Landslide Susceptibility Mapping Using Logistic Regression Model Based On Information Value for the Region Along China-Thailand Railway from Saraburi To Sikhio, Thailand

Abstract

The main purpose of this study is to map landslide susceptibility using the logistic regression model based on information value, for the region along China-Thailand Railway from Saraburi to Sikhio, Thailand. In this study, a total of 60 landslides identified from remotely sensed images were divided into two groups: a group of 80% for training and the left 20% for validation. Landslide hazardous areas were mapped using six landslide controlling factors by logistic regression model based on information value. The performance of the model was evaluated by Receiver Operating Characteristic (ROC) curve. The results showed the model could provide 81.8% and 79.4% success and prediction rates respectively, meaning the map behaved good performance. Furthermore, the two factors of river networks and geotechnical types had a higher impact on the occurrence of landslides compared with other factors. This landslide susceptibility map can be used for preliminary railway construction and landslide mitigation.