Due to the unique climate and frequent human activities, the loess area in Yining County, Xinjiang, China experiences many landslides. In this study, the formation mechanisms and controlling factors of the landslides in a typical loess area in Yining were investigated, including two catastrophic landslides in 2017. Based on 0.6 m-resolution satellite images, the landslides in this area were identified using artificial visual interpretation methods. Nine influencing factors (elevation, slope angle, slope aspect, topographical position index, distance to a road, distance to a river, distance to a fault, rainfall and normalized difference vegetation index) were selected to assess the landslide susceptibility using a logistic regression (LR) model. Finally, the potential of the LR model for assessing landslide susceptibility was evaluated using a receiver operating characteristic (ROC) curve. The landslide susceptibility assessment results obtained from the LR model are consistent with the actual landslide distribution. The LR model provides a powerful method for assessing the landslide susceptibility in this area. The research methods and results can provide references for the prevention and mitigation of landslide disasters across the entire Yili Prefecture of China.
Thematic collection: This article is part of the Role of water in destabilizing slopes collection available at: https://www.lyellcollection.org/cc/Role-of-water-in-destabilizing-slopes