Abstract
A methodology for detecting slope failure areas from high-resolution satellite images and digital elevation model (DEM) is introduced and the applicability of the method is examined. The pre- and post-earthquake IKONOS images and the DEM observed in the epicentral area of the 2004 Niigata-ken Chuetsu earthquake (MJ6.8) are used. Slope failures are visually observed in the image since vegetations are flowed away and soils are exposed on the surface in most of slope failure areas. Difference of Normalized Difference Vegetation Index (NDVI) between the pre- and post-earthquake IKONOS images is used to evaluate slope failures. The areas where NDVI is remarkably decreased are extracted as slope failure areas. DEM is used to overlay the images precisely and to reduce mis-detection on flat areas. The distribution of the detected areas shows good agreement with the slope failure areas manually detected from the aerial photographs. Almost 85% of the slope failure areas are correctly detected in the analysis.
