Landslide susceptibility mapping using a neuro-fuzzy

This paper develops and applied an adaptive neuro-fuzzy inference system (ANFIS) based on a geographic information system (GIS) environment using landslide-related factors and location for landslide susceptibility mapping. A neuro-fuzzy system is based on a fuzzy system that is trained by a learning algorithm derived from the neural network theory. The learning procedure operates on local information, and causes only local modifications in the underlying fuzzy system. The study area, Boun, suffered much damage following heavy rain in 1998 and was selected as a suitable site for the evaluation of the frequency and distribution of landslides. Boun is located in the central part of Korea. Landslide-related factors such as slope, soil texture, wood type, lithology, and density of lineament were extracted from topographic, soil, forest, and lineament maps. Landslide locations were identified from interpretation of aerial photographs and field surveys. Landslide-susceptible areas were analyzed by the ANFIS method and mapped using occurrence factors. In particular, we applied various membership functions (MFs) and analysis results were verified using the landslide location data. The predictive maps using triangular, trapezoidal, and polynomial MFs were the best individual MFs for modeling landslide susceptibility maps (84.96% accuracy), proving that ANFIS could be very effective in modeling landslide susceptibility mapping. Various MFs were used in this study, and after verification, the difference in accuracy according to the MFs was small, between 84.81% and 84.96%. The difference was just 0.15% and therefore the choice of MFs was not important in the study. Also, compared with the likelihood ratio model, which showed 84.94%, the accuracy was similar. Thus, the ANFIS could be applied to other study areas with different data and other study methods such as cross-validation. The developed ANFIS learns the if-then rules between landslide-related factors and landslide location for generalization and prediction. It is easy to understand and interpret, therefore it is a good choice for modeling landslide susceptibility mapping, which are also of great help for planners and engineers in selecting highly susceptible areas for further detail surveys and suitable locations to implement development. Although they may be less useful at the site-specific scale, where local geological and geographic heterogeneities may prevail, the results herein may be used as basic data to assist slope management and land use planning. For the method to be more generally applied, more landslide data are needed and more case studies should be conducted.