Landslide susceptibility analysis and verification using the Bayesian probability model

The Bayesian probability model, using the weight-of-evidence method, was applied to the task of evaluating landslide susceptibility using GIS data. The location chosen for the study was the Janghung area in Korea, which suffered substantial landslide damage following heavy rain in 1998. Using the location of the landslides as well as topographic factors such as soil, forest and land use, the weight-of-evidence method was used to calculate each factor's rating. Tests of conditional independence were performed for the selection of the factors, allowing the large number of combinations of factors to be analyzed. For each factor rating, the contrast values, W+ and W–, were overlaid for landslide susceptibility mapping. The results of the analysis were verified using the observed landslide locations and, among the combinations, the slope, aspect, curvature, soil material and wood types show the best results.