GEOMORPHOMETRIC CHARACTERIZATION OF MASSIVE LANDSLIDES IN AFGHANISTAN

Digital terrain analysis of mountain topography is widely utilized for mapping landforms, assessing the role of surface processes in landscape evolution, estimating the spatial variation of erosion, and for assessment and mapping of hazards. Numerous geomorphometry techniques exist to characterize terrain surface parameters, although their utility to characterize the spatial hierarchical structure of the topography and permit an assessment of the impact of landslides and other hazards in Afghanistan is not well developed due to data integration, scale and analytical reasoning limitations. Consequently, our objective was to characterize the morphometric characteristics of massive landslides in Afghanistan and evaluate the utility of geomorphometry to assess landscape impact potential. We used ASTER high-resolution satellite imagery and digital elevation models to map newly discovered massive landslides in Afghanistan. Geomorphometric analysis consisted of producing global statistics, hypsometric, and altitude function information. We also applied scale-dependent geomorphometric and object-oriented analysis to characterize the hierarchical spatial structure of the landslides. Specifically, we generated terrain objects that represent slope facets and elemental forms based upon curvature. Object-oriented analysis was used to characterize object properties accounting for object size, shape, and morphometry. The spatial overlay and integration of terrain objects at various scales defined the nature of the hierarchical organization. Terrain segmentation and the integration of multi-scale terrain information permited assessment of process domains and natural hazard potential. Our results reveal the ability to map and differentiate landslides with different landscape impact potential. Such information is critical for planning and decision making in military and redevelopment efforts in Afghanistan.