Geological Society of America, Toronto, October 1998

Alteration induced debris flows at Pico de Orizaba, Mexico

HUBBARD, Bernard E., SHERIDAN, Michael F.,
Department of Geology, SUNY at Buffalo, Buffalo, NY 14260
Carrasco-Nunez Gerardo, Instituto de Geologia, UNAM, Mexico, DF, Mexico

Pico de Orizaba (Citlaltepetl) has a history of multiple catastrophic collapses (Carrasco-Nunez, 1993). The largest collapse produced a cohesive debris flow containing clays and other minerals derived from hydrothermal alteration of the volcano edifice (Carrasco-Nunez and others, 1993). The modern edifice exposes zones of hydrothermally altered rock that we identified using an AVIRIS dataset acquired 26 November 1994. Conversion of the original dataset to the broadband equivalents of the VNIR and SWIR channels of Landsat TM provides easy visualization with good resolution. The resulting image was geometrically corrected, re-projected to a UTM base, and fit to a DTM constructed from a TIN derived from a similar topographic base. This multispectral image facilitates their identification of clay and iron minerals in altered rocks by their distinct spectral characteristics. The ratio of the AVIRIS equivalents of TM bands 3/1 have high brightness values where these minerals are present. This image also displays outcrops of iron oxide or hydroxide minerals (limonite, hematite, geothite or jarosite) on several steep slopes around the summit cone, including the Torrecillas peaks south of the Citlaltepetl crater and the Espolon de Oro (Sarcofago) peak north of the Citlaltepetl crater. The ratio of the AVIRIS equivalents of TM bands 5/7 indicate the presence of clay minerals or alunite in small linear zones on the eastern sector of the Citlaltepetl cone. This suggests structural control by a northeast trending fault or other pre-existing fracture. Exposures of highly weathered lavas, breccias and volcaniclastic deposits in deeply incised barrancas on the more humid and tropical eastern valleys of the proximal zone, also present a hazard for earthquake-induced debris flows.

Mexican Geophysical Union, Puerto Vallarta, October 1999

CARTOGRAFÍA GEOLÓGICA APLICADA A LA EVALUACIÓN DE PELIGROS VOLCÁNICOS. EL CASO DEL VOLCÁN CITLALTÉPETL (PICO DE ORIZABA).

Carrasco-Núñez, G.^¨1, Hubbard, B.², Sheridan, M.² y Galicia, C.¹

1 UNICIT, Canpus UNAM Juriquilla, 76230, Qro. gerardoc@conin.unicit.unam.mx.
2 Dept. Geology SUNY at Buffalo, NY 14260, EUA bhubbard@eng.buffalo.edu; mfs@eng.buffalo.edu

Una adecuada evaluación de los peligros volcánicos debe estar basada primordialmente en la reconstrucción de la historia eruptiva del volcán en estudio, para lo cual es necesario establecer una estratigrafía detallada de las diferentes unidades de roca que lo conforman. La presentación de la información geológica a través de mapas geológicos convencionales normalmente no es suficiente para mostrar la naturaleza y potencial eruptivo de un volcán, ya que muchos de los depósitos representativos de eventos catastróficos, o son fácilmente erosionables o bien su tamaño es tan pequeño que su registro geológico es escaso o nulo, resultando comúnmente despreciable para fines cartográficos. Es por ello que, la información cartográfica geológica debe adecuarse en función de las características propias del volcán, como son: su estructura, estilo eruptivo, dimensiones, diversidad y abundancia de las unidades de roca existentes, escala, etc. Es también importante considerar la incorporación de mapas de depósitos representativos de eventos mayores, así como un número suficiente de columnas estratigráficas representativas ubicadas alrededor del volcán, con lo cual se podrán identificar con mayor precisión la magnitud, frecuencia y posible ciclicidad de los eventos eruptivos que ocurrieron en el pasado. El empleo de las herramientas modernas en las que se conjunta un exhaustivo trabajo de campo con la fotogeología y el procesamiento digital de imágenes de satélite con los modelos digitales de elevación del terreno, ha facilitado enormemente el manejo de información diversa (geología, geoquímica, geomorfología, demografía, etc.) que permite una visualización integral del vulcanismo y su potencial amenaza a las áreas circundantes.

En este trabajo se presentan diversos mapas geológicos a diferente escala y con diferente detalle cartográfico, así como mapas individualizados de depósitos mayores asociados con la actividad eruptiva reciente del volcán Citlaltépetl, y una serie de columnas estratigráficas que, junto con la aplicación de modelos de simulación de diversos procesos volcánicos, han permitido definir una zonificación preliminar de los peligros volcánicos para el volcán Citlaltépetl.

American Geophysical Union, San Francisco, December 1999

A comparison of two spectral analysis methods for identifying and mapping hydrothermal alteration zones at Pico de Orizaba (Citlaltepetl), Mexico.

B.E.Hubbard ¹, D.R.Zimbelman ², J.K.Crowley ³, M.F.Sheridan¹, G.Carrasco-Nunez⁴

    1. Department of Geology, State University of New York at Buffalo, Buffalo, NY, USA
    2. G. O. Logic, White Salmon, WA, USA
    3. U. S. Geological Survey, Reston, VA, USA
    4. Unidad de Investigacion en Ciencias de la Tierra, UNAM, Queretaro, Mexico

Pico de Orizaba (Citlaltepetl) has a history of multiple edifice collapses, the last of which was clearly related to hydrothermal alteration (Carrasco-Nunez et al., 1993). This study utilizes Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) data to map altered rock exposures, first, by using band-ratio analysis techniques applied to Landsat TM-equivalent channels generated from the AVIRIS imagery, and second, by applying a spectral shape-fitting algorithm to the 224-channel AVIRIS image cube.

The 3/1 ratio of the AVIRIS TM band-equivalents was effective for enhancing areas enriched in limonite, hematite, geothite and jarosite. The ratio of the AVIRIS equivalents of TM bands 5/7 produced only marginal results after using several different contrast enhancements and correcting for path radiance effects. This was due in part to the similar spectral response of pixels dominated by dry vegetation in desert grass and subalpine zones with clay-rich altered rocks at higher elevations. Improved results were obtained after applying two image processing techniques: RGB to IHS conversion (Knepper & Raines, 1985) and Directed Principal Component Analysis (DPCA - Fraser & Green, 1987), both useful for identifying vegetated pixels and masking their contributions to the ratio image histograms.

The band-ratio results were validated after calibrating the original AVIRIS radiance data to scaled apparent reflectance and removing path radiance and sun-angle effects. The Spectral Angle Mapper (SAM) algorithm (Kruse et al., 1993) was used to match the shape of AVIRIS pixel spectra to a known library of mineral end-members (Clark et al., 1993). SAM mapping of the hydroxyl-bearing minerals in the SWIR (2.0-2.5 um) produced the most conclusive results verified by fieldwork, analysis of AVIRIS spectra, XRD and laboratory reflectance spectra of representative samples. After masking the effects of clouds, SAM mapping showed major areas affected by advanced argillic alteration, with kaolinite/smectite mixtures as the dominant pixel end-members. This type of alteration covers much of the south sector of the Citlaltepetl cone, and extends down to the remains of the Torecillas edifice and the Espolon de Oro peak (Sarcafago). Alteration on the north sector (most evident in the illuminated portion of the crater) is masked by snow and ice fields of the Jamapa Glacier. Additional fieldwork is needed to determine the extent to which this material represents in situ rock outcrops, or transported talus from higher elevations. Pixels dominated by kaolinite, Na-montmorillonite and natroalunite are restricted to the uppermost part of the summit cone. SAM mapping of the iron-bearing minerals was less conclusive because of the marked similarity in reflectance spectra between the different forms of hematite, geothite and jarosite. However, the results partly match those areas identified in the band-ratios. Jarosite- or geothite-bearing rocks appear to be restricted to the remains of the Torecillas edifice, whereas hematite dominated pixels appear at all lower elevations in the vicinity of both Torecillas and Sarcofago.

Application of a GIS Model for Predicting the Impact of Medium to Large Debris Flows from Pico de Orizaba (Citlaltepetl), Mexico.

B.E. Hubbard ¹, M.F. Sheridan¹, G. Carrasco-Nunez², K.M Scott³

1. Department of Geology, State University of New York at Buffalo, Buffalo, NY, USA
2. Unidad de Investigacion en Ciencias de la Tierra, UNAM, Queretaro, Mexico
3. Cascades Volcano Observatory, U. S. Geological Survey, Vancouver, WA, USA

Pico de Orizaba has produced numerous debris flows in the past. The magnitude and origin of these debris flows vary from large-scale edifice collapses, the transformation and remobilization of pyroclastic flows and pyroclastic deposits, to moderate and small-scale earthquake- and storm-triggered slope failures. On January 3, 1920 earthquake triggered landslides occurred in the rugged highlands 30 km northeast of Pico de Orizaba. These landslides transformed to debris flows along the Huitzilapan-Pescados drainage. Laboratory analysis of a sample of this deposit near Barranca Grande, indicates the composition of a cohesive debris flow with clay-size particles ~ 5 percent of the matrix.

The LAHARZ model of Iverson et al. (1998) was used to constrain the volume of this debris flow and determine the possible impact that future events of this magnitude and larger would have on the more populated drainages to the east and south of Pico. The model uses statistically derived scaling factors to plot the cross-sectional and planimetric areas of debris flows on a DEM. Two criteria were used to constrain the volume of the 1920 seismogenic event: 1) minimum planimetric area required to produce a runnout past the Rio Pescados highway bridge 30 km downstream, and 2) minimum cross-sectional area to produce hydraulic radii of 40 to 65 m deep, as reported by Camacho (1922) and Flores (1922). Simulations were made on a georectified, Defense Mapping Agency DEM with a ~90 m grid spacing and +- 30 m vertical accuracy. Results of our best-fit simulations were based on a volume of 4.4 x 10⁷ m³, which resulted in a planimetric area of 2.3 x 10⁷ m² and a cross-sectional area of 6.2 x 10³ m². Such a model debris flow produces a maximum flood wave height of 64 m above the town of Barranca Grande, which was completely destroyed in 1920.

To evaluate hazards in other areas, debris flow inundation zones of four magnitudes were simulated along the Jamapa and Tliapa drainages north and east of Pico. These volumes represent debris flows equivalent to the Teteltzingo lahar from Pico (1.8 x 10⁹ m³), Electron mudflow from Mount Rainier (2.5 x 10⁸ m³), 1920 seismogenic debris flow (4.4 x 10⁷ m³), and the 1897 Bolum Creek debris flow from Mount Shasta (1.5 x 10⁶ m³). The latter three volumes were also simulated along the Metlac-,Orizaba-, and Carbonera-Rio Blanco drainage network to the south of Pico. Terraces associated with these drainages are compatible with passage of debris flows with magnitudes in the range of those modeled. The town of Coscomatepec appears to be mainly at risk from the largest model debris flows of the order of magnitude of the Teteltzingo lahar. In contrast, the city of Orizaba is at high risk from all but the smallest modeled debris flows. This city is located on the floodplain of the Orizaba river. The hydraulic radius of the largest model lahar (10⁸ m³) through downtown Orizaba would be about 46 m. At the same location, the maximum flood wave height of a debris flow comparable to the 1920 seismogenic event would be about 16 m.