THE USE OF HIGH-RESOLUTION SATELLITE IMAGERY FOR DERIVING GEOTECHNICAL PARAMETERS APPLIED TO LANDSLIDE SUSCEPTIBILITY

摘要

In order to mitigate hazards of mass failure, the first step is the identification of potentially unstable slopes, resulting in a landslide susceptibility map. Satellite imagery is an important component in the derivation of critical parameters. Susceptibility maps can be constructed in a variety of ways, including multivariate statistics, heuristic models, and geotechnical models. Each method has been shown to successfully identify hazardous slopes. The method employing geotechnical data has attracted attention, but the problem is a lack of critical parameters such as angle of internal friction or cohesion. These data may be available for few selected slopes, but typically not over a spatially coherent area. This severely limits the use of geotechnical models for landslide susceptibility maps. This study addresses the potential of using remote sensing and in particular high-resolution satellite imagery to derive values for important variables such as angle of internal friction and soil cohesion that are necessary for the geotechnical approach and specifically for the application of the infinite slope method to landslide susceptibility assessment. A 2002 Quickbird image was analyzed with respect to land cover, and then reclassified according to specific geotechnical parameters, in particular: cohesion; angle of internal friction; and root cohesion. The latter has been shown to be of high significance in slope stability. The resulting map spatially depicts the factor of safety or F-value over the study area. This map is compared with a 2005 Quickbird image in which multiple failures that occurred after a fire and subsequent rain storms are visible. The correspondence between the F-value map and the Quickbird image is quite visible, but further studies in refining the parameters are needed. This research suggests that satellite images can be successfully used to derive reasonable values for critical parameters in a geotechnical stability model, thereby increasing the utility of the geotechnical approach in landslide susceptibility mapping.