Volcanic hazard zonation of the Nevado de Toluca volcano, Mxico

L. Capra
a,
, G. Norini
a
, G. Groppelli
b
, J.L. Macas
c
, J.L. Arce
d
a
Centro de Geociencias, Universidad Nacional Autnoma de Mexico, Campus Juriquilla, 76230 Queretaro, Mexico
b
C.N.R. - Istituto per la Dinamica dei Processi Ambientali, Milano, Italy
c
Instituto de Geofsica, UNAM, CU, Mexico DF, Mexico
d
Instituto de Geologia, UNAM, CU, Mexico DF, Mexico
A B S T R A C T A R T I C L E I N F O
Article history:
Received 18 December 2007
Accepted 19 April 2008
Available online 6 May 2008
Keywords:
Nevado de Toluca volcano
Trans-Mexican Volcanic Belt
volcanic hazard
computer simulations
The Nevado de Toluca is a quiescent volcano located 20 km southwest of the City of Toluca and 70 kmwest of
Mexico City. It has been quiescent since its last eruptive activity, dated at 3.3 ka BP. During the Pleistocene
and Holocene, it experienced several eruptive phases, including ve dome collapses with the emplacement of
block-and-ash ows and four Plinian eruptions, including the 10.5 ka BP Plinian eruption that deposited
more than 10 cm of sand-sized pumice in the area occupied today by Mexico City. A detailed geological map
coupled with computer simulations (FLOW3D, TITAN2D, LAHARZ and HAZMAP softwares) were used to
produce the volcanic hazard assessment. Based on the nal hazard zonation the northern and eastern sectors
of Nevado de Toluca would be affected by a greater number of phenomena in case of reappraisal activity.
Block-and-ash ows will affect deep ravines up to a distance of 15 km and associated ash clouds could
blanket the Toluca basin, whereas ash falls from Plinian events will have catastrophic effects for populated
areas within a radius of 70 km, including the Mexico City Metropolitan area, inhabited by more than
20 million people. Independently of the activity of the volcano, lahars occur every year, affecting small
villages settled down ow from main ravines.
2008 Elsevier B.V. All rights reserved.
1. Introduction
The Quaternary Mexican volcanismis concentrated along the Trans-
Mexican Volcanic Belt (TMVB) (Fig. 1A), a continental volcanic arc that
has been active since 14 Ma (Ferrari et al., 1994). At least 14 active
volcanoes are present in Mexico, and most of these are located along the
TMVB (Fig. 1A). At present, only Volcn de Colima and Popocatpetl
exhibit persistent activity, with small eruptive columns (up to 58 kmin
each case) and short-runout pyroclastic ows that have not affected
populated areas (Saucedo et al., 2005; Macas and Siebe, 2005; Macas
et al., 2006). The other volcanoes, including Nevado de Toluca (hereafter,
NdT), are quiescent and apparently represent no threat to the
surrounding populations. However, the sudden reactivation of El
Chichn volcano in 1982 after a 550 yr quiescence period killed over
2000 people (Macas et al., 2008), suggesting that long-term dormant
volcanoes can become active in a very short time, with catastrophic
consequences due to previous dearth of basic studies, hazard maps,
emergency information programs, etc. The Popocatpetl and Colima
volcano hazard maps were the rst maps prepared in Mexico as a
consequence of the eruptive crises that occurred in these volcanoes in
1994 and 1991, respectively (Macas et al., 1995; Del Pozzo et al., 1996;
Sheridanet al., 2001a; Navarroet al., 2003). More recently, Sheridanet al.
(2001b, 2004) produced a hazard map for the Pico de Orizaba volcano in
which the hazard delineation was based primarily on ow simulations
that tookintoaccount themorerecent eruptiveactivity. At present, these
maps represent a fundamental scientic document that civil defense
authorities use incase of future volcanic crises. For the case of Nevado de
Toluca, the onlyavailable hazardmapwas reported by Capra et al. (2000,
2004) and Aceves Quesada et al. (2007). These maps were mostly based
on the stratigraphic record and morphology of the volcano.
The Nevado de Toluca is a quiescent volcano, located 20 km
southwest of Toluca, and 70 km west of Mexico City (Fig. 1B). The
volcano has been silent since its last eruptive activity, dated at 3.3 ka
BP(Macas et al., 1997), althoughminor fumarolic activity was reported
during the nineteenth century (Bloomeld and Valastro, 1977). The
reactivation of NdT could threaten more than 20 million people,
including the Mexico City metropolitan area, that some 10.5 ka years
ago was blanketed by more then 10 cm of pumice from the Upper
Toluca Pumice eruption, one the most violent Plinian eruption
occurred during the Holocene (Fig. 1B) (Cas and Wright, 1988; Arce
et al., 2003). The aimof this work is to present a hazardassessment that
includes future eruptive scenarios deduced from a detailed geological
map (Garcia-Palomo et al., 2002; Bellotti et al., 2004; Norini, 2006)
coupled with computer ow simulations for debris avalanches, block-
and-ash ows, falls, and lahars. All simulations were carefully tested
based on spatial distributions and thickness of past ows, and in some
cases the paleotopography was restored in digital elevation models
(DEMs) to better reproduce past events. The resulting hazard zonation
benets from accurate calibration and validation of the numerical
Journal of Volcanology and Geothermal Research 176 (2008) 469484
Corresponding author.
E-mail address: lcapra@geociencias.unam.mx (L. Capra).
0377-0273/$ see front matter 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.jvolgeores.2008.04.016
Contents lists available at ScienceDirect
Journal of Volcanology and Geothermal Research
j our nal homepage: www. el sevi er. com/ l ocat e/ j vol geor es
models coupled with high-resolution geological data and reliable
computer ow simulations. This compiled methodology can better
predict the extent of products of future volcanic activity in contrast to
previous research, which tended to be based either on numerical
modeling or on geological mapping (Waythomas and Waitt, 1998;
Moreno, 2000; Sheridan et al., 2004; Soeld, 2004).
2. The Nevado de Toluca volcano
2.1. Morphostructural features
The Nevado de Toluca is 4680 m high and is characterized by an
open crater, 1.5 to 2 km in diameter, EWelongated, which developed
in response to intense tectonic activity (Tenango Fault System, Fig. 2;
Garca-Palomo et al., 2000, Norini et al., 2004; Bellotti et al., 2006;
Norini et al., 2006). The Ombligo dome, residing in the crater's interior,
separates two lakes (Moon Lake and Sun Lake, Fig. 3A).
The volcano shows striking morphological differences on its anks
and two main morphological domains can be dened (Norini et al.,
2004). The southern ank of Nevado de Toluca volcano has an irregular
morphology, relatively at anddissectedby deeprectilinear valleys with
NNWSSE strikes associated to the TaxcoQueretaro Fault System
(Garca-Palomo et al., 2000; Bellotti et al., 2006). The altitude of the this
ank varies between 3800 and 2500 m a.s.l. and more than 50% of this
area has slopes greater than 20 (Norini et al., 2004). Ravines are deeply
eroded with depths up to 450 m (Barranca del Muerto, Fig. 2). All these
features led previous authors to consider this area as the remains of an
older volcanic structure called Paleonevado (Cantagrel et al., 1981;
Garca-Palomo et al., 2000). The western, northern, and eastern NdT
anks, in conjunction with the crater area, depict the second mor-
phological domain, and constitute the present active cone of the NdT
volcano. The northern and northwestern sectors of the volcano are
gentle, with slopes of 68, mostly formed by pyroclastic fans (Fig. 3B,C)
on the Toluca basin. The main drainages are radial with respect to the
cone. The eastern sector has a gentler slope but is more dissected, with
ravines up to 70 m deep (Arroyo Grande ravine, Fig. 2) due to the
presence of the Tenango Fault System (Garca-Palomo et al., 2000), an
active left-lateral transtensive structure (Norini et al., 2006).
The present morphological arrangement of the volcano may have a
great inuence on the hazard assessment, because the distribution
and runout distances of volcanic ows will depend primarily on the
ow dynamics and topography.
2.2. Stratigraphic record
The volcano started to grow at 2.6 Ma with andesitic to dacitic
effusive activity that ended at 1.1 Ma and led to the formation of
Paleonevado edice (Cantagrel et al., 1981; Garcia-Palomo et al., 2002;
Martnez-Serrano et al., 2004). After an intense erosive stage that
originated the emplacement of voluminous epiclastic sequences,
including two sector collapses (Capra and Macas, 2000), magmatic
activity was renewed 42 ka ago whit the formation of the recent
active cone of the NdT and the emplacement of the Pink Pumice Flow
(PPF) deposit (Macas et al., 1997). Fig. 4 shows a detailed geological
map compiled from previous works (Garcia-Palomo et al., 2002;
Bellotti et al., 2004; Norini, 2006), representing the last 50 ka of
volcanic activity that was characterized by different eruptive phases
(Fig. 5), including ve dome collapses dated at 37, 32, 28, 26.5, and
13 ka (Macas et al., 1997; Garcia-Palomo et al., 2002), at least three
lateral collapses (Macas et al., 1997; Norini, 2006), and four Plinian
eruptions at 36 ka (Ochre Pumice) (Garcia-Palomo et al., 2002), 21.7 ka
(Lower Toluca Pumice, LTP) (Bloomeld et al., 1977; Capra et al., 2006),
12.1 ka (Middle Toluca Pumice, MTP) (Arce et al., 2005), and 10.5 ka
(Upper Toluca Pumice, UTP) (Macas et al., 1997; Arce et al., 2003). The
eruptive sequence is crowned by a phreatomagmatic surge deposit
dated at 3.3 ka BP (Macas et al., 1997).
2.3. Origin, distribution and magnitude of major volcanic events
We next describe in detail past eruptions and associated deposits
related to the activity of the active cone of NdT (b50 ka), to illustrate
expected events in case of renewed volcanic activity. The activity of
the Paleonevado, characterized by voluminous lava ows, ended
approximately 1.2 Ma ago, a time interval too long to be considered as
a possible locus of a new eruption.
Dome collapse events have been characterized by summit dome
growth and destruction, associated or not with an explosive
component. The 37- and 28-ka events correspond to major dome-
destruction episodes that originated lithic-rich block-and-ash ow
deposits that traveled up to 20 km from the source emplacing up to
30 mthick deposits (Fig. 6A). These deposits are grouped in the map in
one unit because they usually represent vertical sections, being
impossible to lay out their individual distribution on the geological
map. Main outcrops are limited to quarries where material is
extracted for construction (Fig. 3C); it is thus difcult to determine
Fig. 1. A) Map of the Trans-Mexican Volcanic Belt (TMVB) showing the location of the Nevado de Toluca and other active volcanoes. Abbreviations are: Ce: Ceboruco; CVC: Colima
Volcanic Complex; Pa: Parcutin; NT: Nevado de Toluca; Jo: Jocotitln; Mx: Mexico City; Iz: Iztacchuatl; Po: Popocatpetl; PdO: Pico de Orizaba; CdP: Cofre de Perote. B) Landsat
image (RGB combination) of the central sector of the TMVB. Mexico City is located 70 km northwest of Nevado de Toluca and 40 km northeast of Popocatpetl, both active volcanoes.
White dotted lines refer to the 10-cm isopach map of three mayor plinian eruptions occurred at the Nevado de Toluca volcano: the Lower Toluca Pumice (LTP, at 21 ka BP), the Middle
Toluca Pumice (MTP, at 12.5 ka B.P.) and the Upper Toluca Pumice (UTP, at 10.5 ka BP) that covered the area occupied today by Mexico City. Abbreviations are: TF: Tenango Fault; TLF:
Tenango Lava Flow; CVF: Chichinautzin Volcanic Field; CR: Las Cruces Range. Black dotted line refers to the main Tenango fault intersecting Nevado de Toluca Volcano.
470 L. Capra et al. / Journal of Volcanology and Geothermal Research 176 (2008) 469484
the exact distribution of such deposits. Despite this limitation, it is
clear that these two main events were radially distributed as they crop
out along all main ravines. Macas et al. (1997) estimated a total
volume of 3 km
3
for each event, which includes several ow units
that correspond to a VEI of 4 (Volcanic Explosivity Index, Newhall and
Self, 1982). The deposit associated with the younger 13 ka BP dome
collapse crops out mainly on the eastern and northeast sectors. It is
easy to recognize because of its brilliant gray color and its stratigraphic
position, between the Lower Toluca Pumice (LTP) and the Upper
Toluca Pumice (UTP) Plinian fall deposits. It consists of a main-channel
facies up to 10 m thick, with several ow units of clasts-rich block-
and-ash owdeposits, and a lateral facies up to 4 mthick consisting of
a sand-sized unit with small amounts of pumices (Fig. 3C). This ow
traveled a maximum distance of 15 km and was dispersed mostly to
the north-eastern sector of the volcano. Avolume estimationyielded a
value of 0.11 km
3
(D'Antonio et al., 2008), which corresponds to a VEI
of 3, being a moderate eruption. All of these pyroclastic ows show a
similar runout that corresponds to an H/L (relation between the drop
height and the maximum runout) of 0.12. The 13-ka event was used to
calibrate computer simulations for pyroclastic ows because of its
stratigraphic control and good exposures. Previous works (Metcalfe
et al., 1991; Macas et al., 1997; Newton and Metcalfe, 1999; Caballero
et al., 2000) reported an ash owdeposit on the lacustrine sequence of
the Toluca basin. This deposit probably represents the product of an
Fig. 2. Shaded digital elevation model of the Nevado de Toluca Volcano (50 m in pixel resolution) showing main morphological and structural features. Black lines indicate the fault
traces related to the Tenango Fault System (Norini et al., 2006).
471 L. Capra et al. / Journal of Volcanology and Geothermal Research 176 (2008) 469484
ash cloud associated to the main block-and-ash ow. For such kind of
deposits, an H/L of 0.1 was estimated.
During the growth of the modern edice, the extrusion of the 28-ka
summit dome, produced the collapse of the edice with the
emplacement of a debris avalanche deposit followed by a sequence
of block-and-ash ows (Caballero, 2007). The best examples of this
collapse are outcrops on the eastern sector of the volcano (Fig. 6C). The
resulting debris avalanche deposit, primarily outcropping on the
Zaguan ravine, has a volume of 0.35 km
3
and an H/L value of 0.14. Other
debris avalanche deposits were observedonthe northwest andeastern
valleys, with estimated volumes of 0.3 km
3
and similar H/L ratios of
0.14 (Arroyo Grande and El Nopal deposits, Fig. 4, Norini, 2006).
The UTP represents the most violent and voluminous eruption of
NdT, with the emission of 8 km
3
of magma (Dense Rock Equivalent)
and the formation of a 42-km-high Plinian column that dispersed
material toward the ENE, blanketing the area occupied today by
Mexico City with a layer of pumice more than 10 cm thick (Arce et al.,
2003) (Fig. 1B). Associated with this eruptive column were collapsing
pumice ows that emplaced on main ravines (Fig. 4 and 7a). The
OmbligoDomewas extrudedat theendof this activity(Arceet al., 2003).
Reconstructions of this event give a VEI =6 (Arce et al., 2003), whichmay
be comparable to the Plinian phase of the 1991 Pinatubo eruption that
had climatic global effects. In fact, Arce et al. (2003) argued that the
Younger Dryas period, the last major glacial event at the limit of
Holocene, in Mexico occurred exactly after the UTP eruption. The LTP
(Fig. 8b) and the MTP represent minor Plinian eruptions, with volumes
(DRE) of 0.8 km
3
and 1.8 km
3
and column heights of 24 km and 21 km,
respectively, which dispersed material toward the east, over the Toluca
basin (Fig. 1B). These values give a VEI of 4 for such eruptions (Table 1).
Few data are available for phreatomagmatic events of NdT,
although this type of eruptive activity may occur in the future,
because two permanent lakes are situated in the crater area. The only
Fig. 3. A) Picture showing the crater of the volcano, with the Ombligo Dome, emplaced after the UTP plinian eruption, separating the Sun and the Moon Lakes. Picture taken fromthe
western rim of the crater; B) Panoramic viewof the NdT fromthe east. Notice the large amphitheater formed after the last sector failure and dome collapses, subsequently shaped by
glacier activity. C) Panoramic viewof the El Refugio quarry on the northern sector of the volcano, one of the best exposures of the eruptive sequence of the last 40 ka. Here is exposed
the 13 ka BP block-and-ash ow deposit that consists of up to ve different units, and it is crowned by the UTP.
472 L. Capra et al. / Journal of Volcanology and Geothermal Research 176 (2008) 469484
Fig. 4. Geological map showing the distribution of the younger (b42 ka BP) volcanic sequence of the NdT. The vertical black arrow in the legend indicates that block-and-ash ow
deposits are interbeded with Plinian deposits. For the UTP and the MTP deposits, only pumice owdeposits are represented. The evident absence of recent volcaniclastic deposits on
the southern sector of the volcano could be due to the high topographic gradient that is favoring erosion. Abbreviations on the map refer to debris avalanche deposits (DAD): AG:
Arroyo Grande; EZ: El Zaguan; EN: El Nopal (Modied from Bellotti et al., 2004).
Fig. 5. Diagram showing a simplied stratigraphic column of the b42 ka BP eruptive sequence and their C
14
ages (with relative error) highlighting volcanic quiescent intervals of
8000 yrs (i.e. between the LTP and the 13 ka BP dome collapse) (data compiled from Macas et al., 1997; Garcia-Palomo et al., 2002).
473 L. Capra et al. / Journal of Volcanology and Geothermal Research 176 (2008) 469484
known phreatomagmatic deposit was dated at 3.3 ka and has been
described as a surge deposit, dark gray in color, up to 1 m thick, lithic
rich, with a laminated basal layer and a more dilute upper horizon
(Macas et al., 1997) (Fig. 6B). This deposit crops out in only one point
on the western ank of the volcano. This eruption is probably
associated with a low- to medium-magnitude phreatomagmatic
eruption with the formation of a small eruptive column.
Primary lahars have been always associated with mainly eruptive
events, and originated from the remobilization of falls, pumice ows,
block-and-ash ows, and debris avalanche deposits in main ravines
and up to a distance of 15 km from the crater (Figs. 4 and 8A).
Finally, during the past century, heavy rains have remobilized large
amounts of unconsolidated material generating secondary lahars. For
example, in 1952, a rain-triggered ood affected the village of Pueblo
Nuevo, ontheeasternsector of thevolcano(personal communicationfrom
local people). In this same drainage, local authorities are now arranging
sand dikes to prevent future oods (Fig. 8B) that occur every rainy season.
3. Possible hazardous scenarios
Based on the stratigraphic record of the NdTeruptive activity during
the last 42 ka BP, at least 12 main eruptions have occurred at intervals
varyingfrom1to8ka (Fig. 5). The analysis of the composite stratigraphic
section of the volcano allows assessment of possible scenarios in the
case of a future eruption of the volcano. Plinian fallouts rest directly on
top of thick paleosols aroundthe volcano, withalmost no other previous
pyroclastic ow deposits at their base (Fig. 7C). This might indicate that
before Plinian eruptions at NdT, the crater was open, without a central
dome inside as attested by the small amounts of accidental clasts in the
lower parts of the Plinian fallouts (Arce et al., 2003; Arce et al., 2005;
Capra et al., 2006). At present, the volcano has an open crater with the
0.001 km
3
Ombligo Dome on its interior that was emplaced just after
the UTP eruption. Most of the block-and-ash ow deposits recorded in
the stratigraphy of the volcano were probably generated by discrete
collapses of central domes, as those observed in the 28 ka event. A
Fig. 6. Pictures showing some examples of pyroclastic deposits at NdT. A) Section at Zacango quarry, showing the 37 and 28 ka BP block-and-ash owdeposits crowned by the LTP and
UTP pumice fall deposits. B) 3.3 ka BP pyroclastic ow at Raices quarry. C) El Zaguan debris avalanche deposit.
474 L. Capra et al. / Journal of Volcanology and Geothermal Research 176 (2008) 469484
similar sequence of events was observed in other volcanoes, where a
growing dome and the associated hydrothermal system weakened the
main edice, inducing its collapse (Voight et al., 2002). At NdT, the dome
remnants reect a growing path guided by tectonic lineaments,
indicating a structural control during the magma ascent, as for the
direction of main collapses, which was guided by geometry and
kinematics of tectonic structures (Norini, 2006).
Based on the age of the described eruptions, a regular period of
recurrence cannot be easily identied, because this depends on the
type of volcanic activity prior to each quiescent period. Generally, it
can be stated that the volcano has remained silent up to 8 ka followed
by explosive eruptions. This volcano behavior indicates the need to
carefully evaluate its potential hazards. Several other volcanoes
exhibited long quiescent periods before cataclysmic eruptions, such
as Tambora volcano (Stothers, 1984; Sigurdsson and Carey, 1989).
Based on these observations, and according to the present morphol-
ogy of the volcano, the most probable scenario may be a small
pheatomagmatic explosion as the 3.3 ka BP event (vent opening) that
might be followed by 1) explosive activity with the formation of a
sustained column (VEI 46); or 2) opening of new eruptive fractures
(associated with active faults) and dome growth with the collapse of
the old edice, followed by the destruction of the new dome with
emplacement of block-and-ash ow deposits. Independently of the
potential eruptive scenario, supercial water will remobilize uncon-
solidated volcaniclastic material forming lahars. Because of the
absence of lava ows in the stratigraphic record of the recent cone,
they were not included in the hazard evaluation.
4. Computer simulations
The main hazardous events of NdT, in accordance with its eruptive
history are debris avalanches, pyroclastic ows (block-and-ash ows),
Fig. 7. Pictures of main fall deposits. A) UTP sequence showing intercalation of pumice fall and owdeposits. B) Detail of the LTP, characterized by a fall basal unit and a sequence of fall
and surge layers atop. C) Picture displaying fall deposits resting directly on top of thick organic paleosols.
475 L. Capra et al. / Journal of Volcanology and Geothermal Research 176 (2008) 469484
lahars and fallouts (Table 1). Several computational ow simulations
are discussed in the literature (Valentine and Wohletz, 1989; Dobran
et al., 1994; Itoh et al., 2000), but the programs used here for volcanic
ows are those that are particularly suitable to ow over real
topography and that give the best reproducibility. Several works
have already showed the importance of the DEM resolution on
computational routines (Stevens et al., 2002; Hubbard et al., 2006;
Davila et al., 2007). For the present work, computational routines were
performed on a DEM with a resolution of 50-m pixel obtained from
vectorial data at 1:50,000 scale (INEGI) (Fig. 2).
4.1. Dry granular ow: FLOW3D and TITAN2D
Block-and-ash ows are gravity-driven, partially uidized, and
composed of a basal dry-grain ow layer that follows topographic
features, and an upper, more dilute, turbulent layer that can reach
greater distances than the underlying ow. Debris avalanche ows are
dry granular ows originated by the collapse of a volcanic edice.
Different models have been proposed to explain the mobility of such
ows and how the energy generated dissipates during owage (i.e.
Hayashi and Self, 1992). The relationship between the drop height and
the maximum runout (H/L), also known as the apparent coefcient of
friction (Hsu, 1975), better describes their mobility, which generally
increases as the mass increases (Dade and Huppert, 1998). Interstitial
uid is generally absent or present at b10% of the total mass, so it does
not play a dominant role during transport.
Two routines are used to simulate debris avalanches and the basal
grain ow of pyroclastic density currents: FLOW3D (Kover, 1995) and
TITAN2D (Patra et al., 2005). The main difference between these two
routines is that FLOW3Dcalculates the shear resistance based on basal
friction (a0), viscosity (a1) and turbulence (a2) to reproduce sliding
block trajectory (velocity and maximum runout), but obviating the
collapsing mass. In contrast, TITAN2D simulates the change in ow
thickness because it considers the initial volume of the collapsing
mass driven downslope by gravity where the resistance forces are
given by basal and internal friction angles of the collapsing material.
Comparisons between these two routines are provided for some
examples. Table 2 describes the parameters used during simulations.
4.1.1. Block-and-ash ows
Computer ow simulations were calibrated based on the distribu-
tion of the 13 ka block-and-ash deposit, since this represents the
youngest and best-exposed event. Although it does not represent the
most catastrophic event, it may represent the most realistic scenario
Fig. 8. A) Textural features of debris ow deposits at the Ojo de Agua laharic sequence. B) Sand dikes to reduce ood effects close to Pueblo Nuevo village (black arrow).
476 L. Capra et al. / Journal of Volcanology and Geothermal Research 176 (2008) 469484
(D'Antonio et al., 2008). The TITAN2D simulations were performed to
obtain the ow thickness and distribution based on a mass volume of
0.11 km
3
and an estimated basal friction angle between 8 and 10
(Table 2). These values were determined essentially based on the mass
runout of past ows (H/L) and compared with values used for the same
type of ows in previous works to obtain an estimation of the basal
friction angle (Rupp et al., 2006). The internal friction angle was xed
at 30, normal value for granular material, even if is not too relevant for
small granular ow(Hutter et al., 1995). Fig. 9A shows the distribution
of simulated ows where their depth is also indicated. These results
are similar to the thickness and runout of the observed deposits
(Fig. 4). The largest runout ows are observed in shallow ravines with
gentle topographic gradients, which these ows can easily inundate
(e.g. Cienega ravine, Fig. 9). The ow trajectories are different with
respect to past events, because of the actual conguration of the
crater. For instance, the Zacango ravine, one of the most studied places
because of its widely exposed pyroclastic sequence, probably will be
not greatly affected by block-and-ash ows in the future. In fact, a
dacitic dome is acting as a topographic barrier forcing any gravita-
tional ows far away from this ravine (white star in Fig. 4).
As described above, lacustrine sediments on the Toluca basin are
interbedded with ash layers, probably emplaced by the ash clouds
accompanying the main pyroclastic ows. As previously reported
(Malin and Sheridan, 1982; Saucedo et al., 2005), the energy cone can
be used to estimate the possible extension of this heavily diluted
density current. In this case, we used an H/L ratio of 0.1 (Fig. 9B).
4.1.2. Debris avalanches
To simulate this type of ow, we calibrated the simulation
parameters for TITAN2D considering the El Zagun and Arroyo Grande
debris avalanche deposits as prototypes (Fig. 4). We choose the Arroyo
Grande deposit to reconstruct the paleotopography and perform simu-
lations to exactly reproduce its extent (Fig. 10). In fact, debris avalanche
deposits are capable to completely change the topography, making it
impossible to simulate past ows over present DEMs. The paleotopo-
graphy of the Arroyo Grande gully was restored by eliminating the
thickness of the Arroyo Grande deposit and of the following overlapping
volcanic succession, as measured in several stratigraphic sections, after
which simulations with the FLOW3D and TITAN2D routines were
performed. The best-t H/L was 0.14, fromwhicha basal frictionangle of
8 was extrapolated (Table 2). For the FLOW3D routine, viscous and
turbulent coefcient were set equal to 0 (Table 2). Bothsimulations gave
similar results, with ows moving about 9 km from the source, and
stayedconned into the valley where thicknesses of 100 mare observed
on the proximal area. The concordance between these two simulations
suggests that for such low-volume (0.3 km
3
) granular ows, the
traveled distance does not depend strictly on the mass volume, as
previously generalized by other authors for large rock-falls (Dade and
Huppert, 1998). This point will be analyzed further in the next section,
because it has important hazard assessment implications. TITAN2D was
also used to simulate debris avalanche ows on other ravines, where
known deposits crop out (Fig. 10C, D), giving similar results between
simulated outputs and observed deposits.
4.2. Lahars: LAHARZ
Heavy rains are the most common triggering mechanism of lahars,
although other water sources are possible, such as the rupture of crater
lakes (Manvilleet al., 1999), temporaryvolcanic dams (Macas et al., 2004;
Capra, 2007), and glacial water outburst (Gudmundsson et al., 1998).
Depending on the sediment concentration, debris ows can form and
gradually transformto hyperconcentrated ows (Scott, 1988). LAHARZ is
a semi-empirical model, designed as a rapid, objective, and reproducible
automatedmethodfor mappingareas of potential lahar inundationbased
on the ow volume (Schilling, 1998; Iverson et al., 1998). It has already
been used by several authors to produce hazard maps (Sheridan et al.,
2001b; Samaniego et al., 2004; Hubbard et al., 2006).
Inorder to simulate the most probable scenario, the minimumand
maximum ow volumes used were 500,000 m
3
and 5,000,000 m
3
respectively (Fig. 11). The minimum value corresponds to the 1952
Pueblo Nuevo lahar deposit, while the maximum value represents
older lahar deposits exposed on the northern ank of the volcano
(Fig. 4) and associated with the 37 ka, 28 ka, and 13 ka dome collapse
events. The LAHARZ routine simulates the distal lahar inundation
zones (LIZ, Fig. 11), therefore, the red area on the volcano edice
corresponds to an H/L value of 0.17 (LPHZ, Fig. 11), which limits the
proximal area where erosive processes take place and where lahars
start to bulk and transform from uviatile or hyperconcentrated
ows to debris ows. The main parameter that determines the limit
of this proximal area is the break in slope that separates the main
cone from the volcanic apron where deposition takes place. Based on
the nal map, lahars will primarily affect the eastern and northern
sector of the volcano, including several villages. On the southern
anks, ravines are so deep that lahars cannot inundate vast areas, but
contrarily they do have larger runouts.
4.3. Fallout: HAZMAP
HAZMAP is a computer program for simulating the diffusion,
transport and sedimentation of volcanic particles in the atmo-
sphere in two dimensions from discrete point source (Macedonio
et al., 2005, http://datasim.ov.ingv.it/). The programdetermines the
particle settling velocity and based on eruption parameters (mass
discharge, column height, wind prole) and on physical conditions
(atmospheric diffusion coefcient and column shape) the mass
distribution is obtained. Specically, for the present case, we based
the simulation on the PC2 layer of the UTP eruption (VEI 6, Arce
et al., 2003), using two different wind proles that dominate the
area, one eastward (during spring and summer), and one westward
Table 1
Eruptive parameters of major eruptions during late PleistoceneHolocene
Type Age Runout
(km)
H/L 10-cm
isopach
area
(km
2
)
Column
height
(km)
Volume
(km
3
)
VEI Ref
Name
Dome collapse 37 ka yrs 17 0.12 n.a. n.a. 3 4 1
Dome collapse 28 ka yrs 17 0.12 n.a. n.a. 3 4 1
Flank collapse 28 ka yrs 9 0.14 n.a. n.a. 0.35 3 2
El Zagun
Plinian eruption 21.7 ka yrs n.a. n.a. 300 24 0.8 4 3
LTP
Plinian eruption 12.5 ka yrs n.a. n.a. 40 21 1.8 4 4
MPF
Dome collapse 13 ka yrs 15 0.12 n.a. n.a. 0.13 3 1, 5
El Refugio
Plinian eruption 10.5 ka yrs n.a. n.a. 2000 42 8 6 6
UTP
Abbreviations are: LTP: Lower Toluca Pumice; MPF: Middle Toluca Pumice; UTP: Upper
Toluca Pumice; n.a.; not applicable; VEI: Volcanic Explosivity Index. References are; 1,
Macas et al. (1997); 2, Caballero (2007); 3, Capra et al. (2006); 4, Arce et al. (2005); 5:
D'Antonio et al. (2008); 6: Arce et al. (2003).
Table 2
Parameters used for ow simulations with FLOW3D, TITAN2D and LAHARZ routines
Debris avalanches Pyroclastic ow Lahars
H/L (a
0
) 0.14 0.12
0.1 (ash could)
n.a.

b
8 1012 n.a.
v
0
(m/s) 0 15 n.a.
Volume (m
3
) 3.510
5
1.110
5
510
5
510
6
Abbreviations are:
b
: basal friction angle; v
0
: initial velocity; n.a. not applicable.
477 L. Capra et al. / Journal of Volcanology and Geothermal Research 176 (2008) 469484
Fig. 10. Image showing computer simulations for debris avalanches. Simulation of the Arroyo Grande debris avalanche with (A) FLOW3D and (B) TITAN2D model. Both simulations
were performed over a reconstructed paleotopography (see text for details). Note that distribution and runout are practically the same in both simulations. (C) Distribution of debris
avalanche deposits at NdT that is comparable with the TITAN2D simulations (D).
Fig. 9. Pyroclastic owsimulations. A) TITAN2D simulation for pyroclastic ows showing owthickness on main ravines. The parameters were calibrated with the 13 ka BP El Refugio
deposit. Higher thicknesses are exposed on the proximal areas on the cone slopes. CR: Cienega Ravine; ZR: Zaguan Ravine. B) Energy cone obtained with the FLOW3D application for
an H/L value of 0.1, the area represents the possible extension of ne ash related to pyroclastic ow.
478 L. Capra et al. / Journal of Volcanology and Geothermal Research 176 (2008) 469484
(during fall and winter). The meteorological data were obtained
from the Air Resource Laboratory web page (http://www.arl.noaa.
gov/ready/amet.html). The particle settling velocity was calculated
based on the granulometric distribution of F2 pyroclastic ow
emplaced just after the PC2 layer (Arce et al., 2003), which could
give a good approximation of the column grain-size characteristics
(Pfeiffer et al., 2005). Table 3 reports values used during simulation.
Fig. 12Ashows the isopachmapreproducingthe PC2layer. Black dots
indicate thickness measured in the eld (Arce et al., 2003), indicating a
good approximation of the simulated fallout. For distal areas it is much
more difcult to compare real and simulated data because of erosion of
the nest material, but previous authors reported about 10 cm of fall
deposits in Mexico City, as can be extrapolated from the simulation
(Fig. 12C). Fig. 12B shows the result of the simulation towards SW, wind
directionthat occasionallycharacterizes thearea, evenif noneof thefour
plinian eruptions so far described have this dispersal axis.
Fig. 11. Volcanic hazard zonation for lahars at NdT volcano. The LPHZ corresponds to the Lahar Proximal Hazard Zone, where lahars might formby eroding material on the main cone;
LIZ corresponds with the lahar inundation zone based on ow volumes as indicated in the legend.
Table 3
Main parameters used for fall simulations with HAZMAP routine
Number of settling velocity particles classes 17
Total mass emitted from point source (kg) 2E
12
Number of vertical steps 100
Height of the column (m) 40000
Diffusion coefcient (m
2
/s) 8000
Suzuki coefcient (A and ) 2.5 1
479 L. Capra et al. / Journal of Volcanology and Geothermal Research 176 (2008) 469484
5. Discussion
5.1. Hazard zonation
The nal hazard zonation of NdT volcano was compiled based on
the simulated scenarios previously described. A main problem for the
zonation is how to dene the different hazard levels. One potential
approach is to identify the more hazardous zones based on the
number of events that might affect it. In this sense, the main cone
would fall within the highest hazard zone, and distal ravines would be
exposed to the lowest hazard, since only lahars can inundate it. But
evidently, if no magmatic activity occurs at the volcano, lahars will
Fig. 12. Fallout simulation obtained with HAZMAP computer routine with same input parameters (Table 3), but with two different wind proles, A) towards NE and B) towards SW
(for explanation, see the text). Black dotted circles in gure a indicate fall deposit thickness measured for the PC2 layer of the UTP eruption (fromArce et al., 2003). Isopach values are
in meter .C) Landsat image showing the distribution of the simulated 10 cm isopach from a VEI 6 eruption.
480 L. Capra et al. / Journal of Volcanology and Geothermal Research 176 (2008) 469484
Fig. 13. Volcanic hazard map of NdT volcano displaying the areas of pyroclastic ows, fall, debris avalanches and lahars. The dimension of the gray circles is proportional to the number
of inhabitants, as indicated in the scale. Abbreviations are: LPHZ, lahar proximal hazard zone; LIZ, lahar inundation zone.
481 L. Capra et al. / Journal of Volcanology and Geothermal Research 176 (2008) 469484
represent the most probable hazard in the area. Based on these
considerations, two different zonations are proposed.
The rst refers to the actual quiescent state of the volcano and
considers only possible erosive processes that generate lahars. This
scenario corresponds with the lahar hazard delineation previously
described (Fig. 11). Towns such as San Juan de las Huertas, Zacango,
and Villa Guerrero will be affected by lahars produced by heavy rains,
as in the 1952 event. Towns such as San JuanTilapa and Tenango might
also be affected by this hazard, because they are located on the distal
fans of these main ravines susceptible to oods during heavy rains. In
the proximal area, erosive processes could cause signicant damage,
including landslides of the walls of main ravines, where several small
villages are settled. In fact, especially on the southern sector of the
volcano, planar areas are often occupied by greenhouses for cultiva-
tion of plants, and many people settled their houses on the steep
anks of ravines. At least 50,000 people could be affected by this type
of phenomenon.
The secondhazardzonation(Fig. 13) depicts the scenarios incase of a
future magmatic activity. The map delineates different zones based on
the types of events affecting certain areas. By comparing the FLOW3D
and TITAN2D simulations for debris avalanche, an energy line with H/L
of 0.14 was traced to include all possible trajectories and maximum
runout. But considering that past sector collapses have occurred to the
western, northwestern and eastern anks of NdT (Figs. 4 and 10), as the
volcano responses to the geometry and kinematics of the active
basement faults (Norini et al., 2006), the most feasible sectors for
collapse directions are the east-northwest (red dashed line, Fig. 13).
Even if block-and-ash ows could only affect ravines up to a 15 km
from the summit, the associated ash-cloud fallouts (H/L=0.1) could
threaten areas up to a distance of 30 km, including the towns of
Tenango, Metepec, Ixtapan de la Sal and Toluca, where approximately
500,000 people live (demographic data fromINEGI, http://www.inegi.
gob.mx/inegi/default.aspx).
Plinian activity represents the most catastrophic scenario, because
over a 1 m thick layer of pumice could blanket an area within a 30 km
radius fromthe vent that would damage the City of Toluca, capital of the
State of Mexico. But inadditiontolocal damage, suchaneruptive scenario
might completely paralyze the central part of the country affecting more
than 30 million of inhabitants, including Mexico City. A large number of
fatalities could occur at several rural villages over a radius of more than
70 kmdue tothe depositionof N10 cmof ash(Fig. 12), whichwill result in
the collapse of roofs, and health problems. Main airports such as those
located in Mexico City and Toluca will stop operations, as in 1997 when a
small ash plume from Popocatpetl volcano caused the Mexico City
airport to shut down for some hours, causing signicant economic loses.
Finally, ne ashes will be dispersed on the atmosphere provoking,
darkness and temporally climate alteration such as during the 1982
Chichneruptionthat causedthe temperature of the globe todecrease ca.
0.5 (Angell and Korshover, 1983).
Dome and ank collapses should be more predictable because the
volcano currently lacks a summit dome. In fact, in order to experience
this type of activity, a new dome must grow inside or outside the
crater. In contrast, the stratigraphic record (Figs. 5 and 7C) indicates
that Plinian eruptions have taken place with an open conduit similar
to the present morphology of the volcano that will be probably
preceded by phreatic and phreatomagmatic eruptions.
5.2. Computer simulation
An important result of this work resulted from the comparison
between the FLOW3D and TITAN2D routines for conned debris
avalanches. This analysis showed that the trajectory and runout of this
hazard is very similar (Fig. 10A,B) with both routines, even if FLOW3D
does not take into account the mass volume. This is an important
result because it supports the idea that volume does not likely affect
the travel distance of low-volume, conned, granular ows. By
plotting H/L values vs. volume of volcanic debris avalanches (Fig. 14,
data taken from Hayashi and Self, 1992), it is clear that a direct
relationship between volume and runout can be observed only for
voluminous debris avalanches. In contrast, low-volume debris
avalanches are scattered in the plot. This contrasting behavior could
indicate that the mobility of small masses is probably more dependent
on other factors, such as the material type, degree of fracturing,
presence of pyroclastic material, or clay fraction in the pre-failure
mass (Siebert et al., 1987; Ui, 1983) with respect to the mass volume
(Dade and Huppert, 1998). Based on this result, we concluded that the
energy line concept is still useful for hazard assessment of small
volcanic debris avalanches less than 1 km
3
in volume, a method that
Fig. 14. Diagram showing the relation between the H/L and volume for volcanic debris avalanches (data from Hayashi and Self, 1992).
482 L. Capra et al. / Journal of Volcanology and Geothermal Research 176 (2008) 469484
can be easily applied without the time-consuming calibration and
validation of numerical models.
6. Conclusions
The Nevado de Toluca hazard assessment benets from both
geological and computational analyses. The availability of an excellent
stratigraphic record allows evaluation of the expected scenario and
better constraint of the input parameters for numerical modeling of
hazardous volcanic processes. Computer simulations always give an
output, but these do not necessarily correspond to a good scenario of
the volcano eruptive behavior. For the case of the Nevado de Toluca
hazard zonation, the integration between geological data and
computer routines provides validation and calibration for numerical
models. This is the basic condition for reliability and good condence
of the hazard assessment.
Currently, NdT is in a quiescent state without any evidence of
renewed activity; however, NdT may become active in the future, with
seismic and volcanic events that can have catastrophic results for the
surrounding communities. Based on the stratigraphic record, the
volcano remained quiescent for periods up to 8000 years, being now
impossible to predict when the next eruption will be. The population
is now completely vulnerable to such events, because the volcano is
not monitored and no specic civil protection plans are known to be
reliable and ready for use. In this sense, the hazard zonations
presented here provides the rst and most basic document for 1)
land use planning; 2) civil protection defense plans preparation; and
3) volcanic crisis management.
The actual density of the seismic network in the area is too low to
ensure detection of the rst seismic signals associated with renewal of
volcanic activity, andno others types of monitoring (as interferometry for
deformationof the edice, or gravimetry, etc.) are conducted on a regular
basis. In the case of renewed volcanic activity, there is the possibility that
threatenedpopulations will be not alertedintimeandwill not receive the
most useful information. Thus, we encourage better surveillance of NdT
and other dormant volcanoes in the TMVB. Most of these volcanoes
shouldbe better studiedtounderstandtheir potential hazards, whichcan
be assessed in deep only after detailed geological work.
Acknowledgements
This work is dedicated to Armando Garcia-Palomo, our best friend
that largelycontributedtothe knowledge of Nevado de Toluca volcano.
The work was supported by CONACYT grant (projects 37889 and
46340) to L. Capra. The Ministry of Foreign Affairs of Italy and SRE of
Mexico provided travel assistance to Gianluca Groppelli and Gianluca
Norini. We greatly acknowledge Fernando Bellotti, Lizeth Caballero,
Micaela Casartelli, Marco D'Antonio, Andrea Gigliuto, Riccardo Lunghi,
Anna Merlini, and Damiano Sarocchi for their help during the
eldwork. Special thanks to Mike Sheridan and Marcus Bursik who
kindly provided TITAN2D and FLOW3D softwares, and Antonio Costa
for assistance during HAZMAP simulation. Marina Manea and Emilio
Nava also helped during data processing. Revisions by Alicia Felpeto,
Joan Marti and an anonymous reviewer substantially improved the
manuscript.
References
Aceves Quesada, F., Martindel Pozzo, A.L., Lopez Blanco, J., 2007. Volcanic hazards zonation
of the Nevado de Toluca Volcano, Central Mexico. Natural Hazards 41, 159180.
Angell, J.K., Korshover, J., 1983. Global temperature variations in the troposphere and
stratosphere, 195882. Monthly Weather Review 111, 901921.
Arce, J.L., Macas, J.L., Vzquez-Selem, L., 2003. The 10.5 ka Plinian eruption of Nevado de
Toluca volcano, Mexico: stratigraphy and hazard implications. Geological Society of
America Bulletin 115 (2), 230248.
Arce, J.L., Cervantes, K.E., Macas, J.L., Mora, J.C., 2005. The 12.1 ka Middle Toluca Pumice:
a dacitic Pliniansubplinian eruption of Nevado de Toluca in Central Mexico.
Journal of Volcanology and Geothermal Research 147 (12), 125143.
Bellotti, F., Capra, L., Groppelli, G., Merlini, A., Norini, G., 2004. Sector collapses in
composite volcanoes and relationships with the geological evolution and the
structural setting: an example from Nevado de Toluca Volcano (Mexico). IAVCEI
General Assembly 2004. Volcanism and its Impact on Society, Pucon, Chile,
November 15th19th, 2004.
Bellotti, N., Capra, L., Groppelli, G., Norini, G., 2006. Tectonic evolution of the Toluca
Basin and its inuence on the eruptive history of the Nevado de Toluca Volcano.
Journal of Volcanology and Geothermal Research 158, 2136.
Bloomeld, K., Valastro, S., 1977. Late Quaternary tephrochronology of Nevado de Toluca
volcano, central Mexico. Overseas Geology and Mineral Resources 46, 15.
Bloomeld, K., Snchez Rubio, G., Wilson, L., 1977. Plinian Eruptions of Nevado de
Toluca. Volcanology Geological Rundschau 66 (1), 120146.
Caballero, L., 2007. Anlisis textural del depsito de avalancha de escombros El
zagun, volcan Nevado de Toluca: dinmica de transporte y mecanismos de
emplazamiento. Master Thesis, Universidad Nacional Autnoma de Mxico, Mexico
City, pp. 176.
Caballero, M., Macias, J.L., Urrutia-Fucugauchi, J., Lozano-Garcia, S., Castaeda, R., 2000.
Volcanic stratigraphy and palaeolimnology of the Upper Lerma Basin during Late
Pleistocene and Holocene. Sedimentology 30, 5771.
Cantagrel, J.M., Robin, C., Vincent, P., 1981. Les grandes etapes d'evolution d'un volcan
andesitigque composite: Exemple du Nevado de Toluca. Bulletin of Volcanology 44,
177188.
Capra, L., 2007. Volcanic Natural Dams: identication, stability and secondary effects.
Natural Hazard 43, 4561.
Capra, L., Macas, J.L., 2000. Pleistocene cohesive debris ows at Nevado de Toluca
Volcano, central Mexico. Journal of Volcanology and Geothermal Research 102 (1/2),
149167.
Capra, L., Arce, J.L., Macas, J.L., Garca-Palomo, A., 2000. Mapa de peligros volcnicos del
Volcn Nevado de Toluca, Mxico. 2 Reunin Nacional de Ciencias de la Tierra,
Puerto Vallarta, Mxico, p. 353.
Capra, L., Groppelli, G., Lunghi, R., Norini, G., 2004. Volcanic hazards assessment of
Nevado de Toluca volcano (Mexico). 32nd International Geological Congress,
Florence, Italy, p. 84.
Capra, L., Carreras, L., Arce, J.L., Macas, J.L., 2006. The lower Toluca pumice: a 21,700 yr B.P.
plinian eruption of the Nevado de Toluca volcano, Mexico. In: Siebe, C., Macas, J.L.,
Aguirre-Diaz, G. (Eds.), NeogeneQuaternary Continental Margin Volcanism: a
Perspective from Mexico. The Geological Society of America, Penrose Conference
Series, vol. 402, pp. 155174.
Cas, R.A.F., Wright, J.V., 1988. Volcanic Successions (Modern and Ancient), Second
Impression. Unwin Hyman, London, 528 pp.
Dade, W.B., Huppert, H.E., 1998. Long-runout rockfall. Geology 26 (9), 769864.
D'Antonio, M., Capra, L., Sarocchi, D., Bellotti, F., 2008. Reconstruccin del evento eruptivo
asociado al emplazamiento del ujo piroclstico El Refugio hace 13 ka, volcn Nevado
de Toluca (Mxico). Revista Mexicana de Ciencias Geolgicas 25 (1), 115134.
Davila, N., Capra, L., Gavilanes, J.C., Varley, N., Norini, G., Gmez-Vzquez, A., 2007.
Recent lahars at Volcn de Colima (Mexico): drainage variation and spectral
classication. Journal of Volcanology and Geothermal Research 165, 127141.
Del Pozzo, A.L., Sheridan, M.F., Barrera, D., Hubp, J.L., Vzquez, L., 1996. Mapa de peligros
Volcn de Colima. Instituto de Geofsica, UNAM, Mexico D.F, Mexico.
Dobran, F., Neri, A., Todesco, M., 1994. Assessing the pyroclastic owhazard at Vesuvius.
Nature 367 (10), 551554.
Ferrari, L., Garduo-Monroy, V.H., Pasquar, G., Tibaldi, A., 1994. Volcanic and tectonic
evolution of central Mexico; Oligocene to present. Geofsica Internacional 33,
91105.
Garca-Palomo, A., Macas, J.L., Garduno, V.H., 2000. Miocene to Recent structural
evolution of the Nevado de Toluca volcano region, Central Mexico. Tectonophysics
318, 281302.
Garcia-Palomo, A., Macas, J.L., Arce, J.L., Capra, L., Espndola, J.M., Garduno, V.H., 2002.
Geology of Nevado de Toluca Volcano and Surrounding Areas, Central Mexico.
Geological Society of America Map and Chart Series, MCH089.
Gudmundsson, M.T., Sigmundsson, F., Bjrnsson, H., 1998. Icevolcano interaction on
the 1996 Gjlp subglacial eruption, Vatnajokul, Iceland. Nature 389, 954957.
Hayashi, J.N., Self, S., 1992. A comparison of pyroclastic ow and debris avalanche
mobility. Journal of Geophysical Research 97 (B6), 90639071.
Itoh, H., Takahama, J., Takahashi, M., Miyamoto, K., 2000. Hazard estimation of the
possible pyroclastic ow disasters using numerical simulation related to the 1994
activity at Merapi Volcano. Journal of Volcanology and Geothermal Research 100
(14), 503516.
Hubbard, B., Sheridan, M.F., Carrasco-Nuez, G., Diaz-Castellon, R., Rodriguez-
Elizarrars, S.R., 2006. Comparative lahar hazard mapping at Volcan Citlaltepetl,
Mexico using SRTM, ASTER and DTED-1 digital topographic data. Journal of
Volcanology and Geothermal Research 160, 99124.
Hsu, K.J., 1975. Catastrophic debris stream (Sturzstroms) generated by rockfalls.
Geological Society of America Bulletin 86, 129140.
Hutter, K., Koch, T., Pluuss, C., Savage, S.B., 1995. The dynamics of avalanches of granular
materials from initiation to runout. Part II: experiments. Acta Mechanica 109,
127165.
Iverson, R.M., Schilling, S.P., Vallance, J.W., 1998. Objective delineation of lahar-
inundation hazard zones. Geological Society of America Bulletin 110 (8), 972984.
Kover, T.P., 1995. Applications of a digital terrain model for the modeling of vol-
canic ows: a tool for volcanic hazard determination. Master Thesis, SUNY, Buffalo,
62 pp.
Macedonio, G., Costa, A., Longo, A., 2005. A computer model for volcanic ash fallout and
assessment of subsequent hazard. Computers & Geosciences 31, 837845.
Macas, J.L., Siebe, C., 2005. Popocatpetl's crater lled to the brim: signicance for
hazard evaluation. Journal of Volcanology and Geothermal Research 141, 327330.
483 L. Capra et al. / Journal of Volcanology and Geothermal Research 176 (2008) 469484
Macas, J.L., Carrasco-Nuez, G., Delgado, H., Del Pozzo, A.L., Siebe, C., Hoblitt, R.P.,
Sheridan, M.F., Tilling, R.I., 1995. Mapa de peligros volcnicos del Popocatpetl.
Instituto de Geofsica, UNAM, Mexico D.F., Mexico.
Macas, J.L., Garcia-Palomo, A., Arce, J.L., Siebe, C., Espindola, J.M., Komorowski, J.C.,
Scott, K.M., 1997. Late PleistoceneHolocene Cataclysmic Eruptions at Nevado de
Toluca and Jocotitln Volcanoes, Central Mexico. In: Kowallis, B.J. (Ed.), Proterozoic
to Recent Stratigraphy, Tectonics, and Volcanology, Utah, Nevada, Southern Idaho
and Central Mexico. BYU Geology Studies, pp. 493528.
Macas, J.L., Capra, L., Scott, K.M., Espndola, J.M., Garca-Palomo, A., Costa, J.E., 2004. The
May 26, 1982, breakout ow derived from failure of a volcanic dam at El Chichn
Volcano, Chiapas, Mexico. Bulletin of the Geological Society of America 116,
233246.
Macas, J.L., Saucedo, R., Gavilanes, J.C., Varley, N., Velasco Garcia, S., Bursik, M.I., Vargas
Gutierres, V., Cortes, A., 2006. Flujos piroclsticos asociados a la actividad explosiva
del volcn de Colima y perspectivas futuras. GEOS 25 (3), 340351.
Macas, J.L., Capra, L., Arce, J.L., Espindola, J.M., Garca-Palomo, A., Sheridan, M. in press.
Hazard map of El Chichn Volcano, Chiapas, Mexico. Journal of Volcanology and
Geothermal Research.
Malin, M.C., Sheridan, M.F., 1982. Computer-assisted mapping of pyroclastic surges.
Science 217, 637640.
Manville, V., White, J.D.L., Houghton, B.F., Wilson, C.J.N., 1999. Paleohydrology and
sedimentology of a post-1.8 ka breakout ood from intracaldera Lake Taupo, North
Island, new Zealand. Geological Society of America Bulletin 111, 14351447.
Martnez-Serrano, R.G., Schaaf, P., Sols-Pichardo, G., Hernandez-Bernal, M.S., Hernan-
dez-Trevio, T., Morales-Contreras, J.J., Macas, J.L., 2004. Sr, Nd, and Pb isotope and
geochemical data fromthe Quaternary Nevado de Toluca volcano, a source of recent
adakitic magmatism, and the Tenango Volcanic Field, Mexico. Journal of
Volcanology and Geothermal Research 138, 77110.
Metcalfe, S.E., Street-Perott, A., Perrott, A., Harkness, D.D., 1991. Palaeolimnology of the
Upper Lerma Basin, Central Mexico: a record of climatic change and anthropogenic
disturbance since 11600 yr BP. Journal of Paleolimnology 5, 197218.
Moreno, H., 2000. Mapa de peligros del volcn Villarrica, Escala 1:75.000, Regiones de la
Araucania y de Los Lagos. Sernageomin, Chile.
Navarro, C., Cortes, A., Tellez, A., 2003. Mapa de peligros Volcn de Fuego de Colima.
Observatorio Volcanolgico, Universidad de Colima, Colima, Mexico.
Newhall, C.G., Self, S., 1982. The volcanic explosivity index (VEI): an estimate of
explosive magnitude for historical volcanism. Journal of Geophysical Research 87,
12311238.
Newton, A.J., Metcalfe, S.E., 1999. Tephrocronologyof the Toluca Basin, central Mexico.
Quaternary Science Reviews 18, 10391059.
Norini, G., 2006. Relazione tra attivit tettonica e vulcanismo. Universit degli Studi di
Milano, PhD Thesis, 198 pp.
Norini, G., Groppelli, G., Capra, L., De Beni, E., 2004. Morphological analysis of Nevado de
Toluca volcano (Mexico): new insights into the structure and evolution of an
andesitic to dacitic stratovolcano. Geomophology 62, 4761.
Norini, G., Groppelli, G., Lagmay, A.M.F., Capra, L., 2006. Recent left-oblique slip faulting
in the eastern-central Trans-Mexican Volcanic Belt. Seismic hazard and geody-
namic implications. Tectonics 25, TC4012.
Patra, A.K., Bauer, A.C., Nichita, C.C., Pitman, E.B., Sheridan, M.F., Bursik, M., Rupp, B.,
Webber, A., Stinton, A.J., Namikawa, L.M., Renschler, C.S., 2005. Parallel adaptive
simulation of dry avalanches over natural terrain. Journal of Volcanology and
Geothermal Research 139, 122.
Pfeiffer, T., Costa, A., Macedonio, G., 2005. A model for the numerical simulation of
tephra fall deposits. Journal of Volcanology and Geothermal Research 140, 273294.
Rupp, B., Bursik, M.I., Namikawa, L., Webb, A., Patra, A., Saucedo, R., Macas, J.L.,
Renschler, C., 2006. Computational modeling of the 1991 block and ash ows at
Colima Volcano, Mexico. Geological Society of America Special Paper 402, 237252.
Samaniego, P., Monzier, M., Robin, C., Schilling, S., Eissen, J.P., 2004. Lahar hazard
assessment at Nevado de Cayambe volcano (Ecuador). Cities on Volcanoes, Quito,
Ecuador, p. 28.
Saucedo, R., Macas, J.L., Sheridan, M.F., Bursik, M.I., Komorowski, J.C., 2005. Modeling of
pyroclastic ows of Colima Volcano, Mexico: implications for hazard assessment.
Journal of Volcanology and Geothermal Research 139 (12), 103115.
Schilling, S.P., 1998. LAHARZ: GIS programs for automated mapping of lahar-inundation
hazard zones. U.S. Geol. Surv. Open File Rep., vol. 93638. 80 pp.
Scott, K.M., 1988. Origin, behavior, andsedimentologyof lahars andlahar-runout ows inthe
ToutleCowlitz River System. U.S. Geological Survey Professional Paper, vol. 1447-A(74).
Sheridan, M.F., Hubbard, B., Bursik, M.I., Abrams, M., Siebe, C., Macias, J.L., Delgado, H.,
2001a. Gauging short-term volcanic hazardz at Popocatpetl. EOS, Transaction,
American Geophysical Union 82 (16), 187188.
Sheridan, M.F., Carrasco-Nuez, G., Hubbard, B., Siebe, C., Rodriguez-Elizarrars, S.R.,
2001b. Mapa de peligros volcanicos del volcan Citlaltepetl (Pico de Orizaba).
Instituto de Geologia, UNAM, Mxico D.F., Mxico.
Sheridan, M.F., Hubbard, B., Carrasco-Nuez, G., Siebe, C., 2004. Pyroclastic ow hazard
at Volcn Citlaltpetl. Natural Hazards 33, 209221.
Siebert, L., Glicken, H., Ui, T., 1987. Volcanic hazards from Bezymianny- and Bandai-type
eruptions. Bulletin of Volcanology 49, 435459.
Sigurdsson, H., Carey, S., 1989. Plinian and co-ignimbrite tephra fall from the 1875
eruption of Tambora volcano. Bulletin of Volcanology 51, 243270.
Soeld, D., 2004. Eruptive history and volcanic hazards of Volcan San Salvador.
Geological Society of America Special Paper 375, 147158.
Stevens, N.F., Manville, V., Heron, D.W., 2002. The sensitivity of a volcanic owmodel to
digital elevation model accuracy: experiments with digitized map contours and
interferometric SAR at Ruapehu and Taranaki volcanoes, New Zealand. Journal of
Volcanology and Geothermal Research 119, 89105.
Stothers, R.B., 1984. The great Tambora Eruption in 1815 and its Aftermath. Science 224
(4654), 11911198.
Ui, T., 1983. Volcanic dry avalanche deposits identication and comparison with
nonvolcanic debris stream deposits. Journal of Volcanology and Geothermal
Research 18, 135150.
Valentine, G., Wohletz, K.H., 1989. Numerical models of Plinian eruption columns and
pyroclastic ows. Journal of Geophysical Research 94, 18671887.
Voight, B., Komorowski, J.C., Norton, G.E., Belousov, A., Belousova, M., Boudon, G.,
Francis, P.W., Franz, W., Heinrich, P., Sparks, R.S.J., Young, S.R., 2002. The 26
December (Boxing Day) 1997 sector collapse and debris avalanche at Soufrire Hills
Volcano, Montessat. In: Druitt, T.H., Kokelaar, P. (Eds.), The Eruption of Soufriere
Hills Volcano, Montserrat, from 1995 to 1999. The Geological Society of London,
Memories, London, pp. 363407.
Waythomas, C.F., Waitt Jr., R.B., 1998. Preliminary Volcano-Hazard Assessment for
Augustine Volcano, Alska. USGS Open-File Report, vol. 98106. 39 pp.
484 L. Capra et al. / Journal of Volcanology and Geothermal Research 176 (2008) 469484