A Landslide Susceptibility Evaluation of Highway Disasters Based on the Frequency Ratio Coupling Model

A landslide disaster, especially a highway landslide, may greatly impact the transport capacity of nearby roads. Keeping highways open, in particular, is crucial for supporting functioning economy, society and people. Therefore, evaluating susceptibility particularly important. In this paper, city Laibin, Guangxi Zhuang Autonomous Region China, was taken as study zone. According to data on 641 disaster points measured field basic evaluation area, nine factors—the elevation, slope, aspect, height difference, plan curve, profile precipitation, Topographic Wetness Index (TWI) vegetation coverage—were selected. We coupled Frequency Ratio (FR) model, Analytic Hierarchy Process (AHP), Logistic Regression (LR), Back Propagation Neural Network (BPNN) Support Vector Machine (SVM) evaluate landslides, with Receiver Operating Characteristic (ROC) curve used analyze precision these models. The ROC showed that accuracy five models greater than 0.700 thus had certain reliability. Among them, FR-LR model highest accuracy, at 0.804. protocol presented here can therefore provide reference studies other areas.