Factor-based probability model for vulnerability assessment of slopes subjected to earthquakes

This study explores the slope vulnerability assessment subjected to earthquakes using an environmental factor-based probability model based on data from the September 21, 1999, Chi-Chi earthquake-stricken areas of Taiwan. Satellite images were used to identify and interpret the collapsed area in the study area, and strong ground motion station data were used to evaluate the PGA. A DEM was used for topographic analysis to establish 6333 slope units. These slope units were classified into 12 categories based on environmental factors. Finally, a deterministic probability model based on the classification of environmental factors was presented. Landslide fragility functions were estimated for the 12 slope classifications. The results indicated that the important factors affecting the Chi-Chi earthquake were geology, aspect, and gradient. Geological factors indicate that the landslide probabilities of dense soil and soft rock are higher than those of hard and brittle rocks. The strong and weak aspects of the slope direction indicated that the difference in landslide probability was between about 0.05 and 0.1. The gradient factor indicates that steep slopes have a higher landslide probability.