You are on page 1of 63

DEGREE PROJECT IN ENVIRONMENTAL ENGINEERING,

SECOND CYCLE, 30 CREDITS


STOCKHOLM, SWEDEN 2020

Exploring Urban Rainwater


Harvesting in the city of
Madrid:
applying GIS based MCDA expert tools.

SILVIA BÖRJESSON

KTH ROYAL INSTITUTE OF TECHNOLOGY


SCHOOL OF ARCHITECTURE AND THE BUILT ENVIRONMENT
Exploring Urban Rainwater Harvesting
in the city of Madrid:
applying GIS-based MCDA expert tools

Silvia Börjesson

Supervisor:
Lina Suleiman

Examiner:
Maria Håkansson

Degree Project in Urban and Regional Planning


M.Sc. in Environmental Engineering and Sustainable Infrastructure
School of Architecture and Built Environment
Department of Urban and Regional Studies
SE-100 44 Stockholm, Sweden
Abstract...
Due to climate change, water resources’ scarcity and distribution variability have generated a
growing interest in sustainable water management in recent years. In addition, the growing
interest in implementing nature-based solutions for urban resilience leads to the development
of decentralized water supply systems such as rainwater harvesting systems (RWHS) as a
complementary resource to conventional centralized water supply systems. The study is
motivated by the climatic risks that the city of Madrid is subject to face in the upcoming years
such as pronounced summer droughts and urban heat waves, leading to the reduction of water
resources availability. This study aims to use Multi-criteria decision analysis and Geographic
Information systems as tools for locating optimal space for the installation of pond harvesting
systems (PHS) in Madrid for water recollection for non-potable purposes in the city.

Attaining this aim, two objectives were set. Firstly, to evaluate and select the most relevant
criteria for the installation of PHS and secondly, to generate a map of the most suitable
locations for installation of PHS in the city of Madrid through a proposed GIS-MCDA
methodology and a complimentary evaluation for each possible solution, to obtain a global
vision of the applicability of PHS in Madrid. Through the first objective, several criteria were
set for PHS installation, namely: distance to the river, rainfall, slope, soil characteristics, and
land use. These criteria were applied to Madrid through a GIS-MCDA methodology, using
these two tools’ synergy to obtain a suitability map for PHS installation. Two criteria weightings
will be performed to evaluate the model’s robustness by modifying the criteria’ weights
resulting in two different suitability scenarios. The discussion will analyze the results obtained
considering the two scenarios and propose the most suitable location clusters identified.
Finally, the conclusion will reflect the study’s most important findings and open the door to
further research on the topic, such as the design, operation infrastructure, drainage logistics
distribution, and other modifications in Madrid’s current water management system.

Keywords...
Rainwater Harvesting Systems, Pond Harvesting Systems, Nature-based facilities, Climate
Change Adaptation, Urban Resilience, Multicriteria Decision Analysis, GIS tools, ArcGIS,
Madrid, Spain, Sustainable water management solutions.

2
Sammanfattning...
Som en konsekvens av den pågående klimatförändringen och dess inverkan i distributionen
av vattenresurser har det skapats ett växande intresse för en hållbar vattenförvaltning de
senaste åren. Detta växande intresse leder till implementationen av naturbaserade lösningar
för urban resiliens och till en utveckling av decentraliserade vattenförsörjningssystem som
regnvattenuppsamling (RWHS) som en komplementerande resurs till konventionella
centraliserade vattenförsörjningssystem.

Huvudsyftet med denna studie är att täcka forskningsklyftan och att genom användning av
verktyg som multikriterieanalys och geografiskt informationssystem (GIS) lokalisera en
optimal plats för installation av dammar för regnvattenupptagning som del av stadsplanering
för att skapa en hållbar vattenförvaltning i staden.

För att uppnå detta, sattes två mål. Det ena var att utvärdera och välja de mest relevanta
kriterierna för installation av dammuppsamlingssystem (PHS), det andra målet var att
generera en karta över de lämpligaste platserna för installation av PHS i Madrid (Spanien)
genom en föreslagen GIS-MCDA-metod och en kompletterande utvärdering för varje möjligt
resultat i syfte att få en global vision av tillämpligheten av PHS i Madrid.

De främsta kriterierna som utsågs för att göra utvärderingen av uppsamlingssystems


installationen var: avstånd till floden, nederbörd, sluttning, markegenskaper och
markanvändning. Dessa kriterier tillämpades genom en GIS-MCDA-metodik, och med hjälp
av dessa verktygs synergi erhölls en lämplighetskarta för PHS-installationer. Två
kriterieviktningar kommer att utföras för att utvärdera modellens robusthet och vilket även
kommer att resultera i två olika lämplighetsscenarier.

Diskussionen kommer bestå av en analys av resultaten med hänsyn till de två scenarierna,
och föreslå de två lämpligaste lokaliseringsklustren som identifierats. Slutligen kommer
slutsatsen att återspegla studiens viktigaste resultat och öppna dörren för ytterligare forskning
i ämnet, såsom design, driftinfrastruktur, dräneringslogistikdistribution och andra modifieringar
i Madrids nuvarande vattenledningssystem.

3
Acknowledgments...
I want to dedicate this space to thank my thesis supervisor Lina for guiding me through this
thesis research, for her patience and understanding, and especially for her valuable feedback
and comments. This thesis could not have been possible without her flexibility and endurance.

Thanks to KTH for providing all the necessary resources to complete two years of an excellent
master’s program and for granting me experiences on a personal and academic level that I
will always carry with me. Finally, I would like to thank my friends and family for their love and
support, especially Clea and Juan, for lifting me up in the most challenging times. I hope that
the reader can learn and enjoy the content of this thesis as much as I did.

Silvia Börjesson.

4
Table of Contents...
INTRODUCTION................................................................................................................... 8
STUDY BACKGROUND... ....................................................................................................10
RAINWATER HARVESTING SYSTEMS: TERMS AND APPLICATIONS.................................................................................. 10
URBAN WATER MANAGEMENT IN THE CONTEXT OF M ADRID........................................................................................ 12
AIMS, OBJECTIVES, AND STRUCTURE................................................................................16
SCOPE AND LIMITATIONS ............................................................................................................................................ 16
STRUCTURE OF THE STUDY ......................................................................................................................................... 17
METHODOLOGY.................................................................................................................18
RESEARCH EVALUATION ............................................................................................................................................. 19
LITERATURE REVIEW FOR THE SELECTION OF SITE SUITABILITY CRITERIA .................................................................... 20
GIS APPLICATION FOR DETERMINATION OF POTENTIAL SITE LOCATION OF PHS IN M ADRID ......................................... 21
1. Raw spatial data collection ..................................................................................................................... 21
2. Creation of basic maps............................................................................................................................. 21
3. Determination of site suitability scores.................................................................................................. 21
4. Standardization of criteria maps............................................................................................................ 21
5. Criteria weighting..................................................................................................................................... 21
6. Map overlaying ......................................................................................................................................... 22
7. Suitability Map and Robustness Evaluation .......................................................................................... 22
RESULTS... ........................................................................................................................23
D ETERMINATION OF CRITERIA FOR SITE SUITABILITY SELECTION OF PHS .................................................................... 23
Factors ................................................................................................................................................................ 25
Constraints ......................................................................................................................................................... 25
D EVELOPMENT OF A LOCATION SUITABILITY MAP FOR PHS IN M ADRID ....................................................................... 27
1. Raw spatial data collection ..................................................................................................................... 27
2. Creation of basic maps............................................................................................................................. 27
3. Standardization and determination of site suitability scores.............................................................. 31
4. Creation of factor and constraint maps ................................................................................................. 31
5. Criteria weighting..................................................................................................................................... 36
6. Map overlaying ......................................................................................................................................... 37
7. Final Suitability Map................................................................................................................................ 39
ROBUSTNESS E VALUATION......................................................................................................................................... 40
DISCUSSION... ...................................................................................................................43
UNCERTAINTIES AND RISKS......................................................................................................................................... 49
CONCLUSIONS... ...............................................................................................................50
REFERENCES... .................................................................................................................52
ANNEXES... .......................................................................................................................57
ANNEX I..................................................................................................................................................................... 57
ANNEX II.................................................................................................................................................................... 57
ANNEX III ................................................................................................................................................................ 59
ANNEX IV ................................................................................................................................................................ 61

5
List of Figures
Figure 1: Roof Harvesting Systems (RHS) ........................................................................ 11
Figure 2: Pond Harvesting Systems (PHS)........................................................................ 11
Figure 3: PHS example in Augustenborg........................................................................... 12
Figure 4: Example of RWHS in the BBVA Headquarters.................................................... 14
Figure 5: Example of RWHS in the Park between C/Alfonso XII and C/Paraguay ............... 14
Figure 6: RWHS example in La Atalayuela green park ...................................................... 14
Figure 7: Methodology flowchart ....................................................................................... 19
Figure 8: List of Keywords used for the identification of relevant studies ............................ 20
Figure 9: Outline identifying the main green spaces in the city of Madrid............................ 26
Figure 10: Workflow of the Model Builder in ArcGIS .......................................................... 27

List of Tables
Table 1: Selected criteria for rainwater harvesting according to different authors................ 23
Table 2. Model Builder environments properties................................................................ 28
Table 3: Land use attribute table reclassification values .................................................... 34
Table 4: Factor ranking positions ...................................................................................... 36
Table 5: Pairwise comparison for the selected factors ....................................................... 37
Table 6: Weights assigned to the selected factors ............................................................. 37
Table 7: Map overlaying weight and suitability scores description for each factor ............... 38
Table 8: Numerical Suitability values for installation of PHS in Madrid................................ 40
Table 9: Factor ranking positions for the robustness evaluation ........................................ 40
Table 10: Pairwise comparison for the robustness evaluation ............................................ 41
Table 11: Weights assigned for the robustness evaluation................................................. 41
Table 12: Numerical Suitability values for Suitability Map 2................................................ 43

List of Maps
Map 1: Rain map interpolation using Kriging ..................................................................... 30
Map 2: Rain map interpolation using IDW ......................................................................... 30
Map 3: Factor Rain Map ................................................................................................... 33
Map 4: Factor Slope Map.................................................................................................. 33
Map 5: Factor Soil Map..................................................................................................... 33
Map 6: Factor distance to river Map .................................................................................. 33
Map 7: Land use constraints Map ..................................................................................... 35
Map 8: Suitability Map for PHS in the city of Madrid........................................................... 39
Map 9: Suitability Map nº2 based on robustness evaluation............................................... 42
Map 10: PHS Suitability Map - Northern clusters ............................................................... 44
Map 11: PHS Suitability Map - Southern clusters............................................................... 46
Map 12: PHS Suitability Map 2 - Northern clusters ............................................................ 47
Map 13: PHS Suitability Map 2 - Southern clusters............................................................ 48

6
Acronyms and abbreviations
AEMET Spanish State Meteorological Agency
AHP Analytical Hierarchy Process
CNIG Spanish National Geographic Information Centre
CR Consistency Ratio
ETRS European Terrestrial Reference System 1989
GIS Geographic Information Systems
IDW Inverse Distance Weighting
IGME Geological and Mining Institute of Spain
MAPAMA Spanish Ministry of Agriculture and Fisheries, Food and Environment
MCDA Multi-Criteria Decision Analysis
PCM Pairwise Comparison Matrix
PHS Pond Harvesting Systems
RHS Roof Harvesting Systems
RWHS Rainwater Harvesting Systems
UTM Universal Transverse Mercator
WO Weighted Overlay

7
Introduction...
limate change is the primary environmental threat the world is facing today, and cities
play an essential role as one of the leading causes of greenhouse gas emissions and,
at the same time, the most vulnerable to the adverse effects of climate change. Even though
cities represent only 2% of the planet’s surface, they are responsible for consuming more than
75% of the natural resources and 80% of the greenhouse gasses emitted into the
atmosphere 1. On top of this and according to the U.N. 2, the overall growth of the world’s
population could add another 2.5 billion people to urban areas by 2050. Given the unstoppable
processes of urbanization and the high number of people living in cities, the pressure to
introduce actions to mitigate and adapt to the effects of climate change in urban environments
is growing both at a legislative and infrastructure level.

Measures and strategies taken towards the fight against climate change can be divided into
two main pillars: mitigation and adaptation 3. Mitigation measures comprise strategies related
to limiting the intensity of climate change by reducing CO2 emissions. In contrast, adaptation
to climate change is defined as a set of strategies to reduce the vulnerability of both natural
(forests, farmlands, coastal areas) and anthropogenic systems (cities, economy, societies)
from the effects of climate change 4.

All cities will face some impact due to climate change. Some of the main risks that affect
European cities are the general increase in temperatures, heat waves, extreme precipitation
events, droughts, and an increase in sea level 5. Firstly, a warmer climate and heat waves
could translate into increased mortality and morbidity 6 and a challenge for public health. Other
impacts include floods derived from rising sea levels (most likely in central, northern, and
southern Europe) 7and a water increase in the river courses (most likely in central Europe).
Subsequently, climate change could also lead to the spread of new viruses and a possible
increase in the risk of accidents due to extreme events such as storms and floods 8. All these

1 Fernandez García, F. & Martilli, A., 2021, ‘El clima urbano: aspectos generales y su aplicación en el
área de Madrid’, Revista Índice (50), p 21.
2 United Nations, 2018, ‘Revision of World Urbanization Prospects’, Department of Economic and Social

Af fairs, United Nations: New York, NY, USA


3 Oltra, C & Marín, R., 2013, ‘Los retos en la adaptación al cambio climático en entornos urbanos’, ISSN

2013-9004, pp. 311-330


4 Smit, B. & Wandel, J. ,2006, ‘Adaptation, adaptive capacity and vulnerability’, Global Environmental

Change, 16, pp. 282-292.


5 Oltra, C & Marín, R., 2013, ‘Los retos en la adaptación al cambio climático en entornos urbanos’, ISSN

2013-9004, p. 5
6 Sari Kovats, R. & Hajat, S., 2008, ‘Heat Stress and Public Health: A Critical Review’, Public and

Environmental Health Research Unit (PEHRU), London School of Hygiene and Tropical Medicine,
London WC1E 7HT, Annual Review of Public Health Vol. 29, pp. 41-55
7 Oltra, C & Marín, R., 2013, ‘Los retos en la adaptación al cambio climático en entornos urbanos’, ISSN

2013-9004, pp. 311-330


8 Ciscar, J.C., Soria, A., M. Goodess, C. B. Christensen, O., Iglesias, A., Garrote, L., Moneo, M.,

Quiroga, S., Feyen, L., Dankers, R., Nicholls, R. & Richards, J., 2009. "Climate change impacts in
Europe. Final report of the PESETA research project," JRC Working Papers JRC55391, Joint Research
Centre (Seville site), pp. 31-38

8
impacts will have repercussions on various aspects such as infrastructure, public health, the
well-being of the population, and the management of resources.

In the case of southern European and Mediterranean cities, two major threats have more
likelihood of occurrence: droughts and floods9. Trends and evidence regarding these risks are
clear. On the one hand, droughts will present greater frequencies and longer spatial and
temporal duration, reducing the water supply capacity for the population 10. On the other hand,
floods tend to increase, meaning a greater probability of torrential storms with high intensity
and precipitation volumes, generating overflow problems in urban areas 11.

Considering this new climatic reality, Southern European cities must be prepared to withstand
both long months of drought and torrential rains. This scenery requires hydraulic planning of
“extremes” 12 that can guarantee water supply in long periods of drought and significant urban
demands.

Traditionally, conventional water planning engineering structures such as dams, reservoirs,


channels, and pipelines (among others) have been applied to alleviate the adverse effects of
droughts and floods in Mediterranean urban areas 13. Recently, other measures have been
incorporated into sustainable water management, such as flood parks, river restoration
methods, rainwater harvesting systems, or the adaptation of urban design to these
atmospheric extremes 14.

Many diverse terms have been used to refer to sustainable water management and alternative
solutions to manage urban rain, such as sustainable urban drainage systems, best
management practices, urban rainwater harvesting systems, water sensitive urban design15,
blue infrastructures, and nature-based solutions. In this study, rainwater harvesting systems
refer to nature-based sustainable water management through green and blue infrastructures
for cities 16.

9
Oliva Cañizares, A, Sánchez Almodóvar, E & Marcos Palacios, M.J., 2021, ‘La adaptación a los
extremos atmosféricos y al cambio climático mediante los sistemas urbanos de drenaje sostenible
(SUDS) y soluciones basadas en la naturaleza (SBN): propuesta en el término municipal de Callosa
de Segura (Alicante, España)’, Universitat d’Alacant, pp. 9-20.
10 Oliva Cañizares, A, Sánchez Almódovar, E & Marcos Palacios, M.J., 2021, ‘La adaptación a los

extremos atmosféricos y al cambio climático mediante los sistemas urbanos de drenaje sostenible
11 AEMET, Agencia Estatal de Meteorología, 2021, ‘Proyecciones climáticas para el siglo XXI’,

http://www.aemet.es/es/serviciosclimaticos/cambio_climat
12 Oliva Cañizares, A, Sánchez Almodóvar, E & Marcos Palacios, M.J., 2021, ‘La adaptación a los

extremos atmosféricos y al cambio climático mediante los sistemas urbanos de drenaje sostenible
(SUDS) y soluciones basadas en la naturaleza (SBN): propuesta en el término municipal de Callosa
de Segura (Alicante, España)’, Universitat d’Alacant, pp. 9-20.
13 Olcina Cantos, J. Oliva Cañizares, A., 2020, ‘Medidas estructurales versos cartografía de inundación

en la valoración del riesgo en áreas urbanas: el caso del barranco de las Ovejas (Alicante, España)’,
Cuadernos Geográficos, 59 (2), pp. 199-220.
14 Olcina Cantos, J, 2019, ‘Evidencias e incertidumbres del Cambio Climático y de los riesgos asociados

en el litoral mediterráneo español’, Boletín de la Real Sociedad Geográfica, CLIV, pp. 9-34
15
Suleiman, L., Olofsson, B, Saurí, D. & Palau-Rof, L., 2020, ‘A breakthrough in urban rain-harvesting
schemes through planning for urban greening: Case studies from Stockholm and Barcelona’,
ELSEVIER vol. 51, pp. 2-4
16
Agudelo-Vera, C. M., Leduc, W. R. W. A., & Melsa, A. R., 2012, ‘Harvesting Urban Resources
Towards More Resilient Cities’, Resources Conservation and Recycling 64, pp: 3–12.

9
Rainwater harvesting systems are probably the most ancient practice to cope with water
supply needs 17. As a result of new technologies, many countries are implementing modernized
methods of rainwater harvesting systems to undertake water demand pressures mainly
associated with the environmental, climatic, and societal changes cities are experiencing 18.

This study is based on the need to combine measures to cope with the predicted droughts
that will affect South European urban areas and, in particular, Madrid. In this framework, this
study will contemplate implementing urban rainwater harvesting systems in the city based on
the hypothesis of water shortage forecasts in Madrid.

Study Background...
Rainwater Harvesting Systems: terms and applications

ainwater Harvesting Systems (RWHS) allow the collection and storage of rainwater
from rooftops, terraces, courtyards, and other impervious surfaces for on-site use19.
They are considered one of the most promising ways to improve water supply in regions with
present or predicted water shortage problems. Many articles have addressed the multiple
benefits RWHS can provide in densely populated cities, such as flood control, aesthetic and
leisure value, being more adapted to local conditions, and reducing water use 20. RWHS are
increasingly perceived as a safe alternative for drinking water supply in developing countries,
while in developed countries, RWHS are mainly used for non-potable uses 21. In this context,
the implementation of RWHS is considered an accomplice in the fight against climate change,
either due to water harvesting for upcoming periods of droughts or to the mitigation of floods
that may cause both health and physical risks to humans and ecosystems 22.

RWHS are generally implemented through rainwater dams, water retention ponds, rooftop
harvesting systems, water storage tanks, percolation tanks, runoff strips, water reservoirs, and
cisterns, among others 23. According to Zabidi, H. A. (2020), the most commonly used RWHS

17
Campisano, A., Butler, D., Ward, S., J. Burns et al., 2017, ‘Urban rainwater harvesting systems:
Research, implementation and future perspectives’, Water Research 115, pp. 195-209.
18 Amos, C., Rahman, A. & Gathenya, J.M., 2016, ‘Economic analysis and f easibility of rainwater

harvesting systems in urban and peri-urban environments: a review of the global situation with a special
f ocus on Australia and Kenya’, Water 8 (4), p. 149.
19
Campisano, A., Butler, D., Ward, S., J. Burns, et al., 2017, 'Urban rainwater harvesting systems:
Research, implementation and future perspectives', Water Research 115, pp. 195-209.
20
García Soler, N., Moss, T. & Papasozomenou, O., ‘Rain and the city: Pathways to mainstreaming
rainwater harvesting in Berlin’, ELSEVIER vol. 89, pp. 96-100
21
Suleiman, L., Olofsson, B, Saurí, D. & Palau-Rof, L., 2020, ‘A breakthrough in urban rain-harvesting
schemes through planning for urban greening: Case studies from Stockholm and Barcelona’,
ELSEVIER vol. 51, pp. 2-4
22
Montañez Gil, S.J., 2014, ‘Desarrollo de una herramienta computacional para priorizar la ubicación
de tanques de recolección de agua de escorrentía en lugares con riesgo de inundación: Caso Bogotá’,
Universidad de los Andes, facultad de ingeniería civil y ambiental, pp. 8
23 Zabidi, H.A., Weng Goh, H., Kiat Chang, C., Weng Chan, N. & Azazi Zakaria, N., 2020, ‘A Review of

Roof and Pond Rainwater Harvesting Systems for Water Security: The Design, Performance and Way
Forward’, Water 2020 12(11), pp. 3-22

10
in urban areas can be divided into two main categories: roof harvesting systems (RHS) and
pond harvesting systems (PHS). In both cases, urban uses for the collected water are non-
potable and vary from toilet flushing, laundry, irrigation, terrace leaning, and other cleaning
uses to reduce drinking water consumption from centrally supplied sources.24 RHS are more
common and have been applied previously in Spain, mainly regarding Spanish Mediterranean-
weather environments 25. In the case of RHS, the water collection takes place in open or closed
tanks, in some cases above ground (Figure 1) and others in underground buried storage tanks.

Figure 1: Roof Harvesting Systems (RHS) Figure 2: Pond Harvesting Systems (PHS)

In the case of PHS, the storage is made in an open natural reservoir and harvests the water
from the land surface or directly from the rainfall. Hence, it needs a catchment surface where
the runoff water can easily be captured, such as public areas or common land, using
gravitational force.

Studies 26 show how the installation of urban rainwater harvesting systems increases the water
self-sufficiency of cities and helps to delay the need to construct new centralized water
infrastructures. This study will focus on the location suitability of PHS, intended to be a
reservoir for flowing water during rainy season peaks and for non-potable uses such as
irrigation to improve the sustainable management of urban water resources. An image of a
PHS example in Augustenborg, Malmö can be found in Figure 3 below.

24
Butler, D., Ward, S., J. Burns, M., Friedler, E., DeBusk, K., N. Fisher-Jeffes, L., Ghisi, E., Rahman, A.,
Furumai, H., Han, M. & Campisano, A., 2017, ‘Urban rainwater harvesting systems: Research,
implementation and future perspectives’, p. 2
25
Farreny, R., Morales-Pinzón, T., Guisasola, A., Tayá, C., Rieradevall, J. & Gabarrell, X., 2011, ‘Roof
selection f or rainwater harvesting: Quantity and quality assessments in Spain’, Water Research Volume
45, Issue 10, pp. 3245-3254
26 Stef fen, J., Jensen, M., Pomeroy, C.A., Burian, S.J., 2012, ‘Water supply and stormwater

management benefits of residential rainwater harvesting in U.S. cities’, J. Am. Water Resour. Assoc.
49 (4), pp. 810-824

11
Figure 3: PHS example in Augustenborg

Source: https://climate-adapt.eea.europa.eu/metadata/case-studies/urban-storm-water-management-in-
augustenborg-malmo

Urban water management in the context of Madrid


The city of Madrid constitutes a particularly vulnerable urban environment due to its population
size that accounts for 3.334.730 inhabitants in 2020 27 and regional climatic conditions
characterized by sweltering summers and cold winters. On top of this, Madrid is located in the
center of the Iberian Peninsula, where the effects of Climate Change will be especially severe,
according to the European Environment Agency 28, with an increase in heat waves and
summer hydric stress.

In this context, the City of Madrid’s Energy and Climate Change Plan 29 identifies the main
climatic trends and impacts that the city is expected to face over the following decades. These

27
Instituto Nacional de Estadística (INE), https://www.ine.es/jaxiT3/Datos.htm?t=2881. Accessed
07/10/2020
28
EEA ,2012, ‘Urban adaptation to climate change in Europe. Challenges and opportunities for cities
together with supportive national and European policies’, EEA Report No 2/2012, pp 148
29
Ayuntamiento, D.M., 2014, ‘Plan de Energía y Cambio Climático de la ciudad de Madrid-Horizonte
2020’, Ayuntamiento de Madrid, pp. 24-30

12
include a predicted generalized decline of average annual precipitation, meaning fewer rainy
days and more extreme precipitation events, an increase in the monthly mean temperatures
and daily minimums, and an increase in the monthly mean maximum temperatures, meaning
an increase in the maximum temperatures and a decrease in the number of frost days, with a
consequent increase in the duration of heatwaves.

Concurrently, and because of these changes in the climate conditions, there will be shortages
of resources such as water. As a result of this shortage, the monetary cost of this resource
will increase, as will its opportunity cost since it cannot be used to the detriment of human
consumption in the upcoming times of extended droughts.30

In this scenario, it is imperative to deeply study the adoption of sustainability criteria in the
management of water resources in Madrid and work together from all its areas and expertise
in incorporating adaptation strategies against climate change. As mentioned above, the trend
towards an increase in temperature in the city and a decrease in average rainfall is a fact that
could soon become a cause of genuine concern if the appropriate measures are not taken. 31

Developments in changing the approach to water management in Madrid started purposely in


2006, when the Council of the City of Madrid published the Management and Efficient Use of
Water Ordinance 32, establishing among other criteria the minimum percentage of permeable
surfaces in open spaces, both public and private. Subsequently, many municipal documents33
34 advocating sustainable water management came forth mainly in green areas. More recently,

the Madrid + Natural program (inherent to the national strategic plan for Air Quality
Improvement) follows the current trend towards the benefits of using rainwater harvesting
systems in cities, being one of the recommended solutions for the better management of water
resources.

Some practical examples related to RWHS in Madrid are the BBVA Headquarters, the park
between Calle Alfonso XII and Calle Paraguay, and the Atalayuela green park. These three
examples apply different types of RWHS for sustainable water management. In the first case,
BBVA Headquarters integrate permeable areas to accumulate water in buried infiltration tanks,
as shown in Figure 4. Regarding the park between the Streets of Alfonso XII and Paraguay,
water gardens equipped with buried infiltration tanks and filter drains were incorporated, as
shown in Figure 5. The case of Atalayuela green park was a novel case for the city, intended
to incorporate RWHS in urban environments and enable a design focused on preserving
natural patterns and increasing vegetated and permeable surfaces in the park. For this

30
Ayuntamiento, D.M., 2018, ‘Bases y Directrices generales del plan de inf raestructura verde y
biodiversidad de la ciudad de Madrid’, Plan de infraestructura verde y biodiversidad, Ayuntamiento de
Madrid, 2018
31
Ayuntamiento, D.M., 2018, Resumen Ejecutivo del Diagnóstico de Situación del Plan Estratégico,
Plan de inf raestructura verde y biodiversidad, Ayuntamiento de Madrid, 2018
32
Ayuntamiento de Madrid, 2006, ‘Ordenanza de Gestión y Uso Ef iciente del Agua en la Ciudad de
Madrid’, Ayuntamiento de Madrid 22/06/2006 num. 5709, pp. 2410-2443.
33
Ayuntamiento de Madrid, 2007, ‘Criterios para una jardinería sostenible en la ciudad de Madrid’, Área
de Gobierno de Medio Ambiente y Servicios a la Ciudad
34
Ayuntamiento de Madrid, 2009, ‘Buenas Prácticas en Arquitectura y Urbanismo para Madrid. Criterios
bioclimáticos y de eficiencia energética’, Área de Gobierno de Urbanismo y Vivienda del Ayuntamiento
de Madrid.

13
purpose, rain gardens were introduced along the walkways, as shown in Figure 6, with filtering
drains to accumulate water for irrigation. In all three cases, a reduction of the discharge volume
to the city’s sanitation systems was estimated between 83% 35 and 92% 36, compared to
conventional grey infrastructure.

Figure 4: Example of RWHS in the BBVA Headquarters

Figure 5: Example of RWHS in the Park between C/Alfonso XII and C/Paraguay

Figure 6: RWHS example in La Atalayuela green park

Source: The Basic Guide for the Design of Sustainable Stormwater Management systems in Green Zones and
other Public Spaces (2018) developed by the Madrid City Council.

Furthermore, the Green Strategy and Biodiversity Plan (2017) determines among its lines of
action the incorporation of RWHS in newly built infrastructures such as roads and city plazas
and on infrastructures undergoing or subject to reformation works to improve the management
of water runoff, reduce water input to irrigation, and improve the maintenance and
conservation of green areas.

35
Andrés-Valeri, V. C., Perales-Momparler, S., Sañudo-Fontaneda, L. A., Andrés-Domenech, I., Castro-
Fresno, D., & Escuder-Bueno, I., 2016, ‘Sustainable Drainage Systems in Spain’, En Wiley-Blackwell
(Ed.), Sustainable Surface Water Management: A Handbook for SUDS, pp. 355-370
36
Ayuntamiento de Madrid, 2018, ‘Guía Básica de Diseño de sistemas de Gestión Sostenible de
Aguas Pluviales en Zonas Verdes y otros Espacios Públicos’,Área de Gobierno de Medio Ambiente y
Movilidad, Dirección General de Gestión del Agua y Zonas Verdes, pp. 5

14
The National Energy and Climate Change Plan (2020) 37 and in line with the city plans such
as Madrid 360 (2019) 38 , stresses, among other proposed actions, the need to incorporate
Rainwater Harvesting Systems in the urban environment as one of the main pillars for
sustainable management of urban water resources. In this context, identifying potential sites
for installing urban rainwater harvesting systems is a crucial step towards maximizing water
availability in urban areas. 39

Rainwater harvesting systems contribute to runoff management in situ and reinforce


governance regulations from Spanish and European legislation 40 . These regulations address
sustainable water management as measures for the mitigation and adaptation to the
consequences of climate change, protection against floods and droughts, and the
reinforcement and support of the need to integrate sustainable water solutions in cities.

One of the main challenges in including urban water harvesting systems in general, and for
the case of this study of PHS, is to identify the most fitting locations to maximize the
advantages these nature-based facilities can offer. In this context, GIS tools for identifying
optimum sites for water harvesting schemes have been addressed in many studies, playing
an essential role in analyzing optimal locations: reducing time, cost and providing digital data
banks for similar dilemmas 41.

Finding the most suitable site for PHS installation is a complex study that includes social,
environmental, legislative, economic, and technical considerations. The inclusion of such
diverse parameters makes site location a crucial step in the decision-making process for the
success of RWHS 42. Spatial planning dilemmas that require the consideration of multiple and
complicated criteria can be solved through a GIS-based Multi-Criteria Decision Analysis
(MCDA).

This study presents a geographic information system (GIS)-based Multi-Criteria Decision


Analysis (MCDA) framework to identify potential sites for the installation of pond harvesting
systems (PHS) based on biophysical attributes (criteria) in the city of Madrid, Spain. The
results of the suitability analysis will be given in map format, and two criteria weightings will be
performed to evaluate the model’s robustness by modifying the criteria’ weights resulting in
two different suitability map scenarios.

37
Ministerio para la Transición Ecológica, 2020, ‘Plan Nacional Integrado de Energía y Clima (PNIEC)’,
Gobierno de España
38
Área de Gobierno de Medio Ambiente y Movilidad, 2019, ‘Avance de la Estrategia de Sostenibilidad
Ambiental - Madrid 360, Gobierno de Madrid
39
Mbilinyi, B.P., Tumbo, S.D., Mahoo, H.F. & Mkiramwinyi, F.O., 2007, ‘GIS-based decision support
system for identifying potential sites for rainwater harvesting‘, Elsevier Journal of Physics and Chemistry
of the Earth, 32, pp. 1074-1081.
b
Ukiriguru National Agricultural Research Centre, Mwanza, Tanzania Available online 3 August 2007
40 Ayuntamiento de Madrid, 2018, ‘Guía Básica de Diseño de sistemas de Gestión Sostenible de Aguas

Pluviales en Zonas Verdes y otros Espacios Públicos’, Área de Gobierno de Medio Ambiente y
Movilidad, Dirección General de Gestión del Agua y Zonas Verdes, pp. 5
41 Al-shabeeb, A., 2016, ‘The Use of AHP within GIS in Selecting Potential Sites for Water Harvesting

Sites in the Azraq Basin—Jordan’, Journal of Geographic Information System 8, pp. 73-88.
42 Al-Adamat, R., Al-Ayyash, S., Al-Amoush, H., Al-Meshan, O., Rawajifih, Z., Shdeifat, A. & Al-Fajarat,

M., 2012, ‘The combination of indigenous knowledge and geo-informatics for water harvesting sittin in
the Jordanian Badia’, Journal of Geographic Information Systems, 4, pp. 366-376.

15
Aims, objectives, and structure...
s stated above Madrid presents a series of risks derived from climate change, such
as changes in the distribution and recurrence of rainfall and rise in average
temperatures, resulting in pronounced summer droughts and urban heat waves, leading to the
reduction of water resources availability.

In this context, the city has been adopting specific legislation throughout the years to
incorporate sustainable water management practices in the city. Among the contemplated
measures the city of Madrid established a minimum percentage of permeable surfaces in open
spaces (both public and private), encouraging the collection of rainfall for both aquifer recharge
and non-potable uses, the inclusion of RWHS for the better management of water resources,
and implementing correct runoff management, limiting the amount of water discharge to
sewers.

Even though sustainable water management is addressed in the city at both a practical and
legislative level, decision support tools for optimally located PHS are still not widely applied
and are less covered in Madrid by academic and urban planning research. The study is
motivated by the climatic threats that the city of Madrid is subject to face in the following years
and aims to use Multi-criteria decision analysis and Geographic Information systems as tools
for locating optimal space for the installation of PHS in Madrid for water recollection for non-
potable purposes, and to serve as input to urban planning for sustainable water management
in the city.

To accomplish this aim, the study focuses on two main objectives, namely:
1. Reviewing relevant literature to urban planning and sustainable water management for
determining the criteria of site selection for PHS
2. Applying the identified criteria for generating suitability maps for optimal spatial
locations of PHS facilities in the city of Madrid

Finally, a robustness analysis of the decision rules in the MCDA will be applied by modifying
the criteria weighting to analyze the research subjectivity.

Scope and limitations


The study focuses on identifying the most optimal clusters for the implementation of PHS in
the city of Madrid through GIS-MCDA tools. However, no site visits were carried out within the
scope of this study to ensure PHS technical viability in the identified clusters. Moreover, this
study does not look into the design parameters, operation infrastructure, or drainage logistics
of the PHS, which should be contemplated for the materialization of the project. It should also
be noted that the results of this study are framed in the context of the city of Madrid, but the
applied methodology can be replicated to other cities with similar challenges.

16
Structure of the study
The study will be structured as follows:

The content can be divided into four defined points. Firstly, a description of the methodology
will describe the main steps to organize and conduct the study. The methodology and the
results will address each of the objectives separately since the first objective will be met
through qualitative research and the second through quantitative and practical research.

The second point focuses on the first objective of this study. For this, a desk-based collection
and qualitative analysis of practical cases, tools, and methodologies related to the spatial
suitability location of pond harvesting systems will take place, evaluating how these
technologies have been studied on a global level. Through this analysis, several spatial criteria
will be selected for the third step of the study.

The third step applies the knowledge from the second step to extrapolate it to the specific case
of Madrid. It will include information describing the obtention, manipulation, and analysis
applied through GIS and MCDA for the defined criteria.

Finally, the conclusion serves as a closing to summarize main insights, results, errors, and
gaps or improvement points.

17
Methodology...
he present study has been developed using qualitative and quantitative data analysis
under an exploratory research framework. The primary focus of this study is to use a
spatial Multi-Criteria Decision Analysis (MCDA) embedded in a geographical information
system (GIS) developed by the software ArcGIS (ESRI) for the creation of a site suitability
map for the implementation of urban pond harvesting systems (PHS) in the city of Madrid.

GIS tools allow the creation, analysis, management, and mapping of many types of data,
therefore simplifying the search for the suitable site selection of PHS. The application of Multi-
Criteria Decision Analysis (MCDA) has the advantage of combining factual, quantitative
information with expert opinion, therefore integrating a system that transforms spatial and non-
spatial information into a tool to aid the decision-making process 43.

Along with this, the analytical hierarchy process (AHP) as an MCDA method was introduced
by Saaty 44 and has been recognized as the most applicable decision method for identifying
the location of PHS 45. The main idea behind the AHP process is to apply hierarchical
structures to represent the components on any issue and show the links between them, in this
case, applied to the suitable location of PHS.

GIS and MCDA for planning purposes have been widely used among authors. However,
scientific studies and publications regarding the site selection for open rainwater harvesting
systems within an urban area are still rarely done. It is essential to mention that throughout
this study, urban pond harvesting systems (PHS) will be addressed as an open spatial
rainwater harvesting facility for non-potable uses, differentiated from tanks or other forms of
closed rainwater harvesting systems 46.

The methodology is formulated from the description and analysis of the main topic,
establishing the research questions to define the exploratory and output objectives. As stated
above, the main objectives of the study are (i) Reviewing relevant literature to urban planning
and sustainable water management for determining the criteria of site selection for urban pond
harvesting systems and (ii) Applying the identified criteria for generating suitability maps for
optimal spatial locations of PHS in the city of Madrid. To develop the study, the methodology
was divided according to the two main objectives. The first objective was to determine and
limit the criteria to meet the needs of urban pond harvesting systems that will take place

43 Gbanie, S.P., Tengbe, P.B., Momboh, J.S., Medo, J., Tamba, V. & Kabba, S., 2013, ‘Modelling landfill

location using Geographic Inf ormation Systems (GIS) and Multi-Criteria Decision Analysis (MCDA):
Case study Bo, Southern Sierra Leone’, ELSEVIER, pp 3.
44 Saaty, T.L., 1994, ‘How to make a decision: The analytic hierarchy process’, Interfaces 1994, 24, pp.

19 - 43
45 Doulabian, S., Ghasemi Tousi, E., Aghlmand, R., Alizadeh, B., Ghaderi Bafti, A. & Abbasi, A., 2021,

‘Evaluation of Integrating SWAT Model into a Multi-Criteria Decision Analysis towards Reliable
Rainwater Harvesting Systems’, Water 2021, 13, p. 1935.
46 Butler, D., Ward, S., J. Burns, M., Friedler, E., DeBusk, K., N. Fisher-Jeffes, L., Ghisi, E., Rahman,

A., Furumai, H., Han, M. & Campisano, A., 2017, ‘Urban rainwater harvesting systems: Research,
implementation and future perspectives’, pp 2

18
through qualitative desk-based methods. The second step incorporates the quantitative
methods, including the primary processing or collecting and adapting the data related to the
previously selected criteria to obtain the location suitability map for de PHS. The steps and
objectives of the applied methodology can be found in the flowchart below:

Figure 7: Methodology flowchart

Research evaluation
An exhaustive qualitative data analysis was carried out on a local, national, and global scale
to correctly address the objectives. It is relevant to mention that the literature review
methodology lacks certain reliability 47 since it is hard to replicate under the same parameters
and obtain the same results due to its dependence on the researcher’s subjectivity. To mitigate
this reliability issue, the references provided along with the document and the Reference
section provide the reader with the means to assess the trustworthiness of the present study48.

47 Recker, J., 2013, ‘Scientific research in information systems: A beginner’s guide’, https://doi.org/10.1007/97 8-3- 64 2-3 00 48- 6

48 Bhattacherjee, A., 2012, ‘Social Science Research: principles, methods, and practices’, Textbooks collection (Vol. 9).

19
Furthermore, the validity of results has also been identified as a challenge to this study
concerning the qualitative and quantitative applied methodology because of its requirements
in aspects such as correctness, truthfulness, and statement strength 49. Therefore, it is
essential to mention that the quality of the present study relies on the interpretation of the data
collected and attends to two types of validity: external and internal. The external validity was
conducted through a robustness analysis to determine how much the decision maker’s
subjective preferences affect the suitability results. On the other hand, the internal validity was
approached by providing enough evidence to justify the performed interpretations throughout
the study.

Literature review for the selection of site suitability criteria


For this study, the criteria are the basis for evaluating the different variables to consider for
the decision-making process. A desk-based qualitative study method was conducted to
identify relevant datasets. The qualitative information was obtained from different sources of
secondary information, academic or scientific articles, or web pages. To determine the criteria
used in similar studies, a search was conducted through google scholar using the following
keywords:

● Urban Pond Retention Systems

● Urban Rainwater Harvesting (URH)


● Spatial Multi-Criteria Decision Analysis

● Suitable Site location for RWHS


● GIS Suitability Analysis in ArcGIS

● Foundations of Urban Ponds

● Decision Support Systems in GIS

Figure 8: List of Keywords used for the identification of relevant studies

Based on what the academic community formulated in terms of site suitability criteria for
rainwater harvesting systems and, to narrow down the studies to be analyzed, papers were
selected based on their relevance to the present study. Since the study aims to obtain the
most suitable locations for installing pond harvesting systems, the scope of research was
limited to those studies that addressed issues related to open rainwater harvesting. To meet
the first objective of the study, the following steps were applied:

● Identifying main sets of site-selection criteria


● Categorizing and comparing the main site-selection criteria
● Selection and evaluation of site-selection criteria

49 Bhattacherjee, A., 2012, ‘Social Science Research: principles, methods, and practices’, Textbooks collection (Vol. 9).

20
GIS application for determination of potential site location of PHS
in Madrid
A seven-step quantitative methodology was applied to meet the study’s second objective. The
present study’s data analysis and processes were carried out using ArcMap 10.8.1. and its
Model Builder function. The Model Builder was applied to create the multi-step modeling tool
using a series of software command elements (models) and links with data elements to
perform the suitability analysis of the area. To better understand the proposed methodology,
it is necessary to restate that it is intended for the case in which rainwater is collected directly
from the water ponds of the selected area.

The applied seven-step methodology is as follows:

1. Raw spatial data collection


This stage consists of recollecting all the spatial data of the selected area (in this case, Madrid)
previously selected to determine the site suitability criteria step of the study.

2. Creation of basic maps


This second stage consists of the management and creation of basic maps from the selected
spatial data using ArcMap 10.8.1. Moreover, the data will be prepared for its evaluation
through its spatial analyst tools.

3. Determination of site suitability scores


This step consists of establishing the suitability scores for the emplacement of urban pond
harvesting systems. This determination of site suitability scores will, later on, be used to
standardize the scoring of each factor map to allow spatial overlaying.

4. Standardization of criteria maps


Once the suitability scores are determined, the criteria maps are standardized based on their
suitability scoring for PHS implementation.

5. Criteria weighting
The application of MCDA implies the need to assign importance values to each of the
considered factors, and therefore represent the relative importance of a factor in the analysis.
Before combining all the factor maps, each factor was weighted according to their significance
level on the site selection process. The weights were normalized to add to 1, defining the
formula as follows:

𝑤𝑤 = (𝑤𝑤1 , 𝑤𝑤2 , . . . , 𝑤𝑤𝑛𝑛 ), � 𝑤𝑤 = 1

The Analytical Hierarchy Process (AHP) was used to determine the criteria weight in this
study. In this method, every criterion under consideration is ranked in order of preference
according to its suitability for the specific objective, being, in this case, the site selection
suitability for urban pond harvesting systems. Once the ranking was established, the pairwise

21
comparison matrix (PCM) method was used to obtain the weights. The matrix has the following
structure:

1
where 𝑎𝑎𝑖𝑖𝑖𝑖 = 1,𝑎𝑎𝑗𝑗𝑗𝑗 = 𝑎𝑎 , and 𝑎𝑎𝑗𝑗𝑗𝑗 ≠ 0 50. Later, the weights are calculated by dividing the cell
𝑗𝑗𝑗𝑗

values of the pairwise comparison matrix (PCM) by the sum of the column and averaged to
obtain the weights of the factors.

6. Map overlaying
For the Map overlaying step, the Weighted Overlay ArcGis tool is applied based on the concept
of weighted average. The total suitability score will be determined by multiplying the suitability
scoring of each factor by the importance weight assigned to every factor, to finalize by adding
these products over all the criteria51 applying the following formula:

𝑆𝑆 = � 𝑊𝑊𝑖𝑖 𝑆𝑆𝑖𝑖

Being 𝑆𝑆𝑖𝑖 the suitability index and 𝑊𝑊𝑖𝑖 the determined weight for each factor.

7. Suitability Map and Robustness Evaluation


The output of the Weighted Overlay tool will result in the suitability map for the emplacement
of urban pond harvesting systems, according to its level of suitability attending to the selected
area. Ultimately, a robustness evaluation will be conducted to determine the subjectivity of the
criteria weights.

50
Ozturk, D. & Kiliç, F., 2011, ‘Implementation of GIS-Based multicriteria Decision Analysis with VB in
ArcGIS’, International Journal of Inf ormation Technology and Decision Making · November 2011, pp.
1025-1026.
51
Yoon, K.P., & Hwang, C.L., 1995, ‘Multiple Attribute Decision Making: An Introduction, SAGE
Publications, Thousand Oaks, pp. 31–37.

22
Results...
Determination of criteria for site suitability selection of PHS
n the search of relevant studies, numerous papers were discovered where different
biophysical criteria were used to determine site suitability selection for RWHS systems. It
is essential to mention that no studies on the spatial location analysis of RWHS systems using
MCDA in Madrid were found. Most identified studies took place in the Middle East, Africa,
Latin America, or Southern Asia.

The selected studies fulfilled an open system water harvesting function, including dams, water
retention ponds, ponds, water storage tanks, and water retention ponds. Out of the detected
studies, a total of 14 studies were reviewed, and their criteria for site selection suitability are
summarized in the table below:

Table 1: Selected criteria for rainwater harvesting according to different authors

Selected criteria for rainwater harvesting

Land Drainage and Watershed Total nr.


Rainfall Runoff Slope Soil type Aquifers
Authors RWHS Type use/cover rivers size of criteria

Faqe, G.R. et al.


(2019) 52 Dams 1 1 1 1 1 1 6

Ferianto, D &
Abdullah, R (2019) 53 Water retention ponds 1 1 1 1 1 5

Rooftop, ponds, pans


Mat, B et al (2006) 54 and dams 1 1 1 1 4

Senay, G.B. &


Verdin, J.P. (2004)55 Ponds 1 1 1 1 4

Mbilinyi, B.P. et al
(2007) 56 Water storage tanks 1 1 1 1 1 5

Kadam, A.K. et al. Dams, ponds,


(2012) 57 percolation tanks 1 1 1 1 1 5

52 Faqe, G.R., Rasul, A, Ali, A, Fattah, Z & Ahmad, A, 2019, ‘Suitable Site Selection for Rainwater
Harvesting and Storage Case Study Using Dohuk Governorate’, MDPI, pp. 3
53 Ferianto, D & Abdullah, R, ‘Application of geographic information systems (GIS) in the multicriteria

site selection of retention ponds for urban rainwater management’, ReaserchGate, pp. 4
54 Mati, B., De Bock, T., Malesu, M., Khaka, E., Oduor, A., Nyabenge, M. & Oduor, V., 2006,

‘Mapping the potential of rainwater harvesting technologies in Africa. A GIS overview on development
domains for the continent and ten selected countries’, Technical Manual, 6, p. 126
55 Senay, G.B. & Verdin, J.P., 2004, ‘Developing index maps of water-harvest potential in Africa’,

Applied Engineering in Agriculture, pp. 789 - 799


56 Mbilinyi, B.P., Tumbo, S.D., Mahoo, H.F. & Mkiramwinyi, F.O., 2007, ‘GIS-based decision support

system for identifying potential sites for rainwater harvesting’, EL SEVIER, pp. 1074-1081
57 Kadam, A.K., Kale, S.S., Pawar, N.J., Sankhua, R.N., & Pawar, N.J., 2012, ‘Identifying potential

rainwater harvesting sites of a semi-arid, basaltic region of Western India, using SCS-CN method’,
Water Resource Management 26 (9), pp. 2537-2554.

23
Al-Adamat, R. et al Water harvesting
(2010) 58 ponds 1 1 1 1 1 5

Ziadat, F. et al. Dams, ponds, and


(2012) 59 runoff strips 1 1 1 1 1 5

I. Sekar, I. &
Randhir, T.O.
(2007) 60 Water storage 1 1 1 1 4

Weerasinghe, H., Tanks, water


Schneider, U.A. & reservoirs, dams, pits,
Löw, A. (2011) 61 and terrace 1 1 1 1 1 5

Khan, M.D. & Dams, ponds, Ngarim


Khattak, M. (2012) 62 and gully plugs 1 1 1 1 1 5

Moges, G. (2009) 63 Ponds and in situ 1 1 1 1 4

Isioye, O.A. et al, In situ, ponds and


(2012) 64 dams 1 1 1 1 4

Anane, M. et al.
(2012) 65 Water harvesting 1 1 1 1 1 5

Number of times the criteria is used 12 6 11 12 12 8 3 2 66

% as a percentage of the total reviewed


studies 18,18% 9,09% 16,67% 18,18% 18,18% 12,12% 4,55% 3,03% 100,00%

Based on the reviewed studies and papers considering site location for RWHS systems, the
most commonly applied criteria (as a percentage of the reviewed studies) were Rainfall
(18,18%), slope (16,67%), soil type (18,18%), land use/cover (18,18%), drainage and rivers
(12,12%).

Based on Hussein, F. & Mohamed, A. R. (2015) study for the location of water retention ponds,
the criteria were subdivided into two groups: factors and constraints. A factor criterion should

58Al-Adamat, R., Diabat, A., Shatnawi,G., 2010, ‘Combining GIS with multicriteria decision
making for siting water harvesting ponds in Northern Jordan.’, Journal of Arid Environment 74, pp.
1471 - 1477.
59Ziadat, F., Bruggeman, A., Oweis, T., Haddad, N., Mazahreh, S., Sartawi, W. & Syuof, M., 2012, ‘A

participatory GIS approach for assessing land suitability for rainwater harvesting in an arid rangeland
environment’, Arid Land Research and Management, 26 (4), pp. 297-311
60 I. Sekar, I. & Randhir, T.O., 2007, ‘Spatial assessment of conjunctive water harvesting potential in

watershed systems’, Journal of Hydrology, 334 (1), pp. 39-52.


61 Weerasinghe, H., Schneider, U.A. & Löw, A., 2011, Water harvest-and storage-location

assessment model using GIS and remote sensing, Hydrology and Earth System Sciences
Discussions, 8 (2), pp. 3353-3381.
62 Khan, M.D. & Khattak, M., 2012. ‘Sitting of rainwater harvesting locations in District Haripur using

geographic information techniques’, Journal of Himalayan Earth Sciences, 45, p. 2.


63 Moges, G., 2009, ‘Identification of potential rainwater harvesting areas in the central Rift Valley of

Ethiopia using a GIS based approach (Doctoral dissertation)’, Wageningen University, The
Netherlands, p. 83.
64 Isioye, O.A, Shembe, M.W., MNomoh, U.O. & Bako, C.N., 2012, ‘A Multi Criteria Decision Support

System (MDSS) for identifying rainwater harvesting site(s) in Zaria, Kaduna State, Nigeria’,
International Journal of Advanced Scientific Engineering and Technological Research, 1 (1) , pp. 53-
71.
65 Annane, M., Bouziri, L., Limam, A. & Jellali, S., 2012, ‘Ranking suitable sites for irrigation with

reclaimed water in the Nabeul-Hammamet region (Tunisia) using GIS and AHP-multicriteria decision
analysis’, Resources, Conservation and Recycling, Volume 65, pp. 36-46.

24
be understood as a criterion that enhances or detracts suitability for a specified activity under
consideration. A constraint serves as a way to limit alternatives under consideration,
disregarding alternatives that are considered unsuitable due to the presence of a specific
characteristic.

Factors
The goal of the PHS is for it to retain water for non-potable uses. Therefore, the site suitability
for the urban water retention pond is determined by the amount of rainfall in a specific location,
meaning the higher the rainfall, the higher the suitability. The slope of the land must also be
suitable for the pond’s location for it to accumulate water. Therefore the slope must be limited,
meaning the higher the slope, the lower the suitability. The means to supply the retention pond
together with rainfall collection will be through a river. Therefore distance to the river should
be considered. For the distance factor to rivers, the closer the distance, the higher the
suitability. In the case of the soil characteristics, the more impervious the soil, the higher the
suitability.

Constraints
The constraints were set through Boolean logical data. In mathematics, a Boolean function is
a function whose domain comprises binary values of 0 or 1, or true or false, respectively. The
Boolean constraints were set depending on the possibility to construct or not in the given area.
Some areas are not suited for constructing retention ponds, including water, industrial,
commercial, public, military, and other constructed areas.

Given the characteristics of the establishment and the motivation for the implementation of
PHS (mainly irrigation purposes), this study will propose its most optimal location in the green
areas of the city. Regarding green spaces, Madrid counts 18,3 m2 of green areas per
inhabitant and has more than 50% of its streets covered with trees, with 6.000 ha of green
areas in the city. To better understand the distribution of green or unbuilt areas in Madrid,
Figure 9 is attached. The areas marked in the green image correspond to green spaces, the
largest green area in the northern part of the city called el Monte del Pardo.

25
Figure 9: Outline identifying the main green spaces in the city of Madrid

26
Development of a location suitability map for PHS in Madrid
GIS has recently been established as a tool for urban planners to create decision support
systems due to its practicality in analyzing spatial information with the superposition of layers,
creating and offering viable alternatives for the solution of urban dilemmas. The literature
review carried out allowed the identification of case studies in which GIS is applied for the
prioritization of rainwater collection points, which at the same time permitted the identification
of the relevant criteria for the study and the description of similar or standard methodologies
in the study process.

Based on this, the quantitative methodology for the elaboration of the suitability map for Urban
Rainwater Harvesting was applied as follows:

1. Raw spatial data collection


Spatial information regarding the identified criteria, including factors (slope, rainfall, distance
to rivers, soil) and constraints (land use), is collected using GIS. The raw GIS datasets used
for the study and their data type and description can be found in Annex I.

2. Creation of basic maps


The spatial information was loaded and processed using ArcMap 10.8.1. For the study, a
workflow was created using Model Builder. This application allows stringing together
sequences of geoprocessing tools, joining the output of one tool into another tool as an input66.

The figure below describes the workflow and the operations used to obtain the suitability Map.
The diagram has been divided into two phases to differentiate the operations for the
constraints and criterion factors.

Figure 10: Workflow of the Model Builder in ArcGIS

66ArcGIS Desktop https://desktop.arcgis.com/en/arcmap/latest/analyze/modelbuilder/what-is-


modelbuilder.htm, Accessed 13/10/2021

27
Before building the model, some general properties of the study were set under the
Environments tag of the Model properties, aligning the output coordinates, processing extent,
and cell size for all maps.

Table 2. Model Builder environments properties

Model properties

Geodata frame of reference ETRS 89

Datum European Terrestrial Reference System 1989


UTM zone 30 N

Processing extent Shapef ile contour of the city of Madrid


Cell size 5x5

The sections below explain the procedures to generate the criterion maps, both constraint,
and factor.

2.1. Constraints

2.1.1. Land-Use
The land-use layer was used to determine those points where the urban rainwater retention
pond emplacements were not feasible. For the evaluation of land use in Madrid, the Urban
Atlas layer was downloaded from the Spanish National Geographic Information Centre (CNIG)
in shapefile format. Urban Atlas provides land occupancy data for the most significant Spanish
urban areas at a 1:15.000 scale and is part of the Territory services that the European
Copernicus program provides.

The first step of the data transformation process was to transform the shapefile vector layer
into a raster to manipulate the data. This step was carried out using the Feature to raster tool.
Once the Map was in raster format, an analysis of the layers’ Attribute table took place. The
constraints are related to the land use of the area. The areas identified as non-feasible were
the ones attributed with a 0 value according to the Boolean logic: airports, construction sites,
urban fabric, roads, and associated land, industrial, commercial, public, military, and private
land units, Isolated structures, mineral extraction and dump sites, sports and leisure facilities
and water.

On the other hand, areas marked with a 1 allow the insertion of PHS technologies, being the
following: arable land, forests, green urban areas, herbaceous vegetation associations, land
without current use, open spaces with little or no vegetation, pastures, and permanent crops.

28
2.2. Factors

2.2.1. Rain
To prepare the rain map, a shapefile layer containing information about the location of
pluviometry stations in Madrid was downloaded from the State Meteorological Agency
(AEMET) with a 1:25:000 scale. This layer provided information regarding the location and
distribution of pluviometry stations in Madrid, allowing climatological variables to be monitored
and collected. To include this information, pluviometry data was downloaded from the AEMET
in Excel format and was introduced manually in ArcGIS through the layer’s attribute table,
including data of the mean annual Rainfall recorded in mm/year. The downloaded shapefile
contained information about the location. However, the attribute table did not contain data
regarding the amount of rainfall recorded for each station. The source data detailing the rainfall
information of the different stations can be found in Annex II.

Once the annual mean precipitation values were defined for each rainfall station, a Point to
raster conversion took place to have the data in raster format. To create the rain map from the
precipitation stations, the next step was to interpolate the data to obtain a continuous rainfall
field so that every point of the Map has a precipitation value attributed. Spatial interpolation is
a mathematical procedure used to estimate the value of a particular attribute, in this case,
Rain, from values obtained in measurement points to the different cells of the Map. Two of the
most used interpolation processing models are the Inverse Distance Weighting (IDW) and the
Kriging methods. For this study, both methods are performed using ArcGIS tools. They will be
compared to evaluate the model with more accurate estimates depending on the pluviometry
data. The comparison evaluation takes place through visual evaluation, qualitatively
establishing the behavior patterns of both maps 67.

The IDW (Inverse Distance Weighting) tool applies an interpolation method that estimates cell
values by calculating the average of the values in the vicinity of each processing cell. The
closer a point is to the center of the cell being estimated, the more influence or weight it will
have on the average calculation. 68 To perform the IDW, the Inverse Distance Weighting tool
from the Spatial Analyst ArcGIS toolset was applied.

On the other hand, Kriging is an advanced geostatistical procedure that generates an


estimated interpolation surface from a set of scattered points, including z (height) values69.
Regarding the weights, in the case of the Kriging method, they come from a semi-variogram
developed by analyzing the spatial structure of the data. To create the continuous surface,
Kriging uses predictions based on the semi-variogram and the spatial arrangement of nearby
measures values 70. The Kriging tool was selected from the Spatial Analyst ArcGIS toolset to
carry out the Kriging estimation method.

67
Vargas, A., Santos, A., Cárdenas, E & Obregón, N., 2011, ‘Analysis of distribution and spatial
interpolation of rainfall in Bogota, Colombia’, Dyna, año 78, Nro. 167, pp. 152-159
68
ArcGIS Desktop, https://pro.arcgis.com/es/pro-app/latest/tool-reference/spatial-analyst/comparing-
interpolation-methods.htm, Accessed 12/09/2021
69
ArcGIS Desktop, https://pro.arcgis.com/es/pro-app/latest/tool-reference/spatial-analyst/comparing-
interpolation-methods.htm, Accessed 12/09/2021
70
Sheehan, D, 2003, ‘Working With Interpolation Tools: IDW, Spline, and Kriging’, Updated 10/29/03 -
CA, 10/11/2006 -DS, Accessed on 21/11/2021. Web:
https://www.mit.edu/course/1/1.966/OldFiles/www/Lab4

29
Kriging IDW

Map 1: Rain map interpolation using Kriging Map 2: Rain map interpolation using IDW

As observed in Map 1, the interpolation of the data using the Kriging model provides less
abrupt data changes, and therefore a more precise continuous map. On the other hand, the
IDW model produced concentration point errors in the center of the Map, aggregating rain
values surrounding the stations and causing unrealistic model outputs in the vicinity of the
meteorological points 71. For this reason, the study was conducted with the rainfall map
obtained applying the Kriging method in ArcGIS shown in Map 1. Details on the method and
its variables can be found in Annex III.

2.2.2. Soil type


For the soil type map, a geological and lithological base map of Madrid was downloaded from
the Geological and Mining Institute of Spain (IGME) with a scale of 1:1.000.000 providing
geological and lithological units for the Iberian Peninsula and both the Balearic and the Canary
Islands, detailing their specific soil description in the attribute table: dominant rock types,
chronostratigraphic characteristics, and lithologic units. The complete legend of the
downloaded shape can be found in Annex IV.

Regarding the extent of the city of Madrid, four different types of soil were identified: Rocky
granitoids, gravel and silts, conglomerates, and clays. After downloading the Map, a Feature
to raster transformation tool was used to convert the shape into a vector format.

2.2.3. Slope
The Digital Slope Model was downloaded from the Spanish National Geographic Information
Centre (CNIG) in raster format with a pixel size of 5 and a 1:50.000 scale. In the attribute table
accompanying the Map, the slope was detailed in percentage (%)

71 Drešković, N., & Samir, D., 2012, ‘Applying the inverse distance weighting and kriging methods of
the spatial interpolation on the mapping the annual precipitation in Bosnia and Herzegovina’,
International Congress on Environmental Modelling and Software, pp. 229

30
2.2.4. Rivers
A shapefile layer was downloaded from the Spanish Ministry of Agriculture and Fisheries,
Food and Environment (MAPAMA) for the river map. The downloaded cartography included
the delimitation of the hydrographic network of the Iberian Peninsula with a scale of 1:25.000.
Once downloaded and performed spatial analysis evaluations, the layer was transformed into
raster format using the Feature to raster tool.

For this study, the importance of the rivers is related to the distance of these to the PHS since
the intention is that they fulfill both the function of water supply in drier periods and drainage
route in case of overflow. In this context, the Euclidean Distance tool was applied, which
calculates the distance to the closest source for each cell.

3. Standardization and determination of site suitability scores


For the purpose of the study, each variable was given a suitability score based on its
landscape characteristics. Four suitability levels were considered, 4 very high suitability, score
3 for high suitability, 2 for moderate suitability, and 1 for low suitability based on the Al-Adamat
R. et al. method 72. All criterion layers were run through the Reclassify tool to reassign their
values to the suitability scores.

4. Creation of factor and constraint maps

Factors
For rainfall factors, the higher the Rainfall, the higher the suitability. For slope factors, the
higher the slope, the lower the suitability; for the distance to rivers, the closer the distance, the
higher the suitability. Finally, for soil characteristics, rocky granitoids were given the lowest
suitability since they could complicate the construction of the water retention pond, Gravel
sand and slits were given low suitability since they present higher percolation rate,
conglomerates were given moderate suitability, and finally clays for being the most impervious
of the mentioned. 73

Based on these suitability characteristics for the location of the urban pond, a Reclassify was
applied to all the essential maps mentioned in section 2. The suitability scores of each factor
are presented in the table below:

72 Al-Adamat R., Diabat A. & Shatnawi G, 2010, ‘Combining GIS with multicriteria decision making for
siting water harvesting ponds in Northern Jordan’, Journal of Arid Environment, 74, 1471-1477.
73 Hussein, F. & Mohamed, A. R., 2015, ‘Selection of rainwater harvesting sites by using remote

sensing and GIS techniques: a case study of kirkuk, Iraq’, Jurnal Teknologi (Sciences & Engineering)
76:15, pp. 75-81

31
Nr. Factors Suitability value Suitability score Attribute table characteristics

1 Rain Low 1 3,54 – 7,2

Moderate 2 7,2 – 9,8

High 3 9,8 – 12,7

Very high 4 12,7 – 17,7

2 Soil Low 1 Rocky granitoids

Moderate 2 Gravel sand and slits

High 3 Conglomerates

Very high 4 Clays

3 Slope Low 1 > 4.5

Moderate 2 3 - 4.5

High 3 1.5 - 3

Very high 4 < 1.5

4 Distance to rivers Low 1 > 500

Moderate 2 300 - 500

High 3 150 - 300

Very high 4 0 - 150

The figures below represent the factor maps because of the reclassification of the factors
according to the suitability scores:

32
Map 3: Factor Rain Map Map 4: Factor Slope Map

Map 5: Factor Soil Map Map 6: Factor distance to river Map

33
Constraints
The Boolean constraints were set depending on the possibility to construct or not in the given
area. Land uses such as Airports, construction sites, the urban fabric of any density, fast transit
roads, Industrial, commercial, public, military, and private units, mineral extraction and dump
sites, roads, railways sports facilities, and water bodies were set as 0 in the constraint map
since the water retention pond cannot be constructed in those land-use characteristics. In the
case of “Mineral extraction and dumpsites” was set as 0, for it is expected that in these areas,
the amount of contamination may be high due both to the extraction of minerals and its use
as a landfill, so it would not be appropriate for the location of water retention ponds. The Spatial
Analyst tool Reclassify was used to attribute the Boolean values to obtain the final constraint
map. By reclassifying the land use map, the raster was given the value of one or zero,
transforming the original table attributes into the Boolean values as can be seen in the table
below:

Table 3: Land use attribute table reclassification values

Land use attribute table Reclassify value


Airports 0

Arable land 1

Construction sites 0

Continuous urban fabric 0

Discontinuous dense urban fabric 0

Discontinuous low-density urban fabric 0

Discontinuous medium density urban fabric 0

Discontinuous very-low-density urban fabric 0

Fast transit roads and associated land 0

Forests 1

Green urban areas 1

Herbaceous vegetation associations 1

Industrial, commercial, public, military, and private units 0

Isolated structures 0

Land without current use 1

Mineral extraction and dumpsites 0

Open spaces with little or no vegetation 1

Other roads and associated land 0

Pastures 1

Permanent crops 1
Railways and associated land 0

Sports and leisure facilities 0

Water 0

34
As shown in the table, only natural or non-constructed areas were selected as suitable for
PHS implementation. The following Map represents the constraint map with only two scores,
namely 1 and 0. Pixels with a score of 0 indicate a non-feasible location for the urban rainwater
retention pond, while pixel scoring 1 is feasible.

Map 7: Land use constraints Map

35
5. Criteria weighting
The studies reviewed to obtain the criteria to be considered in the PHS suitability mapping
process provided information regarding the selection of criteria and, in turn, provided
information on those criteria most used. Based on the reviewed studies and papers
considering site location for RWHS, the most commonly applied criteria (as a percentage of
the reviewed studies) were rainfall (18,18%), slope (16,67%), soil type (18,18%), land
use/cover (18,18%), drainage and water bodies (12,12%). Therefore, it is important to mention
that to objectify the process of criteria hierarchy when carrying out the AHP, these percentages
were considered when prioritizing between criteria.

According to Al-Adamat (2008), Rainfall is the major component in water harvesting systems.
Therefore, the more Rainfall in any area, the higher the possibilities of harvesting part of it.
Since Rainfall was the most used criterion based on the literature review, scoring 18,18%, it
should occupy the first position in the AHP. Conforming with Al-Adamat and Critchley et al.’s
studies, soil texture is considered an extremely decisive factor regarding suitability site
selection for PHS, primarily when intended to be retained for non-potable uses such as
irrigation since it will significantly influence the percolation factor. Given these studies and the
literature review score of 18,18%, soil type will occupy the second position in the AHP.

On the other hand, according to Critchley et al. (1991) 74 , PHS installation is not recommended
in areas where the slope is higher than 5% due to the economic cost of construction works
and uneven distribution of the runoff. Based on the literature review, slope counted with a
16,67% score in applications. It, therefore, will be placed in the third position of the AHP.
Finally, drainage and water bodies are defined in Al-Adamat’s study as the primary carrier of
surface water and the suppliers for most of the collected surface water runoff in wintertime.
Therefore, distance to rivers is essential both for water supply and for channeling water in
pond overflow. Distance to the river will occupy the fourth and land position in the AHP
hierarchy ranking, obtaining a 12,12% score in the literature review analysis. The final rank
assignment can be found in Table 3 below.

Table 4: Factor ranking positions

Factors Rainfall Soil type Slope Distance to river


Rank 1 2 3 4

Once the factors are ranked, a pairwise comparison matrix is constructed, based on the rating
of the relative preferences established above for the present study, evaluating two criteria at
a time. Each comparison is a two-part question 75, evaluating which factor es more important
and to what extent. The values range from ¼ representing the least important to 1 for equal
importance and 4 for the most important, covering all the possible values in the matrix.

74 Critchley, W., Siegert, K. & Chapman, C., 1991, ‘Water Harvesting, A Manual Guide for the Design
and Construction of Water Harvesting Schemes for Plant Production’, FAO, Rome, Accessed
08/01/2020
75
Bakhtiar, F., Kienberger, S. & Valizadeh Kamran, K., 2015, ‘Sensitivity and Uncertainty Analysis
Approach for GIS-MCDA Based Economic Vulnerability Assessment’, GI_Forum ‒ Journal for
Geographic Information Science, pp. 81-87.

36
Table 5: Pairwise comparison for the selected factors

Criteria Rainfall Soil type Slope Distance to river

Rainf all 1/1 2/1 3/1 4/1

Soil type 1/2 2/2 3/2 4/2

Slope 1/3 2/3 3/3 4/3

Distance to river 1/4 2/4 3/4 4/4

Finally, the pairwise comparison matrix is normalized, and the weights are obtained.

Table 6: Weights assigned to the selected factors

Criteria Rainfall Soil type Slope Distance to river Weight

Rainf all 1,00 2,00 3,00 4,00 45%

Soil type 0,50 1,00 1,50 2,00 25%

Slope 0,33 0,67 1,00 1,33 17%

Distance to river 0,25 0,50 0,75 1,00 13%

Following the calculations performed in the pairwise comparison, it is crucial to undertake a


consistency ratio (C.R.) judgment to check the consistency of the conducted comparisons76.
When all judgments are perfectly consistent, the value is zero. However, according to Saaty 77,
if the C.R. <0.10, the pairwise comparison matrix has an acceptable consistency. The weight
values are valid and can be utilized. Otherwise, if the C.R.≥0.10, the pairwise comparisons
lack consistency, and the matrix needs to be adjusted, and the element values should be
modified 78. The C.R. of the operations carried out in this study is CR=0.04 and therefore
considered acceptable.

6. Map overlaying
The following table summarizes the primary operations carried out throughout the
methodology detailing the suitability scores, standardized suitability stored in the factor maps,
and the weights obtained through the AHP method:

76
Bakhtiar, F., Kienberger, S. & Valizadeh Kamran, K., 2015, ‘Sensitivity and Uncertainty Analysis
Approach f or GIS-MCDA Based Economic Vulnerability Assessment’, GI_Forum ‒ Journal for
Geographic Information Science, pp. 81-87.
77
Saaty, T. L., 1977, ‘A scaling method f or priorities in hierarchical structure’, Journal of Mathematical
Psychology, 15 (3), pp. 234-281.
78
Bakhtiar, F., Kienberger, S. & Valizadeh Kamran, K., 2015, ‘Sensitivity and Uncertainty Analysis
Approach f or GIS-MCDA Based Economic Vulnerability Assessment’, GI_Forum ‒ Journal for
Geographic Information Science, pp. 81-87.

37
Table 7: Map overlaying weight and suitability scores description for each factor

Factor Weight % Suitability Suitability value


score
Rainfall 45 1 low

2 moderate

3 high

4 very high

Soil type 25 1 low

2 moderate

3 high

4 very high

Slope 17 1 low

2 moderate

3 high

4 very high

Distance to river 13 1 low

2 moderate

3 high

4 very high

The final suitability map was obtained by multiplying the suitability factor map with the
constraint map to limit the suitability areas. The applied MCDA method for this study is the
weighted overlay, obtaining the final Map by multiplying the suitability score by the assigned
weight.

38
7. Final Suitability Map
The following maps show the final suitability maps resulting from the multi-criteria analysis for
PHS in the city of Madrid.

Map 8: Suitability Map for PHS in the city of Madrid

39
Based on the selected criteria and the analytical methods used to obtain the Map, the table
below details the percentage of the surface of the city of Madrid according to the suitability
values for the implementation of PHS:

Table 8: Numerical Suitability values for installation of PHS in Madrid

Suitability
Number of pixels % km2
value

Restricted 10.774.061 44,63% 269,351525

Low 1.297.617 5,38% 32,440425

Moderate 9.481.170 39,27% 237,02925

High 2.532.454 10,49% 63,31135

Very high 55.630 0,23% 1,39075

Robustness Evaluation
Ultimately, a robustness evaluation was performed. The main objective of conducting a
robustness evaluation is to determine how much the decision maker’s preferences affect the
geographical suitability results. The robustness evaluation or sensitivity analysis mainly
consists of re-evaluating the factor suitability scoring (1 low suitability - 4 very high suitability)
and the weights assigned to each factor to determine which values are more critical in
determining the final suitability map. This evaluation was carried out by modifying the weights
granted to each factor in the MCDA phase, maintaining the same suitability scores and the
same number of constraints and factors.

In order to carry out the robustness evaluation, the factors were modified as follows: First,
there was a change in the approach, and for the robustness evaluation, instead of filling the
pond with rainwater, it would be filled with the excess of water from the river itself, so the
proximity to the river would occupy the highest of the positions in the hierarchy level. Secondly,
we would have the soil type since, in this sense, it is a very limiting factor due to the
impermeability it provides. Thirdly, we place the slope since the possible locations depend on
it to a large extent. Finally, the rain factor, since the ponds can help in torrential rains or storms
to avoid an excess runoff.

The AHP analysis with the new determined approach will result as follows:

Table 9: Factor ranking positions for the robustness evaluation

Criteria Distance to river Soil type Slope Rainfall


Rank 1 2 3 4

40
Table 10: Pairwise comparison for the robustness evaluation

Criteria Distance to river Soil type Slope Rainfall

Distance to river 1/1 2/1 3/1 4/1

Soil type 1/2 2/2 3/2 4/2

Slope 1/3 2/3 3/3 4/3

Rainf all 1/4 2/4 3/4 4/4

Table 11: Weights assigned for the robustness evaluation

Criteria Distance to river Soil type Slope Rainfall Weight

Distance to river 1,00 2,00 3,00 4,00 0,45

Soil type 0,50 1,00 1,50 2,00 0,25

Slope 0,33 0,67 1,00 1,33 0,17

Rainf all 0,25 0,50 0,75 1,00 0,13

The suitability map 2 will result as follows applying the above weights:

41
Map 9: Suitability Map nº2 based on robustness evaluation

42
Table 12: Numerical Suitability values for Suitability Map 2

Suitability value Number of pixels % km2

Restricted 10.774.061 44,63% 269,351525

Low 3.229.908 13,38% 80,7477

Moderate 8.990.658 37,24% 224,76645


High 1.092.397 4,53% 27,309925

Very high 53.908 0,22% 1,3477

Discussion...
n the previous section, the results for two different types of Weighted Overlay (WO) in the
multi-criteria analysis are shown. Different weights have been changed from one result to
another to study the model’s robustness and give another approach to the suitable locations
of the PHS in Madrid. The robustness evaluation was conducted by modifying the weights
granted to each factor in the MCDA phase, maintaining constraints and suitability scores
intact. Therefore, it was expected to find the main differences between the two maps to be
directly related to the variation of weight of the different factors.

In the first analysis, where rainfall and soil textures were given the highest weights, the areas
with a suitability score of very high are distributed between the city’s northwest and the
southwest areas. Based on these criteria, there is a total of 0,23% or 1,4 𝑘𝑘𝑘𝑘2 of the area with
suitability considered as very high, 10,48% of area or 63,3 𝑘𝑘𝑘𝑘2 considered as high, 39,32%
of the area or 237 𝑘𝑘𝑘𝑘2 suited as moderate and finally 5,38% of surface or 32,4 𝑘𝑘𝑘𝑘2 considered
as low. As can be seen in map 10, there are three distinct very high suitability clusters in the
northwest of the city of Madrid, all of them surrounded by high suitability level areas.

Cluster 1 is located at the entrance of the Manzanares river to Madrid and just before its
widening towards the entrance to the city. Its proximity to an active river could devalue the
properties sought to be obtained from the PHS since its main objective was water harvesting
for irrigation purposes during dry seasons. Cluster 2 It is in the north of Monte del Pardo at its
entrance into the territory of the city of Madrid. Given its location in the center of the northern
area, it could be interesting to evaluate the incorporation of the PHS in that location. The third
Cluster is located bordering the Arrollo de Trofas. This tributary runs along the right side of the
Manzanares river, in the northwestern area of Monte del Pardo. In this case, we are in a similar
situation as with Cluster 1, where the installation of the PHS may be meaningless due to its
pronounced proximity to the stream. Although, in this case, being Arrollo de Trofas a much
smaller water body, it could make sense as a collection point and fulfill its harvesting function
during the rainy periods in the interest of water availability in the drier periods.

Furthermore, and as could be appreciated in Figure 9 regarding the green areas of Madrid,
the Norwest area of Madrid city corresponds with the Monte del Pardo and the regional park
of Cuenca Alta del Manzanares . A state body regulates the management of these areas since
the park is considered National Patrimony; this translates into a highly protectionist and

43
restrictive policy concerning the green area and its surroundings. In this sense, some
complications may arise when introducing the PHS.

Map 10: PHS Suitability Map - Northern clusters

On the other hand, in the southwest area, there are six differentiated clusters (Map 11). The
clusters are surrounded by high suitability level areas and restricted urban or constructed
areas in all six samples. Clusters 4 and 6 are located bordering the passage of river
Manzanares through the urban area of Madrid, in a park area called Madrid Rio. Madrid Rio
is a park surrounded by green spaces, monuments, bridges, pine forests, river water from river
Manzanares, and accompanied by river ponds. Since this space already counts with river
ponds, a very similar technology to that intended to be obtained through PHS, adding this

44
technology could be vague unless it accompanied the existing river ponds regarding water
harvesting.

Additionally, this area where the Manzanares River passes by has footbridges and dams
upstream for public safety reasons, so the recharge of these river ponds is highly controlled.
Cluster 9, on the other hand, presents a valuable location since it is placed in a park called
Parque Linear de Manzanares. This park is also crossed by the Manzanares river but carries
a lower water flow. In this regard, the proximity to the river could be used from the point of
view of overflow discard, and at the same time benefit the park from the water harvesting
functions from the PHS for irrigation.

Cluster 5 is located in the Park of Tierno Galván, another urban park in the city’s center. This
park already has three ponds distributed along its area, so incorporating PHS would not imply
a tremendous additional value to the site, especially regarding irrigation 79. Finally, Cluster 7
and Cluster 8 are located in two small parks: the Julio Alguacil Gómez Forest Park and the
Dehesa Boyal park. The Julio Alguacil Gómez Forest Park responds to the need to implement
innovative urban regeneration strategies integrating social, economic, and environmental
aspects to favor the sustainable development of the city of Madrid 80. In this context, the
implementation of PHS technologies would be ideal for the maintenance of the park as well
as for the implementation of sustainable water management strategies to address added value
in sustainable matters to the park. On the other hand, the Dehesa Boyal park could also benefit
from implementing PHS. However, since the park already has a water body in its area and
considering its reduced size, the benefits become less noticeable than in Cluster 7.

79 Ayuntamiento, D.M., https://www.madrid.es/portales/munimadrid/es/Inicio/El-Ayuntamiento/Parques-y-


jardines/Patrimonio-Verde/Parques-en-Madrid/Parque-Enrique-Tierno-
Galvan/?vgnextoid=e09f680e2383e210VgnVCM1000000b205a0aRCRD&vgnextchannel=38bb1914e7d4e210Vg
nVCM1000000b205a0aRCRD, Accessed 15/11/2021
80 Ayuntamiento, D.M., https://www.madrid.es/portales/munimadrid/es/Inicio/El-Ayuntamiento/Parques-y-
jardines/Patrimonio-Verde/Parques-en-Madrid/Parque-Forestal-Julio-Alguacil-
Gomez/?vgnextoid=1bab840b7ce43610VgnVCM1000001d4a900aRCRD&vgnextchannel=38bb1914e7d4e210V
gnVCM1000000b205a0aRCRD, Accessed 12/11/2021

45
Map 11: PHS Suitability Map - Southern clusters

Regarding Suitability Map 2, corresponding to the suitable areas for PHS, ten different clear
clusters appear after modifying the different weights. For the second Map, the most relevant
weights were assigned to distance to water bodies and soil type resulting in a total of 0,23%
or 1,3 𝑘𝑘𝑘𝑘2 of the area with suitability considered as very high, 4,53% of area or 27,3 𝑘𝑘𝑘𝑘2
considered as high, 37,2% of the area or 224,7 𝑘𝑘𝑘𝑘2 suited as moderate and finally 13,38% of
surface or 80,7 𝑘𝑘𝑘𝑘2 considered as low. In the same way as Suitability Map 1, the areas with
a suitability score of very high are distributed between the northwest and the southwest areas
of the city.

Clusters 1 and 3 of the northwest area of Suitability Map 2 coincide with the ones in Suitability
Map 1. On the contrary, Cluster 2 of Suitability Map 2 is less suitable than Suitability Map 1. It

46
has the same drawbacks as Clusters 1 and 3, which lack practical utility because it is close to
the river.

Map 12: PHS Suitability Map 2 - Northern clusters

For the Southwest Clusters, numbers 5, 6, 7, 8, 9, and 10 are the same as for Suitability Map
1 and therefore have the same characteristics concerning their fitness for emplacement of
PHS. Concerning Cluster 4, its location is in the Campo del Moro Gardens, historic gardens
created by King Philip II, and historical-artistic interest. These gardens are protected as

47
national heritage and possessed by the Spanish crown. Therefore, the location of PHS in
Cluster 4 would be completely unfeasible 81.

Map 13: PHS Suitability Map 2 - Southern clusters

Comparing both Suitability maps, the clusters are almost identical except for Clusters 2 and 4
in Suitability map 2. This shows that the selection of the criteria applied in the study limits the
suitability locations to Madrid’s northwest and southwest areas since we cannot find any

81 Nacional, P., https://www.patrimonionacional.es/visita/jardines-del-campo-del-moro,


Accessed: 15/11/2021

48
Clusters on the Eastern side of the city. The model was processed correctly since the
differences were minor between both maps. For Suitability Map 2, it was possible to appreciate
the weight change regarding distance to water bodies, obtaining two different clusters adjacent
to the river compared to Suitability Map 1. Nevertheless, it is necessary to mention that even
though the Suitability map 2 presented slightly lower very high suitability areas with respect to
the Suitability Map 1 (1,3 𝑘𝑘𝑘𝑘2 vs. 1,4 𝑘𝑘𝑘𝑘2), the clusters were distributed differently, having
the second Map 10 very high suitability clusters compared to the 9 Clusters in the first Map.

Looking at the results, we can analyze the robustness of the model. A model is robust
(concerning some conditions) if changing these conditions does not change the model’s result.
In this case, the conditions changed have been the weights assigned to every factor.
Nevertheless, in the results, the location of the best suitable areas for PHS implementation
adequately correspond to the different weights given to the different factors. Considering the
two suitability maps, the second Map is more restrictive than the first one. There are less
suitable areas in the study area. Therefore the weights established for the second analysis
make a higher restriction when considering constructible areas for PHS.

Given the results, it can be extrapolated that Cluster 7 from Suitability Map 1, coinciding with
Cluster 8 from Suitability Map 2 and Cluster 9 from Suitability Map 1, are the best suitable
Clusters areas for the implementation of PHS can be found. These clusters present
characteristics that favor PHS employment: high rains, clay soils, low slopes, and an adequate
distance to rivers. The proximity of these clusters to green areas facilitates the use of
harvested water for one of the main non-potable uses that motivated the present study, the
irrigation of natural spaces.

These suitability map results only represent the suitable potential areas for PHS
implementation. However, to ensure the technical viability of the execution of the project, it is
imperative to carry out visits to the area and make measurements of all the factors considered
theoretically in this study.

Uncertainties and risks


The main qualitative uncertainties of the study refer to the search of relevant studies and the
selected keywords for the selection of relevant studies. According to Levac et al. (2020) 82, the
identification and selection of relevant studies should be performed by two or more
researchers to review and filter the most relevant articles to analyze independently. In the case
of this study, this research was done solely by the author. Therefore, there is a potential
ambiguity and subjectivity risk in selecting relevant studies.

The quantitative uncertainties in this type of modeling lie in the accuracy of input data and how
well the assumed properties match the needs of the targeted result. There are different types
of errors and uncertainties present in this work. Most of the information provided in the study
data is sourced from the Spanish National Geographic Institute (CNIG), making the
information reliable but rarely perfect. The fact that the information provided is not impeccable
supposes two types of errors: database errors and decision rule uncertainties.

Levac, D., Colquhoun, H. & O’Brien, K.K., 2010, ‘Introduction: Rationality’, Implementation Science,
82

pp. 1-16

49
In this work, the database uncertainty resides in selecting the criteria (factors and constraints)
used in the decision rule, which is the procedure by which criteria are selected and combined
to conclude a detailed evaluation. In this type of uncertainty, we must also consider the
measurement error, which is attributed to the measurement and classification of the
information that has been applied for the elaboration of this work. As the data was directly
downloaded from the mentioned sources (see Annex I), the measurement part was not part
of this study. Therefore, there must also be sources of error that reside in the accuracy of the
content, such as the density of the observations, positional accuracy, attribute accuracy in the
tables, topological topography that can influence the accuracy of the data for slope and
lineage. Nevertheless, these errors were assumed since it has been considered that the data
and the information applied were representative and worthwhile for what was intended to be
done in the study without visits to the field.

It is also important to acknowledge that there is a decision rule uncertainty, which arises from
how the data have been combined and analyzed to obtain the suitability maps. There may be
some uncertainty in the relationships chosen to select the data, such as the relationship
between drainage and distance to the river. For this reason, the robustness analysis took place
to evaluate the importance of the chosen criterion and their relations in the study. Many of the
final results shown in the previous section can drag different errors in the different steps of the
previous models that calculated constraints and distances to certain features of interest. In the
rasterization process, there is a critical error source due to the size of the pixel (see Annex I).
When applying the Euclidean Distance command, it is also essential to acknowledge that the
tool gives information according to a straight-line distance. This may not be possible in a
specific location since steep slopes, hills, or other physical obstacles may occur. All the
methodological and discussed decisions contrived in this paper are attached to a decision risk
factor that arises from the assumptions and arrangement elaborated that might come from a
combination of different sources involved and procedures.

Additionally, the applied methodology could have been approached differently. Conducting
the site selection and weighting criteria through interviews could have helped obtain valuable
insights from agents involved in generating resilience and mitigation strategies in the urban
field. In this way, their guidance could have resulted in more robust decision-making,
identifying barriers, gaps, or potential project applications.

Conclusions...
his study focused on two main objectives, (i) determination of criteria for site suitability
selection of pond harvesting systems and (ii) creation of a location suitability map for
the emplacement of pond harvesting systems in Madrid. The research firstly examined the
criteria for the emplacement of PHS in a general manner, attending to similar studies carried
out worldwide to generate suitability maps for their emplacement in the city of Madrid using
GIS-based MCDA tools.

Attending to the first objective, the qualitative research outcomes indicated the use of rain, soil
type, rivers, slope, and land use as the most applied criteria for PHS. The review of similar

50
studies also contributed to generating the quantitative methodology exposed later to generate
suitability maps.

Regarding the creation of location suitability maps, the relative importance of the different
criteria selected for the GIS-MCDA approach was conducted based on the frequency of
application of the criteria in the selected studies together with desk-based research from the
first objective. The suitability maps locate the most suitable areas for PHS emplacement in
Madrid, being Cluster 7 from Suitability Map 1, coinciding with Cluster 8 from Suitability Map
2 and Cluster 9 from Suitability Map 1. These clusters present characteristics that favor PHS
employment: high rains, clay soils, low slopes, and an adequate distance to rivers. The
proximity of these clusters to green areas facilitates the use of harvested water for one of the
main non-potable uses that motivated the present study, the irrigation of natural spaces.

This study presented the application of a GIS-based MCDA analysis by applying these
methods to a tangible problem that already affects the city of Madrid and is expected to
influence the city more in the following years due to Climate Change. The study’s outcome
has given practical results and proves the utility of GIS-MCDA in urban planning and decision
making. Nonetheless, both the factor determination and the weight selection are subjective
arguments. In order to limit the subjectivity of the study, a robustness analysis was performed
by assigning different weights to the selected criteria, obtaining similar results between the
two maps, therefore, indicating the robustness of the model.

Recommendations could also be made for future urban planners who seek to integrate PHS
into their city. Firstly, for the actual materialization of this study and to limit the areas classified
as very high suitability, a field visit should be carried out to assess more specific biophysical
aspects of the selected clusters, such as technical viability, surrounding flora and fauna, and
a deeper geological and lithological study of the area. Secondly, to obtain greater efficiency
and benefits from PHS implementation, it is advisable to consider the demand associated with
the suitability clusters, maximizing all non-potable uses for nearby areas and other areas that
can access the reuse of the resource.

This study sought to open the gate to further research and implement other technical
sustainable water management practices using nature-based facilities in the city, such as the
design, operation infrastructure, drainage logistics distribution, and other modifications in the
current water management system city of Madrid. At the same time, this study generates the
need to develop further analysis involving social, environmental, and economic aspects such
as impact in social participation in the use of rainwater, environmental impact in the receiving
bodies, cost-benefit and effectiveness analysis in the implementation of PHS.

It can be concluded that the integration of more sustainable water management systems is
vital for cities’ resilience adaptation strategies, especially in areas with water-shortage
forecasts in the upcoming years. This paper demonstrates the applicability of GIS-based
MCDA tools for planning purposes, where multiple information can be combined to create the
visualization and mapping for the most suitable site location for a specific urban dilemma.

51
References...
AEMET, Agencia Estatal de Meteorología, 2021, ‘Proyecciones climáticas para el siglo XXI’,
http://www.aemet.es/es/serviciosclimaticos/cambio_climat, Accessed: 27/09/2021

Agudelo-Vera, C. M., Leduc, W. R. W. A., & Melsa, A. R., 2012, ‘Harvesting Urban
Resources Towards More Resilient Cities’, Resources Conservation and Recycling 64,
pp: 3–12.

Al-Adamat R., Diabat A. & Shatnawi G, 2010, ‘Combining GIS with multi-criteria decision
making for siting water harvesting ponds in Northern Jordan’, Journal of Arid Environment,
74, 1471-1477.

Al-Adamat, R., Al-Ayyash, S., Al-Amoush, H., Al-Meshan, O., Rawajifih, Z., Shdeifat, A. &
Al-Fajarat, M., 2012, ‘The combination of indigenous knowledge and geo-informatics for
water harvesting sitting in the Jordanian Badia’, Journal of Geographic Information Systems,
4, pp. 366-376.

Al-Shabeeb, A., 2016, ‘The Use of AHP within GIS in Selecting Potential Sites for Water
Harvesting Sites in the Azraq Basin—Jordan’, Journal of Geographic Information System 8,
pp. 73-88.

Amos, C., Rahman, A. & Gathenya, J.M., 2016, ‘Economic analysis and feasibility of
rainwater harvesting systems in urban and peri-urban environments: a review of the global
situation with a special focus on Australia and Kenya’, Water 8 (4), p. 149.

Andrés-Valeri, V. C., Perales-Momparler, S., Sañudo-Fontaneda, L. A., Andrés-Domenech,


I., Castro-Fresno, D., & Escuder-Bueno, I., 2016, ‘Sustainable Drainage Systems in Spain’,
En Wiley-Blackwell (Ed.), Sustainable Surface Water Management: A Handbook for SUDS,
pp. 355-370

Annane, M., Bouziri, L., Limam, A. & Jellali, S., 2012, ‘Ranking suitable sites for irrigation
with reclaimed water in the Nabeul-Hammamet region (Tunisia) using GIS and AHP-
multicriteria decision analysis’, Resources, Conservation and Recycling, Volume 65, pp. 36-
46.

ArcGIS Desktop, https://desktop.arcgis.com/en/arcmap/latest/analyze/modelbuilder/what-is-


modelbuilder.htm, Accessed 13/10/2021

ArcGIS Desktop, https://pro.arcgis.com/es/pro-app/latest/tool-reference/spatial-


analyst/comparing-interpolation-methods.htm, Accessed 12/09/2021

Área de Gobierno de Medio Ambiente y Movilidad, 2019, ‘Avance de la Estrategia de


Sostenibilidad Ambiental - Madrid 360, Gobierno de Madrid

Ayuntamiento de Madrid, 2006, ‘Ordenanza de Gestión y Uso Eficiente del Agua en la


Ciudad de Madrid’, Ayuntamiento de Madrid 22/06/2006 num. 5709, pp. 2410-2443.

Ayuntamiento de Madrid, 2007, ‘Criterios para una jardinería sostenible en la ciudad de


Madrid’, Área de Gobierno de Medio Ambiente y Servicios a la Ciudad

52
Ayuntamiento de Madrid, 2009, ‘Buenas Prácticas en Arquitectura y Urbanismo para Madrid.
Criterios bioclimáticos y de eficiencia energética’, Área de Gobierno de Urbanismo y
Vivienda del Ayuntamiento de Madrid.

Ayuntamiento de Madrid, 2018, ‘Guía Básica de Diseño de sistemas de Gestión Sostenible


de Aguas Pluviales en Zonas Verdes y otros Espacios Públicos’, Área de Gobierno de Medio
Ambiente y Movilidad, Dirección General de Gestión del Agua y Zonas Verdes, pp. 5

Ayuntamiento, D.M., 2014, ‘Plan de Energía y Cambio Climático de la ciudad de Madrid-


Horizonte 2020’, Ayuntamiento de Madrid, pp. 24-30

Ayuntamiento, D.M., 2018, ‘Bases y Directrices generales del plan de infraestructura verde
y biodiversidad de la ciudad de Madrid’, Plan de infraestructura verde y biodiversidad,
Ayuntamiento de Madrid, 2018

Ayuntamiento, D.M., 2018, Resumen Ejecutivo del Diagnóstico de Situación del Plan
Estratégico, Plan de infraestructura verde y biodiversidad, Ayuntamiento de Madrid, 2018

Ayuntamiento, D.M., https://www.madrid.es/portales/munimadrid/es/Inicio/El-


Ayuntamiento/Parques-y-jardines/Patrimonio-Verde/Parques-en-Madrid/Parque-Enrique-
Tierno-
Galvan/?vgnextoid=e09f680e2383e210VgnVCM1000000b205a0aRCRD&vgnextchannel=
38bb1914e7d4e210VgnVCM1000000b205a0aRCRD, Accessed 15/11/2021

Ayuntamiento, D.M., https://www.madrid.es/portales/munimadrid/es/Inicio/El-


Ayuntamiento/Parques-y-jardines/Patrimonio-Verde/Parques-en-Madrid/Parque-Forestal-
Julio-Alguacil-
Gomez/?vgnextoid=1bab840b7ce43610VgnVCM1000001d4a900aRCRD&vgnextchannel=
38bb1914e7d4e210VgnVCM1000000b205a0aRCRD, Accessed 12/11/2021

Bakhtiar, F., Kienberger, S. & Valizadeh Kamran, K., 2015, ‘Sensitivity and Uncertainty
Analysis Approach for GIS-MCDA Based Economic Vulnerability Assessment’, GI_Forum ‒
Journal for Geographic Information Science, pp. 81-87.

Bakhtiar, F., Kienberger, S. & Valizadeh Kamran, K., 2015, ‘Sensitivity and Uncertainty
Analysis Approach for GIS-MCDA Based Economic Vulnerability Assessment’, GI_Forum ‒
Journal for Geographic Information Science, pp. 81-87.

Bhattacherjee, A., 2012, ‘Social Science Research: principles, methods, and practices’,
Textbook’s collection (Vol. 9).

Butler, D., Ward, S., J. Burns, M., Friedler, E., DeBusk, K., N. Fisher-Jeffes, L., Ghisi, E.,
Rahman, A., Furumai, H., Han, M. & Campisano, A., 2017, ‘Urban rainwater harvesting
systems: Research, implementation and future perspectives’, pp 2

Campisano, A., Butler, D., Ward, S., J. Burns, et al., 2017, ‘Urban rainwater harvesting
systems: Research, implementation and future perspectives’, Water Research 115, pp. 195-
209.

Ciscar, J.C., Soria, A., M. Goodess, C. B. Christensen, O., Iglesias, A., Garrote, L., Moneo,
M., Quiroga, S., Feyen, L., Dankers, R., Nicholls, R. & Richards, J., 2009. “Climate change
impacts in Europe. Final report of the PESETA research project,” JRC Working
Papers JRC55391, Joint Research Centre (Seville site), pp. 31-38

53
Critchley, W., Siegert, K. & Chapman, C., 1991, ‘Water Harvesting, A Manual Guide for the
Design and Construction of Water Harvesting Schemes for Plant Production’, FAO, Rome,
Accessed 08/01/2020

Doulabian, S., Ghasemi Tousi, E., Aghlmand, R., Alizadeh, B., Ghaderi Bafti, A. & Abbasi,
A., 2021, ‘Evaluation of Integrating SWAT Model into a Multi-Criteria Decision Analysis
towards Reliable Rainwater Harvesting Systems’, Water 2021, 13, p. 1935.

Drešković, N., & Samir, D., 2012, ‘Applying the inverse distance weighting and kriging
methods of the spatial interpolation on the mapping the annual precipitation in Bosnia and
Herzegovina’, International Congress on Environmental Modelling and Software, pp. 229

EEA, 2012, ‘Urban adaptation to climate change in Europe. Challenges and opportunities for
cities together with supportive national and European policies’, EEA Report No 2/2012, pp.
148

Faqe, G.R., Rasul, A, Ali, A, Fattah, Z & Ahmad, A, 2019, ‘Suitable Site Selection for
Rainwater Harvesting and Storage Case Study Using Dohuk Governorate’, MDPI, pp. 3

Farreny, R., Morales-Pinzón, T., Guisasola, A., Tayá, C., Rieradevall, J. & Gabarrell, X.,
2011, ‘Roof selection for rainwater harvesting: Quantity and quality assessments in Spain’,
Water Research Volume 45, Issue 10, pp. 3245-3254

Ferianto, D & Abdullah, R, ‘Application of geographic information systems (GIS) in the multi-
criteria site selection of retention ponds for urban rainwater management’, Research Gate,
pp. 4

Fernandez García, F. & Martilli, A., 2021, ‘El clima urbano: aspectos generales y su
aplicación en el área de Madrid’, Revista Índice (50), pp 21.

García Soler, N., Moss, T. & Papasozomenou, O., ‘Rain and the city: Pathways to
mainstreaming rainwater harvesting in Berlin’, ELSEVIER vol. 89, pp. 96-100

Gbanie, S.P., Tengbe, P.B., Momboh, J.S., Medo, J., Tamba, V. & Kabba, S., 2013,
‘Modelling landfill location using Geographic Information Systems (GIS) and Multi-Criteria
Decision Analysis (MCDA): Case study Bo, Southern Sierra Leone’, ELSEVIER, pp 3.

Hussein, F. & Mohamed, A. R., 2015, ‘Selection of rainwater harvesting sites by using remote
sensing and GIS techniques: a case study of Kirkuk, Iraq’, Jurnal Teknologi (Sciences &
Engineering) 76:15, pp. 75-81

Instituto Nacional de Estadística (INE), https://www.ine.es/jaxiT3/Datos.htm?t=2881,


Accessed 07/10/2020

Isioye, O.A, Shembe, M.W., MNomoh, U.O. & Bako, C.N., 2012, ‘A Multi-Criteria Decision
Support System (MDSS) for identifying rainwater harvesting site(s) in Zaria, Kaduna State,
Nigeria’, International Journal of Advanced Scientific Engineering and Technological
Research, 1 (1), pp. 53-71.

Kadam, A.K., Kale, S.S., Pawar, N.J., Sankhua, R.N., & Pawar, N.J., 2012, ‘Identifying
potential rainwater harvesting sites of a semi-arid, basaltic region of Western India, using
SCS-CN method’, Water Resource Management 26 (9), pp. 2537-2554.

Khan, M.D. & Khattak, M., 2012. ‘Sitting of rainwater harvesting locations in District Haripur
using geographic information techniques’, Journal of Himalayan Earth Sciences, 45, p. 2

54
Levac, D., Colquhoun, H. & O’Brien, K.K., 2010, ‘Introduction: Rationality’, Implementation
Science, pp. 1-16

Mati, B., De Bock, T., Malesu, M., Khaka, E., Oduor, A., Nyabenge, M. & Oduor, V., 2006,
‘Mapping the potential of rainwater harvesting technologies in Africa. A GIS overview on
development domains for the continent and ten selected countries’, Technical Manual, 6, p.
126

Mbilinyi, B.P., Tumbo, S.D., Mahoo, H.F. & Mkiramwinyi, F.O., 2007, ‘GIS-based decision
support system for identifying potential sites for rainwater harvesting ‘, Elsevier Journal of
Physics and Chemistry of the Earth, 32, pp. 1074-1081.

Ministerio para la Transición Ecológica, 2020, ‘Plan Nacional Integrado de Energía y Clima
(PNIEC)’, Gobierno de España

Moges, G., 2009, ‘Identification of potential rainwater harvesting areas in the central Rift
Valley of Ethiopia using a GIS-based approach (Doctoral dissertation)’, Wageningen
University, The Netherlands, p. 83.

Montañez Gil, S.J., 2014, ‘Desarrollo de una herramienta computacional para priorizar la
ubicación de tanques de recolección de agua de escorrentía en lugares con riesgo de
inundación: Caso Bogotá’, Universidad de los Andes, facultad de ingeniería civil y ambiental,
p. 8

Nacional, P., https://www.patrimonionacional.es/visita/jardines-del-campo-del-moro,


Accessed: 15/11/2021

Nations, U., 2018, ‘Revision of World Urbanization Prospects’, Department of Economic and
Social Affairs, United Nations: New York, NY, USA

Olcina Cantos, J, 2019, ‘Evidencias e incertidumbres del Cambio Climático y de los riesgos
asociados en el litoral mediterráneo español’, Boletín de la Real Sociedad Geográfica, CLIV,
pp. 9-34

Olcina Cantos, J. Oliva Cañizares, A., 2020, ‘Medidas estructurales versos cartografía de
inundación en la valoración del riesgo en áreas urbanas: el caso del barranco de las Ovejas
(Alicante, España)’, Cuadernos Geográficos, 59 (2), pp. 199-220

Oliva Cañizares, A, Sánchez Almodóvar, E & Marcos Palacios, M.J., 2021, ‘La adaptación a
los extremos atmosféricos y al cambio climático mediante los sistemas urbanos de drenaje
sostenible (SUDS) y soluciones basadas en la naturaleza (SBN): propuesta en el término
municipal de Callosa de Segura (Alicante, España)’, Universitat d’Alacant, pp. 9-20

Oltra, C & Marín, R., 2013, ‘Los retos en la adaptación al cambio climático en entornos
urbanos’, ISSN 2013-9004, pp. 311-330

Ozturk, D. & Kiliç, F., 2011, ‘Implementation of GIS-Based multi-criteria Decision Analysis
with V.B. in ArcGIS’, International Journal of Information Technology and Decision Making ·
November 2011, pp. 1025-1026.

Saaty, T. L., 1977, ‘A scaling method for priorities in hierarchical structure’, Journal of
Mathematical Psychology, 15 (3), pp. 234-281

Saaty, T.L., 1994, ‘How to make a decision: The analytic hierarchy process’, Interfaces 1994,
24, pp. 19 - 43

55
Sari Kovats, R. & Hajat, S., 2008, ‘Heat Stress and Public Health: A Critical Review’, Public
and Environmental Health Research Unit (PEHRU), London School of Hygiene and Tropical
Medicine, London WC1E 7HT, Annual Review of Public Health Vol. 29, pp. 41-55

Sekar, I. & Randhir, T.O., 2007, ‘Spatial assessment of conjunctive water harvesting
potential in watershed systems’, Journal of Hydrology, 334 (1), pp. 39-52.

Senay, G.B. & Verdin, J.P., 2004, ‘Developing index maps of water-harvest potential in
Africa’, Applied Engineering in Agriculture, pp. 789 - 799

Sheehan, D, 2003, ‘Working With Interpolation Tools: IDW, Spline, and Kriging’, Updated
10/29/03 -C.A., 10/11/2006 -D.S., Web:
https://www.mit.edu/course/1/1.966/OldFiles/www/Lab4, Accessed 21/11/2021.

Smit, B. & Wandel, J., 2006, ‘Adaptation, adaptive capacity and vulnerability’, Global
Environmental Change, 16, pp. 282-292

Steffen, J., Jensen, M., Pomeroy, C.A., Burian, S.J., 2012, ‘Water supply and stormwater
management benefits of residential rainwater harvesting in U.S. cities’, J. Am. Water Resour.
Assoc. 49 (4), pp. 810-824

Suleiman, L., Olofsson, B, Saurí, D. & Palau-Rof, L., 2020, ‘A breakthrough in urban rain-
harvesting schemes through planning for urban greening: Case studies from Stockholm and
Barcelona’, ELSEVIER vol. 51, pp. 2-4

Vargas, A., Santos, A., Cárdenas, E & Obregón, N., 2011, ‘Analysis of distribution and spatial
interpolation of rainfall in Bogota, Colombia’, Dyna, año 78, Nro. 167, pp. 152-159

Weerasinghe, H., Schneider, U.A. & Löw, A., 2011, Water harvest-and storage-location
assessment model using GIS and remote sensing, Hydrology and Earth System Sciences
Discussions, 8 (2), pp. 3353-3381

Yoon, K.P., & Hwang, C.L., 1995, ‘Multiple Attribute Decision Making: An Introduction, SAGE
Publications, Thousand Oaks, pp. 31–37

Zabidi, H.A., Weng Goh, H., Kiat Chang, C., Weng Chan, N. & Azazi Zakaria, N., 2020, ‘A
Review of Roof and Pond Rainwater Harvesting Systems for Water Security: The Design,
Performance and Way Forward’, Water 2020 12(11), pp. 3-22

Ziadat, F., Bruggeman, A., Oweis, T., Haddad, N., Mazahreh, S., Sartawi, W. & Syuof, M.,
2012, ‘A participatory GIS approach for assessing land suitability for rainwater harvesting in
an arid rangeland environment’, Arid Land Research and Management, 26 (4), pp. 297-311

56
Annexes...
Annex I
Raw data information and sources.

Filename Type Description Source Scale


Land use Vector, Land cover data Spanish National Geographic Information 1:15.000
Shapefile Centre (CNIG)

Slope Raster Slope map Spanish National Geographic Information 1:50.000


Centre (CNIG)

Pluviometry Vector, Pluviometry State Meteorological Agency (AEMET) 1:25:000


stations Point

Soil type Vector, Geological and Geological and Mining Institute of Spain 1:1.000.000
Shapefile lithological data (IGME)

Rivers Vector, line River location in Spanish Ministry of Agriculture and 1:25.000
the city Fisheries, Food and Environment
(MAPAMA)

Annex II
Data source detailing the rainfall information of the different pluviometry stations of Madrid.

V D H P R
T P COORDENA COORDENA
V V R B S LONGITU
CÓDIGO ESTACIÓN DIRECCION (8 (8 COD_VIA DA_X_ETRS DA_Y_ETRS LATITUD
(8 (8 (8 (8 (8 D
3) 9) 89 89
1) 2) 6) 7) 8)

Plaza de
28079004 Plaza España X 273600 439579,3291 4475049,263 -3.7122567 40.4238823
España

Entre C/
Escuelas
28079008 Alcalá y C/ O’ X X 18900 442117,2366 4474770,696 -3.6823158 40.4215533
Aguirre
Donell

C/ Arturo
Soria esq. C/
28079016 Arturo Soria X X 798700 445786,1729 4476796,019 -3.6392422 40.4400457
Vizconde de
los Asilos

Calle Farolillo
28079018 Farolillo X 1903 437891,6961 4471832,769 -3.7318356 40.3947825
- C/Ervigio

Casa de
Casa de Campo
28079024 X X X X X X X 905219 436598,5637 4474571,618 -3.7473445 40.4193577
Campo (Terminal del
Teleférico)

Plaza del
Plaza del
28079035 Carmen esq. X X 145800 440346,3619 4474524,357 -3.7031662 40.4192091
Carmen
Tres Cruces.

57
Avd.
Moratalaz
28079036 Moratalaz esq. Camino X X 522000 445245,513 4473237,349 -3.6453104 40.4079517
de los
Vinateros

Avd. Pablo
Cuatro Iglesias esq.
28079038 X X 554550 440033,4632 4477450,211 -3.7071303 40.4455439
Caminos C/ Marqués
de Lema

Avd.
Betanzos
Barrio del
28079039 esq. C/ X X X 514425 439689,0496 4481081,619 -3.7115364 40.4782322
Pilar
Monforte de
Lemos

Avda La
Ensanche de
28079054 Gavia / Avda. X X X X X X 31000516 448033,2263 4469339,044 -3.6121394 40.3730118
Vallecas
Las Suertes

Pza. Elíptica -
28079056 Plaza Elíptica X X X X X X 293200 438991,9161 4470741,548 -3.7187679 40.3850336
Avda. Oporto

Avda. La
28079058 El Pardo X X 348500 434381,6 4485548,7 -3.7746101 40.5180701
Guardia

Parque Juan
Carlos I
(frente
28079059 Juan Carlos I X X X X X X X 447744,4666 4479077,678 -3.6163407 40.4607255
oficinas
mantenimient
o)

-
J.M.D. C/ Fuente
28079102 X X X X X X X 315060 445947,89 4472195,4 3.6356370 40.39979278
Moratalaz Carantona, 8
5

-
J.M.D. C/ Arroyo
28079103 X X X X X X X 74200 439633,65 4466785,76 3.7095247 40.3506278
Villaverde Bueno, 53
6

E.D.A.R. La
28079104 Embajadores X X 262000 442177,66 4468453,71 -3.679722 40.3658333
China

Centro Mpal. Autovía M-30


28079106 X X X X X X X 31001349 437130,68 4476974,19 -3.74 40.442222
De Acústica Km. 21.700

J.M.D. Ctra. de
28079107 X X X X X X X 133455 444215,2 4479200,02 -3.656667 40.462778
Hortaleza Canillas, 2

C.D.M.
28079108 Peñagrande X X X X X X X 514425 439037,69 4480778,49 -3.717881 40.4766333
Peñagrande

J.M.D.Chamb Plaza de
28079109 X X 227900 440772,75 4475799,91 -3.69695 40.4319111
erí Chamberí, 4

28079110 J.M.D.Centro C/ Mayor, 72 X X 497800 439584,05 4473998,69 -3.710792 40.4156

C/ Principe
J.M.D.Chama
28079111 de Vergara, X X 610450 442171,33 4474760,82 -3.6803686 40.4226491
rtin
142

J.M.D.Valleca Avda.
28079112 X X 18600 443329,69 4471846,08 -3.666456 40.396472
s1 Albufera, 42

J.M.D.Valleca Avda.
28079113 X X 18600 443329,69 4471846,08 -3.666456 40.396472
s2 Albufera, 42

Paseo de la
28079114 Matadero 01 X X 230600 440680,48 4471430,55 -3.697631 40.3925444
Chopera, 10

Paseo de la
28079115 Matadero 02 X X 230600 440680,48 4471430,55 -3.697631 40.3925444
Chopera, 10

58
ANNEX III
Kriging process information and detail.

59
60
ANNEX IV
Soil type source data file legend.

61
TRITA – ABE–MBT–2228

www.kth.se

You might also like