GIS-based landslide susceptibility mapping using analytic hierarchy process and artificial neural network in Jeju (Korea)

Abstract

This paper presents a landslide susceptibility analysis in Jeju Island, South Korea using AHP (Analytic Hierarchy Process) and ANN (Artificial Neural Network) methods. To assess the landslide susceptibility, seven factors which affect the landslide occurrence were selected as: slope, aspect, soil type, geological type, rainfall intensity, forest and land cover. According to the relation between the above factors and landslide distribution, the weight value and rating value of each factor were calculated using AHP and ANN. Finally, we get two susceptibility maps using AcrView software through weighted overlay GIS (Geographic Information System) method. The comparative analysis reveals that the slope, rainfall intensity, soil and forest play important roles in landslide occurrence. From the results, we also found that AHP and ANN methods can be used to analyze landslide susceptibility using the data from the different area and the ANN method was more suitable and effective than AHP method.