Abstract
Lahars are among themost serious and far-reaching volcanic hazards. In regions with potential interactions of lahars with populated areas and human structures the assessment of the related hazards is crucial for undertaking appropriate mitigating actions and reduce the associated risks. Modeling of lahars has become an important tool in such assessments, in particular where the geologic record of past events is insufficient. Mass-flow modeling strongly relies on digital terrain data. Availability of digital elevation models (DEMs), however, is often limited and thus an obstacle to lahar modeling.
Remote-sensing technology has now opened new perspectives in generating DEMs. In this study, we evaluate the feasibility of DEMs derived from the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) and the Shuttle Radar TopographyMission (SRTM) for lahar modeling
on Popocatépetl Volcano, Mexico. Two GIS-based models are used for lahar modeling, LAHARZ and a flow-routing-based debris-flow model (modified single-flow direction model, MSF), both predicting areas potentially affected by lahars. Results of the lahar modeling show that both the ASTER and SRTM DEMs are basically suitable for use with LAHARZ and MSF. Flow-path prediction is found to be more reliable with SRTM data, though with a coarser spatial resolution. Errors of the ASTER DEMaffecting the prediction of flow paths due to the sensor geometry are associated with deeply incised gorges with north-facing slopes. LAHARZ is more sensitive to errors of the ASTER DEMthan theMSF model. Laharmodeling with the ASTERDEMresults in a more finely spaced predicted inundation area but does not add any significant information in comparisonwith the SRTMDEM.
Lahars at Popocatépetl are modeled with volumes of 1×105 to 8×106 m3 based on ice-melt scenarios of the glaciers on top of the volcano and data on recent and historical lahar events. As regards recently observed lahars, the travel distance of lahars of corresponding volume modeled with LAHARZ falls short by 2 to 4 km. An important finding is that the travel distance of potential lahar events modeled with LAHARZmay differ by about 2 km when using SRTMor ASTER data because of varying lateral flow-volume distribution. As a consequence, verification and sensitivity analysis of the DEMis fundamental to deriving hazard maps from predicted modeled inundation areas. Because of the global coverage of this type of remote-sensing data, the
conclusion that both SRTM and ASTER-derived DEMs are feasible for lahar modeling opens a wide field of application in volcanic-hazards studies.
Huggel, C