Distribution characteristics of snow and evaluation of snowdrift susceptibility along the Xinku section of Yining-Aksu Railway in China

Abstract

The frequent occurrence of snowdrift hazards in the Tianshan region of China often poses a serious threat to transportation corridors, such as roads and railways. Revealing and understanding the characteristics of snow distribution and using a mixed model to assess the susceptibility to snowdrifts can effectively and quickly evaluate the regional risk of the snowdrift. In this study, the key factors affecting the formation of a snowdrift in the study area, such as snow, are analyzed by utilizing remote sensing interpretation, meteorological monitoring, and field investigation along the Xinyuan-Kutaikelike railway section of Yining to Aksu Railway in China as the study area object. Based on the multi-source data, ten snowdrift susceptibility evaluation factors were extracted with the help of GIS. On this basis, three susceptibility models were used to evaluate the snowdrift susceptibility, namely the subjective evaluation method – Analytic Network Process (ANP) method, the objective evaluation method - CRITIC weighting method, and the combination of subjective and objective Game theory combination method. The evaluation showed that the susceptibility indices obtained by the ANP, CRITIC, and Game theory combination were in the ranges of [0.243, 0.746], [0.202, 0.751], and [0.249, 0.735] respectively, with the percentages of the area of medium susceptibility and above being 67.3%, 67.2%, and 67.5%. All three models exhibited suitable robustness. The predicted results of the three models were verified by the field survey, and it was found that the results of the Game theory combined model were in good agreement with the field survey, which allows it to be suitable for the study of snowdrift susceptibility zoning in the study area. The results revealed that the medium-high susceptibility areas of a snowdrift were mainly in the grassland area of the Small Yultuz Basin and the A'erxian-Bayinguoleng area. The high susceptibility zones were predominately located in the Nalati Ridge, the northern ridge of the Koktag Mountains, the northern ridge of the Huola Mountains, and the pre-mountain alluvial and floodplain areas of these mountain systems. This study provides a scientific basis and methodology for preventing and mitigating snowdrift disasters in cold regions, specifically for major engineering constructions such as transportation corridors.