4th International Conference on Integrating GIS and Environmental Modeling
(GIS/EM4):
Problems, Prospects and Research Needs. Banff, Alberta, Canada, September
2 - 8, 2000.
GIS model for volcanic hazard assessment:
pyroclastic flows at Volcán Citlaltépetl, México
GIS/EM4 No. 69
Michael F. Sheridan
Bernard Hubbard
Gerardo Carrasco-Nuñez
Claus Siebe
Abstract
Volcán Citlaltépetl (Pico de Orizaba) with an elevation of 5675 m is the highest volcano in North America. Its most recent catastrophic events involved the
production of pyroclastic flows that erupted approximately 4000, 8500, and 13000 years ago. The distribution of mapped deposits from these events gives an
approximate guide to the extent of products from potential future eruptions. Because the topography of this volcano is constantly changing computer simulations
were made on the present topography using three computer algorithms: energy cone, Flow2D, and Flow3D. The Heim Coefficient (
),
used as a code parameter for frictional sliding in all our algorithms, is the ratio of the assumed drop in elevation (H) divided by the lateral extent of
the mapped deposits (L). The viscosity
parameter for theFlow2D and Flow3D codes was adjusted so that the paths of the flows mimicked those inferred from the mapped deposits. We modeled two
categories of pyroclastic flows modeled for the level I and level II events. Level I pyroclastic flows correspond to small but more frequent block-and-ash flows that
remain on the main cone. Level II flows correspond to more wide-spread flows from catastrophic eruptions with an approximate 4000 year repose period. We
developed hazard maps from simulations based on a Defense Mapping Agency DEM with a 90 m grid and a vertical accuracy of ± 30 m. Because realistic
visualization is an important aid to understanding the risks related to volcanic hazards we present the DEM as modeled by Flow3D. The model shows that the
pyroclastic flows extend for much greater distances to the east of the volcano summit where the topographic relief is nearly 4,300 m. This study will be used to plot
hazard zones for pyroclastic flows in the official hazard map that is in preparation.
Keywords
Citlaltépetl, computer simulations, DEM, hazard zones, interactive viewing, México, Pico de Orizaba, pyroclastic flows, volcano.
Introduction
Pyroclastic flows are incandescent mixtures of volcanic particles and gas that descend from collapsing eruption columns and volcanic domes at
speeds that can exceed 100 m/s (Sheridan 1979). The societal risk related to pyroclastic flows is a problem that public safety authorities around the
world must face several times a year (Tilling 1989). We have developed a methodology to evaluate the danger from these incandescent avalanches
and clouds by using computer models of their flow paths to map hazard zones. This technique requires an accurate digital elevation model of the
surface and reliable data on the flowing materials to constrain parameters for the computer simulations. Some recent pyroclastic flows that were used
to constrain the effects of terrain on our model parameters are: Mount St. Helens (1980), Unzen (1991), and Colima (1991 & 1994). In this paper we
evaluate the results from three different, but related, flow models: energy cone, Flow2D, and Flow3D. Our procedure greatly extends the forecasting
capability at poorly studied volcanoes by providing a rapid and easily interpreted product that is based principally on topography and flow parameters
from observed eruptions.
|
Figure 1. Location map for Pico de Orizaba. Cities: (G) Guadalajara, (M) Mexico City, (P) Puebla, (V) Veracruz. Active volcanoes: (SJ) San Juan de
Tepic, (CE) Ceboruco, (CO) Colima, (PC) Popocatépetl, (PO) Pico de Orizaba, (SM) San Martín Tuxla, (EC) El Chichón,
(TA) Tacaná.
|
Problem statement
Volcán Citlaltépetl (Pico de Orizaba) is located in the State of Veracruz at the eastern end of the Trans-Mexican Volcanic Belt (Figure 1). Its
elevation of 5675 m creates considerable hazard for gravity-driven materials like avalanches, mudflows, and pyroclastic flows. Although this volcano
has been relatively quiet since the Spanish conquest, there have been three well-documented catastrophic events that involved the production of
pyroclastic flows at approximately 4000, 8500, and 13000 years ago (Carrasco- Nuñez and Rose 1993). The approximate repose period of 4000 years
for catastrophic events justifies preparation for future activity of this type. Several cities with populations between 10,000 and 100,000 are scattered
around the volcano in the states of Veracruz and Puebla, Mexico (Figure 2). This paper is part of a larger effort to create a hazard map of Volcán
Citlaltépetl that is comparable to maps of other active volcanoes in Mexico such as Popocatépetl (Macías et al. 1995) and Colima (Martin del Pozzo
et al. 1995).
|
Figure 2. Detailed location of study area. Volcanic features: (SN) Sierra Negra, (PO) Pico de Orizaba, (LC) Cerro Las Cumbres, (SD) South Derrumbadas, (ND)
North Derrumbadas, (PI) Cerro Pinto, (CP) Cofre de Perote. Cities: (S) Ciudad Serdan, (O) Orizaba, (C) Cordoba, (D) Soledad Doblado, (R) Boca del Rio, (V)
Veracruz, (J) Jose Cardel, (L) Laguna Verde, (X) Xalapa, (P) Perote.
|
Methods
Models
Although the three models used in this paper calculate several aspects of pyroclastic flows, the main characteristics of pyroclastic flows that
concern hazard assessment are runout distance and flow coverage. Because pyroclastic flows destroy all life in their path, the simulated surface
covered by the maximum expected pyroclastic flow defines the hazard zone. All three models in our study depend on an accurate DEM of current
topography for flow calculation and visualization of flow distribution. We designed the output display format of both models and maps to facilitate
interpretation by public safety officials as well as scientists.
Energy cone
Malin and Sheridan (1982) proposed the energy cone model to mimic the 1980 blast eruption of Mount St. Helens. The principle is that the height of the
starting point of the flow (H) ratios to the length of the runout (L) as a type of friction parameter termed the Heim coefficient (m) after Albert Heim, the
originator of the energy line concept. The inclination of the energy cone is an angle defined by arctan (H/L). The intersection of the energy cone, originating
at the eruptive source, with the ground surface defines the distal limits of the flow. The vertical distance (h) between the ground surface and the energy cone
provides a means to estimate of the flow velocity in this model.
|
(Notation 1)
|
v2 = 0.5 g h, where v is flow velocity and g is gravitational acceleration.
|
For our level I simulation we used a Heim coefficient of 0.26 and for the level II hazard we used a value of 0.18. The model results are shown projected on the
topography in Figure 3 and mapped in Figure 4. The energy cone has been applied to the assessment of hazard at Vulcano, Lipari, and Vesuvius (Sheridan and
Malin 1983) among other volcanoes.
Flow2D
A weakness of the energy line models is that it assumes straight-line flow trajectories that pass through topographic obstacles. To resolve this
problem in their study of Colima and El Chichon volcanoes Sheridan and Macías (1992) developed the Flow2D code for use in small computers.
This code assumes that shear resistance (
) depends on both basal friction (a0) and viscosity parameters (a1).
|
(Notation 2)
|
= a 0 + v*a 1
|
Flow2D code mimics actual flow velocity better than does the energy line and it shows the runback of flows off large topographic barriers. It is also easily used
in 2D on personal computers. However this code is limited by the need to tabulate topographic profiles alone each flow path. Another disadvantage is that
Flow2D does not consider lateral movement of the pyroclastic flows. For our Flow2D simulations of the level I pyroclastic flows we used an average value of
0.3 for a0 and 0.01 for a1. For the level II flows we used average value of 0.5 for a0 and 0.01 for a1.
The results of the Flow2D calculations are shown in the hazard maps of Figure 4.
Flow3D
The Flow3D code (Sheridan and Kover 1997) provides velocity histories of particle streams along flow paths in three dimensions. Multiple flow
paths are incremented every 0.1 seconds across triangular elements using as many as three parameters to calculate shear resistance (): basal
friction (a0), viscosity (a1) and turbulence (a2).
|
(Notation 3)
|
= a 0 + v*a 1 + v 2*a 2
|
Flow3D simulations have been verified by laboratory tests and field studies at the 1991 Unzen eruption (Kover and Sheridan 1993) and the 1980 Mount St.
Helens pyroclastic flows. It has been applied to risk assessment at several other volcanoes including the creation of hazard maps at Popocatépetl (Macías et al.
1995) and Volcán Colima (Martin del Pozzo et al. 1995) and risk probability at Volcán Colima (Sheridan and Macías 1995, Saucedo et al. 1997). For the
level I pyroclastic flows we used a0 as 0.17 and a1 as 0.01. For the later level 2 flows we used a0
as 0.15 with no viscous effects. Bit-mapped and color coded overlays of multiple themes were used to produce realistic images and maps. The interactive platform of
Flow3D allows the observer to adjust the perspective and distance for the desired view. The lifelike appearance of scene facilitates the interpretation by non-professional
observers. The use of cities and towns as a layer in the model allows the estimation of potential loss of life and property in the model simulations.
These results are compared with those of the energy cone in Figures 3 and 4.
GIS procedures
FLOW3D and Energy Cone models were simulated on a topographic model constructed from a Defense Mapping Agency DEM with a 90 m grid spacing and
vertical accuracy of approximately ±30 m. The digital topography takes the form of a Dense Triangular Network (DTN) that, like a Triangular Irregular Network
(TIN), conforms well to the morphology of topographic barriers and resolves the deepest path of descent. But in contrast it resolves watersheds and drainage
networks like grid-based DEMs (Jenson & Domingue, 1988; Jones et al., 1990; Band, 1993).
The final energy cone model was vectorized and filtered to eliminate pixel outliers and small "islands" within hazardous areas. FLOW2D results were
digitized using the 1:50,000 scale INEGE maps as a base and major cities and towns as geographic control points. FLOW3D results were vectorized
directly from a geocoded, raster output of the program, taking into account shadow zones created by the multiple flow traces. All of the above themes
were assembled as vector overlays in ARCVIEW with other themes such as population centers and 500 m contours derived from the DEM.
Hazard map construction
Siebe et al. (1993) mapped the distribution of young pyroclastic flows on the west of Pico de Orizaba. Carrasco- Nuñez and Rose (1995) and
Carrasco- Nuñez (1999) expanded the mapped flows to include those to the east and showed that the repose period between this type of event is
about 4000 years. These mapped deposits provide an outer limit for products of catastrophic eruptions, here called level II hazards. Smaller and more
frequent pyroclastic flow eruptions that have a more limited extent (Hoskuldsson and Robin 1993) are here termed level I hazards. The parameters for
the three models in the level I and level II simulations are set to mimic the distribution of these deposits.
|
Figure 3. Visualization comparison of energy cone (shaded area) and Flow3D (colored flow threads) simulations of Level I and Level II type events. In the Level I
simulation the 4000 yBP deposits are shown in red. In the Level II simulation the 8,500 yBP deposits are shown in blue. Mapped deposits, digitized as vectors,
were color encoded to the topography using a raster bit map.
|
Discussion
There is a general agreement between the hazard maps that the three models produced. The energy cone map is easy to produce and it generates the
most conservative map. Shadow zones behind topographically high area are weak in this model and the perimeter is relatively regular. In contrast, the
Flow3D maps show a much stronger conformation of flow boundaries with present topography and shadow zones are pronounced. The weakness of
this model is that it has a strong dependence on weakly constrained parameters. Flow2D map boundaries generally fall between those of the other
two models.
These maps are very useful for forecasting future pyroclastic flow hazards at Pico de Orizaba. They should help authorities in Mexico to design their mitigation
plans for specific dangerous areas and to discuss the options with the general population. The question arises in the use of models, how much faith can be placed
on any particular model. A conservative approach would be to draw boundaries in a regular fashion to include all towns that lie close to a boarder and to ignore
the shadow zones forecast by the models that are more strongly influenced by topography. This would favor the energy line simulations. However, if the shadow
zones really are protected areas, there would be a considerable saving of resources during an emergency if these areas were treated as an asset rather than a liability.
|
|
Figure 4. Map showing Level I (left image) and Level II (right image) pyroclastic flow simulations. UTM coordinates are UL = 659450E, 2126045N,
LR = 710647E, 2076017N. Cities, towns, villages: (Z) Zoapan, (T) Tlachichuca, (B) Santa Ines, (A) Avalos, (F) San Francisco Independencia, (G)
San Miguel Ocotenco, (H) Ahuatepec de Camino, (S) Ciudad Serdan, (V) San Martin, Arcos Ojo de Agua, San Francisco Cuautlanzingo, (E)
Esperanza, (M) Maltrata, (U) Cuiyachapa, (L) Tetelzingo, (X) Xocotla, (C) Coscomatepec, (P) La Perla, (I) Ixhuatlancillo, (N) Santa Ana Atzacan,
(Y) Moyoapan, (O) metropolitan Orizaba. Topographic barriers: (GO) Cerro Gordo, (AX) Cerro Acaxapo, (C) Chichimeco, (SA) Sarcafago, (SN)
Sierra Negra, (TP) Torrecillas Peaks, (TE) Cerro Tepozteca.
|
Recommendations for future research
It is obvious that the effect of topography on the distribution of pyroclastic flows must be assessed in regard to development of hazard maps.
Existing and new models of pyroclastic flows should be constructed for volcanoes with a high probability of future pyroclastic flow eruption. The best
models for hazard application can then be chosen based on actual prediction performance.
The next step in the GIS assessment of pyroclastic hazards at Pico de Orizaba is to estimate potential loss of life and property using polygon data. A logical
extension beyond this would be to develop an interactive system for computation, visualization and communication on high-performance computers. Such a
system would allow scientists, civil protection officials and the general public to have real-time access to potential hazards with realistic visualization of various
scenarios at various scales. The hardware and interfaces for such a system currently exist at several research facilities in the USA.
Acknowledgments
The authors thank Michael Abrams of JPL for supplying the DEM and other digital data used in this study. Civil protection authorities in the city of
Xalapa and the state of Veracruz provided valuable field assistance, vehicles, and aircraft for the field stage of this study. Special thanks goes to
Jose Luis Murrieta and Sergio Rodriguez Elizarraras of the Universidad Veracruzana in Xalapa for providing hospitality and logistics throughout the
study. This work was conducted with financial aid from NASA grant NAG57579.
References used
Band LE. 1993. Extraction of channel networks and topographic parameters from digital elevation data. In Beven K, Kirkby MJ, editors. Channel Network Hydrology. John Wiley & Sons, pp
13-42.
Carrasco-Nuñez G, Rose WI. 1995. Eruption of a major Holocene pyroclastic flow at Citlaltépetl volcano (Pico de Orizaba), Mexico, 8.5-9.0 ka. J Volcanol Geotherm Res 69:197-215.
Carrasco-Nuñez G. 1999. Holocene block-and-ash flows from summit done activity of Citlaltépetl volcano, eastern Mexico. J Volcanol Geotherm Res 88:47-66.
Hoskuldsson A, Robin C. 1993. Late Pleistocene to Holocene eruptive activity of Pico de Orizaba, eastern Mexico. Bull Volcanol 55:571-587.
Jenson SK, Domingue JO. 1988. Extracting topographic structure from digital elevation data for geographic information system analysis. Photogrammetric Engineering and Remote Sensing.
54 (11):1593-1600.
Jones NL, Wright SG, Maldment DR. Watershed delineation with triangle-based terrain models. J Hydraulic Eng 116 (10):1232-1251.
Kover TP, Sheridan MF. 1993. Numerical Models for pyroclastic flows of Mt. Unzen, Japan. Geol. Soc. Amer. Abstracts with Programs 25:268.
Macías JL, Carrasco-Nuñez G, Delgado H, Martin del Pozzo AL, Hoblitt RP, Sheridan MF, Tilling RI. 1995. Mapa del Peligros del Volcán Popocatépetl. Informe Final, Instituto de Geofisica, UNAM, México, 12 p. & map.
Malin MC, Sheridan MF. 1982. Computer-assisted mapping of pyroclastic surges. Science 217:637-639.
Martin del Pozzo AL, Sheridan MF, Barrera D, Hubp JL, Selem LV. 1995. Mapa de Peligros Volcán de Colima, Instituto de Geofisica, UNAM, México.
Saucedo R, Komorowski J-C, Macías JL, Sheridan MF. 1997. Modeling of pyroclastic flows generated during the 1913 eruption of Colima Volcano, México. EOS, Trans Amer Geophys Union
78:F823.
Sheridan MF. 1979. Emplacement of pyroclastic flows -- A review. In Chapin CE, Elston WE, editors. Ash Flow Tuffs. Geol Soc Amer Special Paper 180:125-136.
Sheridan MF, Kover T. 1996. FLOW3D: A computer code for simulating rapid, open-channel volcanic flows: Proc. UJST workshop on the Technology of Disaster Prevention against Local Severe
Storms, Norman OK. 155-163.
Sheridan MF, Macías JL. 1992. PC software for 2-dimensional gravity-driven flows: Application to Colima and El Chichón Volcanoes, México. Second International Meeting on Volcanology, Colima, México. p. 5.
Sheridan MF, Macías JL. 1995. Estimation of risk probability for gravity-driven pyroclastic flows at Volcán Colima, México. J Volcanol Geotherm Res 66:251-256.
Sheridan MF, Malin MC. 1983. Application of computer-assisted mapping to volcanic hazard evaluation of surge eruption: Vulcano, Lipari, Vesuvius. In Sheridan MF, Barberi F, editors.
Explosive Volcanism. J Volcanol Geotherm Res 17:187-202.
Siebe C, Abrams M, Sheridan MF. 1993. Major Holocene block-and-ash fan at the western slope of ice-capped Pico de Orizaba volcano, México: Implications for future hazards: J Volcanol Geotherm Res 59:1-33.
Tilling RI. 1989. Volcanic hazards and their mitigation: Progress and problems. Reviews of Geophysics 27:237-269.
Authors
Michael F. Sheridan, Professor of Geology
Department of Geology, 876 Natural Science Complex, SUNY at Buffalo, Buffalo, NY 14260.
Email: mfs@geology.buffalo.edu, Tel: +1-716-645-6800 x6100, Fax: +1-716-645-3999.
Bernard Hubbard, Research Assistant
Department of Geology, 876 Natural Science Complex, SUNY at Buffalo, Buffalo, NY 14260.
Email: bhubbard@acsu.buffalo.edu, Tel: +1-716-645-6800 x6100, Fax: +1-716-645-3999.
Gerardo Carrasco -Nuñez, Professor
Unidad de Investigacion en Ciencias de la Tierra, Campus Juriquilla, UNAM, Querétaro, Qro.76001, México.
Email: gerardoc@unicit.unam.mx, Tel: +52 -5-623-4124, Fax: +52 -5-623-4101.
Claus Siebe, Professor
Instituto de Geofísica, UNAM, Ciudad Universitária, CP.. 04510, Coyoacán, México, D.F., México.
Email: csiebe@tonatiuh.igeofcu.unam.mx, Tel: +52-5-622-4146, Fax: +52-5-550-2486.