Abstract
The construction of transportation infrastructure lines alters the natural terrain, leading to uneven redistribution of snow under the influence of snow-drifting, which can compromise road safety. With the ongoing expansion of transportation networks in cold regions, transportation engineers are increasingly focusing on snow-drifting phenomena and the associated disasters. This study, based on a GIS platform, conducts a case study of a transport infrastructure project in Xinjiang to analyze the geographical and environmental conditions along the route. Through on-site monitoring and investigation, factors related to the occurrence of snow-drifting disasters were implied. The WOE (Weight of Evidence) model was selected as the base evaluation model, and the BP-GA algorithm was applied to optimize the weights of evaluation indicators. This led to the establishment of a susceptibility evaluation index system for snow-drifting disasters, improving both the computational efficiency and the accuracy of the evaluation model. The results indicate that the evaluation accuracies of the WOE, WOE-BP, and WOE-BP-GA models were 72.32%, 75.32%, and 85.18%, respectively. The use of the GA-BP algorithm effectively captured the complex nonlinear relationships among various factors, producing evaluation results highly consistent with real-world conditions. This method may efficiently identify high-risk areas of snow-drifting along transportation infrastructure lines, providing valuable insights for disaster prevention using ArcGIS.