Evaluation of Different Landslide Susceptibility Analysis Methods: A Case Study of Bagmati Rural Municipality

Abstract

This study is focused on the evaluation of the different statistical methods of landslide susceptibility analysis in the Himalaya. For the evaluation, Bagmati Rural Municipality is selected as a typical study site. Various statistical methods are used to determine the relationship between thirteen landslide causative factors (viz., slope, aspect, profile curvature, tangential curvature, Normalized Difference Vegetation Index, Normalized Difference Water Index, elevation, distance from fault, distance from river network, distance from road network, geology, land use pattern and soil type) and landslide occurrence. Furthermore, a landslide inventory map of 154 landslides was prepared and 70% landslides were randomly selected for generating a model and the remaining 30% were used for validation proposes. These factor maps and the inventory are analyzed in a geographic information system. The evaluation between the three statistical bivariate methods was performed using the success rate and prediction rate. The study revealed that all susceptibility analysis techniques performed well, with the Certainty Factor Method having the best success and prediction rates for the study area, at 90.53% and 92.07%, respectively.