Landslide Susceptibility Evaluation Based on Weights of Evidence Method for the Zaoshi Reservoir in Shimen County, Hunan Province

Landslide geological disasters occur frequently in the Zaoshi reservoir area, Shimen County of Hunan Province. The evaluation of landslide susceptibility by evidence right method can provide a scientific basis for landslide prevention and control. By using historical field revision data of landslide spots, remote sensing image, topographic map, DEM and geological map, we extracted 9 evidence factors on ArcGIS, including slope, slope shape, relative elevation of slope, vegetation coverage, stratum lithology, the models of slope structure, fault buffer distance, road buffer distance and river buffer distance. The weights of evidence factors are determined and the distribution chart of landslide susceptibility are generated by overlaying posterior probability of evidence layer of each unit. The results show that: (1) the landslide susceptibility in the study area can be subdivided into four categories: high prone area, medium prone area, low prone area and very low prone area, whose area proportions are 7.5%, 20.6%, 54.9% and 17.0% respectively; （2) Based on the success rate curve method, the zoning accuracy is 84.7%, and the evaluation results are consistent with the distribution of disaster points.