Multi-scale analysis of the susceptibility of different landslide types and identification of the main controlling factors

Abstract

Landslides are common and extremely destructive geological hazards in mountainous areas, posing a serious threat to life, property, and socioeconomic development. Landslide susceptibility mapping is a key link in landslide risk management, but existing studies are mostly limited to single landslide types and single scale analysis, which makes it difficult to fully reflect the complexity and diversity of landslide occurrence. This paper takes Zhenxiong County, Yunnan Province as the study area, and constructs a multi-type and multi-scale landslide susceptibility mapping method by integrating information value (IV) model and principal component analysis (PCA) model. It reveals the scale effect of landslide driving force and the impact of this scale effect on the accuracy of susceptibility mapping, and identifies the main controlling factors of landslides at different scales. The results of the study showed that: (1) The influencing factors of the three landslide types have significant scale effects. Natural factors such as topography and geology play a controlling role at the buffer zone scale of 800–1000 m, while inducing factors such as human activities and rainfall mainly play a role at the buffer zone scale of 100–600 m. (2) The scale effect of influencing factors significantly affects the accuracy of landslide susceptibility mapping. The AUC value of soil landslide shows a trend of first increasing and then decreasing with the increase of scale, and the optimal AUC value is 0.815 at the 600-meter scale; the AUC value of rock landslide shows a trend of fluctuating and then increasing with the increase of scale, and the AUC value at the 1000-meter scale reaches the highest value of 0.877; the AUC value of ground collapse fluctuates with the expansion of scale, and the AUC value at the 800-meter scale is 0.883, which is the highest value among all scales. (3) Different types of landslides are controlled by different dominant factors. Slope height, rainfall and engineering geological rock formations are the main controlling factors for soil landslides; topographic factors such as slope height and DEM are the main controlling factors for rock landslides; DEM and engineering geological rock formations are the main controlling factors for ground collapse. The multi-type and multi-scale landslide susceptibility mapping method proposed in this paper can effectively identify the main controlling factors and scale effects of different types of landslides, improve the accuracy and reliability of landslide susceptibility mapping, and provide a scientific basis and decision-making support for landslide disaster prevention and control in Zhenxiong County.