) and elevation (topography) framework. In these past fires. associate to rates of spread and intensities safe and efficient to work. • confirm the location of the opportunities. and properties of vegetation types. as well as the operations conducted. 2009).Wildfire Scenarios: Learning from Experience 127 these fires are analyzed. The use of simulations permits confirmation and adjustment of several parameters which are used for the planning of strategic zones: • map inaccessible areas for the suppression crews. also taking notes on its weak points. The fire-design is a tool to understand how large wildfires perform in a certain territory. This anticipation is not only based on insight gained by people or on mathematical models adjusted to real interaction between fire and a landscape. By specifying the fire types adjusted to the specific properties of the site.e. different theoretical opportunities considering relevant fire types for the zone are studied.

Perimeters of fires in 1949. The counter directive winds reach the higher part of the ridge where the main wind occurs. a second ridge was available and burned). Fire spreads following ridges until the end. Wind direction Uphill backwind Spot fires. as the ignition in 1979 was more to the east than the other two. are all captured in the fire-design. Acknowledging and understanding the fire paths in a lee wind area. The three fires show similar spread patterns with a synoptic situation of northern wind. and 2000. Location: Albiol. northeast Iberian Peninsula. with similar fire paths schedules with analogue synoptic situations and macro topographical conditions (Castellnou et al. to the development of the synoptic situations in the area or to vegetation characteristics. Catalonia. This box describes the fire types and related opportunities. 2009). blown downhill by the main wind flow Figure 4. provides possibilities to identify the suppression potential at the end of a ridge or where the ridge drastically changes direction. Figure 3. 1970. The properties of each fire type in a given scenario due to local topography. Instructions to integrate fire in forest management. . In this way.128 Towards Integrated Fire Management Box 2. spotting fire downhill. Propagation by lee winds. and open to new ridges if available (for example. new fire-runs develop uphill following the lee winds.

Wildfire Scenarios: Learning from Experience 129

Box 2. Continued.

Figure 5. Perimeters of the fire in the main valley of the River Segre, Catalonia.

Figure 6. Topographical fire spread pattern near a canyon.

will provide possibilities to plan the pre-attack needs and extinction resources accordingly. In this way we can search for minor actions (CSM – Critical Sites to be Managed) which will reduce large wildfire spreads. Therefore, studying the repeated local spread path of past wildfires allows managing a fire scenario, creating discontinuities of vegetation disrupting these paths and managing vegetation to manage the whole scenario. One single methodology of gathering general data for the whole European Union should cover the complete range of different fire types, and increase the number and accuracy of fire-designs. A platform of exchanges of information for the whole Europe is needed. Using one common language in the characterization of wildfires will simplify the know-how transfer of specialists/analysts between the regions’ fire fighting projects and between the states of the European Union. With respect to wildfire prevention, this exchange of knowledge allows the creation of a

130 Towards Integrated Fire Management

Box 3. Wildfire Scenarios – l’Empordà, Catalonia, Spain.
The study of large wildfires (those in the country but also nearby) resulted in the identification and confirmation of the fire scenarios which affected each area. From a database of 46 perimeters representing 67 471 ha, we can conclude that the Cap de Creus (the little peninsula in the northwest part of the picture) burns every 26 years mostly by wind-driven fires (also with topographical constraints). The validation of former wildfire control opportunities permits us to extrapolate the lessons learned to each new wildfire scenario. Then, the CSM (Critical Sites to be Managed) can be determined. Every CSM is linked to a potential, planned burning surface that can be protected.

Figure 7. Main CSM as black ellipses, to protect against eventual 5000 ha wildfires.

common working guide. This could help to implement suppression models and their requirements, while establishing an assortment of general opportunities for every fire type in all the regions. With this methodology it is also possible to highlight the different problems which the suppression crews (in different regions) may face. It is possible to search for common/shared solutions within the mix of measures that every suppression system can provide. Thus we can clearly map the problems, and investigate different solutions.

Wildfire Scenarios: Learning from Experience 131

In this manner it is possible to work on preventing the worst possible wildfires in these zones. Besides, we will know beforehand what the critical situations will be, which can occur simultaneously (or within a short period of time) in different parts of Europe. This was the case within the synoptic pattern at the end of July 2009 when there were large wildfires occurring at the same time in different parts of Spain (Castilla La Mancha, Aragón and Catalonia) and one day later in Sardinia (Italy). The study of present and past large wildfires is essential for integrating forest management and suppression actions where both have the same objective: the persistence of structures and species as far as possible. We should focus on the larger wildfires, because for the smaller fires the present suppression methods are adequate. This will allow to plan several actions like pre-attack (locate the places where the fire fighters could have suppression opportunities) and prevention (locate the places where adequate forest management could reduce fire intensity, and by this creating opportunities for the likelihood of successful suppression). This will also assist in deciding where to make use of agricultural breaks, and to determine where houses are most vulnerable. It helps also to reinforce or favor vegetation structures which are more resistant in areas of full alignment (i.e. worst fire behaviour). In the same way as it is possible to place structures which are less vulnerable to crown fires in trigger points, so, the study of wildfire scenarios is also an integral tool for forest management. Nevertheless, it is important to remember that determining where to act and increasing the planned activities is necessary but not sufficient to limit the destructive effects of the large wildfires. The integrated actions will then be completed by an increase in the most effective forest management options in which the risk of wildfires is explicitly incorporated, with expanding forest tracks which facilitate establishment and maintainance of forest structures which are more tolerant to the fire spread and related damage. This will also reduce the capacity for sustaining crown fires because of the more discontinuous nature of the vegetation, which will assist fire suppression in general.

References
Anderson, H.E. 1983. Predicting wind-driven wild land fire size and shape. Intermountain Forest and Range Experiment Station, Ogden, UT, USA. 26 p. Andrews, P.L. 1986. Behave: fire behavior prediction and fuel modeling system - BURN subsystem Part 1. USDA Forest Service, Intermountain Research Station, Ogden, UT, USA. 130 p. Bahro, B., Barber, K.H., Sherlock, J.W., and Yasuda, D.A. 2007. Stewardship and fireshed assessment: a process for designing a landscape fuel treatment strategy. Restoring fireadapted ecosystems. Proceedings of the 2005 national silviculture workshop, Gen. Tech. Rep PSW-GTR-203. Pp. 41–54. Bajocco, S. and Ricotta, C. 2007. Evidence of selective burning in Sardinia (Italy): which land-cover classes do wildfires prefer? Landscape Ecology 23: 241–248. Binggeli, F. 2009. Linking bulletin and video reports for the website. Internal Report IR8.5-1 of the Integrated Project “Fire Paradox”, Project no. FP6-018505. European Commission. 6 p.

132 Towards Integrated Fire Management

Campbell, D. (ed.) 1995. The Campbell Prediction System: A Wild Land Fire Prediction System & Language. 129 p. Castellnou, M. 1996. Reconstrucción de la progresión de los incendios históricos y su efecto modelador en la vegetación: Ribera de Ebro. In: Gonzalez, J.M. (ed.). Seminario sobre incendios forestales. Centre Tecnològic Forestal de Catalunya. Solsona, Spain. Pp. 207–223. Castellnou, M. 2000. Nuevas Metodologías de Prevención de Grandes Incendios. Congreso Forestal Ibérico. Castelo Branco, Portugal. Castellnou, M., Miralles, M., Pages, J. and Pique, M. 2009 Tipificación de los incendios forestales de Cataluña. Elaboración del mapa de incendios de diseño como herramienta para la gestión forestal. Proceedings of the 5th Congreso Forestal Español. Ávila, Spain. Dupuy, J.L. 2009. Fire start and spread. In: Birot, Y. (ed.) Living with wildfires, what science can tell us. EFI Discussion Paper 15. European Forest Institute. Pp. 27–31. Expósito, R. and Cordero, T. 2004. Albiol fire. Did we learn at the third time? In: Viegas, D.X. (ed.). Proceedings of the IV International Conference on Forest Fire Research. Coimbra, Portugal. Finney, M.A. 1998. FARSITE: Fire Area Simulator - Model development and evaluation. RMRS-RP-4. USDA Forest Service Rocky Mountain Forest and Range Experiment Station, Fort Collins, CO, USA. 47 p. Finney, M.A., Sapsis, D.B. and Bahro, B. 2002. Use of FARSITE for simulating fire suppression and analyzing fuel treatment economics. Sugihara, N.G., Morales, M.E., Morales, T.J. (eds.). Proceedings of the symposium: Fire in California ecosystems: integrating ecology, prevention and management. San Diego, CA, USA. Pp. 121–136. Finney, M.A. 2007. A computational method for optimizing fuel treatment locations. International Journal of Wildland Fire. 16: 702–711. GRAF 2009. The Sant Llorenç Savall Forest Fire – Operational Report. Generalitat de Catalunya. 23 p. Herrero, G., Montiel, C., Agudo, J. and Aguilar S. 2009 Assessment document on the main strengths and weaknesses of the legislation and policy instruments concerning integrated wildland fire management in EU, in European Member States and in North African countries. Deliverable D7.1-1-2 of the Integrated Project “Fire Paradox”, Project no. FP6018505. European Commission.117 p. Iniguez, J.M., Swetnam, T.W. and Stephen, R.Y. 2008. Topography affected landscape fire history patterns in southern Arizona, USA. Forest Ecology and Management 256: 295–303. Mazzoleni, S., Giannino, F., Beccarisi, L. and Ricotta C. 2008. Comparison of contamination models. Deliverable D6.2-3 of the Integrated Project “Fire Paradox”, Project no. FP6-018505. European Commission. 43 p. Millan, M., Estrela, M.J. and Badenas, C. 1998. Meteorological processes relevant to forest FIRE dynamics on the Spanish Mediterranean COAST. Journal of Applied Meteorology. 37: 83–100. Montserrat, D. 1998. La predicció dels Grans Incendis Forestals. Catalunya Rural i Agrària. 54: 5–13. Rifà, A. and Castellnou, M. 2007. El modelo de extinción de incendios forestales catalan. Proceedings of the IV International Wildfire Conference. Seville. Spain. Rigolot, E., Fernandes, P. and Rego, F. 2009. Managing Wildfire Risk, Prevention, Suppression. In: Birot, Y. (ed.) Living with wildfires, what science can tell us. EFI Discussion Paper 15. European Forest Institute. Pp. 49–52. Rothermel, R.C. 1991. Predicting behaviour and size of crown fires in the Northern Rocky Mountains. Res. Pap. INT-438. USDA Forest Service, Intermountain Research Station, Ogden, UT, USA. 46 p.

N. C. E. Vega.A. 2009. Jiménez. J.E. FP6-018505. European Commission.Wildfire Scenarios: Learning from Experience 133 Ruíz C. FP6-018505. G. J. and Senra F. . A. European Commission. Taylor. Junta de Anadalucia.Operational Report. 2009. Project no. Study of a 2nd set of largest past fires. Sagarzazu Ansaldo.3-2 of the Integrated Project “Fire Paradox”.R.3-1 of the Integrated Project “Fire Paradox”. and Skinner. 13: 704–709. 27 p. 15 p. and Defossé. Deliverable D8. The Casarabonela 2009/07/14 Fire . Project no. and Pérez. 82 p. Effectiveness of suppression fires: methods and first results. M.. Internal Report IR2. Spatial patterns and controls on historical fire regimes and forest structure in the Klamath Mountains. 2003.S. Ecological Applications.H. 2009.

.

4. Prescribed Burning: Preventing Wildfires and More .

.

Moreover. changing paradigms in nature and landscape policies in Atlantic and Central European countries have propitiated the development of prescribed burning initiatives in the frame of research projects (Goldammer et al. Thus. the use of fire for different management purposes remains rather limited. affecting even the more reluctant countries.4. objectives for the use of this technique have broadened towards more environmental oriented purposes as more experience has been accumulated both in prescribed burning research and management fields. Sweden and Finland) the use of prescribed burning is at present included as a management practice required by forest certification schemes and also for nature conservation of protected areas. However. Spain has also progressed in this direction. but mainly at the regional level.1.1 Overview of Prescribed Burning Policies and Practices in Europe and Other Countries Andrea Lázaro and Cristina Montiel Research Group Forest Policy and Socioeconomics. especially in those countries which were pioneers in its introduction. although still maintained at the experimental level. important changes have taken place for prescribed burning in the last years. The support of Fire Paradox project and the influence of the exchanges between fire professionals of Europe . Moreover. In fact. launched in southern countries in the 1980s (Botelho and Fernandes 1997). Twenty years after the first experimental initiatives of prescribed burning in Europe. the introduction and implementation of prescribed burning practices with different management objectives is a progressive trend in fire and land management policies. prescribed burning has been progressively included in nature conservation policy documents (Rigolot 2005).g. Portugal and France were the first European countries to initiate policy processes aiming to develop prescribed burning practices. The main constraints for prescribed burning introduction and implementation in the continent are fire bans. Catalonia and the Canary Islands have created professional teams in the use of fire for management purposes. Thus. 2007). and mainly for wildfire hazard reduction objectives. 2006).1 Introduction Prescribed burning is a technique that represents the controlled application of fire to vegetation under specific environmental conditions to attain planned resource management objectives (FAO 1986). lack of professional experience and negative public perception (Xanthopoulos et al. complex land structure. In Northern European countries (e. Spain 4. University Complutense of Madrid. based on specific legal frameworks and national systems for the professional accreditation in the use of this technique.

Goldammer et al. . we find it relevant to include also traditional fire use practices since it is considered as an important factor for the development of prescribed burning policies in the future. southern Italy (2009). For this purpose. or by (re)introducing prescribed burning as a modern technique for management purposes within a professional structure (see Box 1). with a few exceptions focusing on some European regions (Pyne 1997.138 Towards Integrated Fire Management and other continents have allowed the barrier against the use of prescribed burning to be overcome in countries which traditionally had maintained a negative attitude towards the use of prescribed burning. 1996) as well as the latter evaluation which determines whether the pre-determined management objectives have been reached and allow for future improvements (Fernandes 2002). The most prominent example in this sense is the development of a professional fire use team in Sardinia (2008) and the development of the first prescribed burning initiatives for fire prevention in the Cilento National Park. it is important to clearly differentiate between traditional burning practices and prescribed burning practices. 4. In this regard the main characteristic that clearly distinguishes prescribed burning from traditional fire use is an adequate planning (Pyne et al. 1 For more information on sources and methods see Lázaro et al. information was collected. that is either regulating traditional fire use and profiting from this experience to develop prescribed burning policies. and identification of the main factors that are key to promote these practices as well as to provide recommendations to overcome the principal constraints for the future. the first aim of fire use practices assessment within Fire Paradox project was to obtain a complete summary of spatial distribution of traditional and prescribed burning in Europe and North African countries. Thus. Botelho and Fernandes 1997. (2008). However when considering both types of practices.2 Inventory and classification of fire use practices The available information about the development and distribution of fire use practices in Europe is scattered and rather scarce. The research on prescribed burning practices and policies carried out within the frame of Fire Paradox has enabled a general overview of prescribed burning spatial patterns at the EU level to be obtained. 2007). based on a questionnaire distributed to national experts with responsibilities in forest and fire management1. At this point.1. and where the possibility to use prescribed burning for fire hazard reduction has similarly aroused technical and political interest. Besides. classified and referenced at the regional level for the EU-27 member states and the North African countries (Morocco and Tunisia). the research has considered neighboring countries in the southern Mediterranean basin by acquiring a general sense of their fire uses. identification of the different existing policy scenarios in European countries towards prescribed burning practices.

vegetation encroachment. 2004) and together with grazing and other traditional cultivation systems have become part of integral components which have shaped Mediterranean landscapes (Naveh 1975). 2000). In southern and eastern European countries.) have increased the risk of traditional fire use practices which have evolved to be at present the main cause for wildfire in Mediterranean countries (FAO 2007. hunting. In addition. In the Medite