Landslide susceptibility assessment in reservoir areas using time-series InSAR: A case study of the Zigui–Yiling region in the three Gorges reservoir

Landslide hazards occur frequently in the Three Gorges Reservoir Region. Therefore, landslide susceptibility assessment is of great importance for hazard identification and risk control. Traditional assessment methods that rely only on static environmental factors cannot fully reflect recent slope activity. This study focused on the Zigui to Yiling section of the Three Gorges Reservoir area. It used the SBAS-InSAR technique to obtain surface deformation information from 2021 to 2024. PCC, VIF, TOL, and IGR were then used to select nine conditioning factors. Based on these factors, CNN, SVM, and RF models were constructed for landslide susceptibility assessment. On this basis, InSAR deformation information was further incorporated into the assessment. The results showed that most deformation rates in the study area fell within the range of -10 to 10 mm/a, while strong deformation was mainly distributed along the riverbanks in Zigui County and in its mountainous areas. Analysis of typical reservoir bank areas showed that slope deformation became significantly stronger during periods of rapid reservoir water level fluctuation. When the monthly rise rate of the reservoir water level reached its maximum, the deformation rates at the three representative points all reached their peak values. Among the three models, the CNN model performed best, with an AUC of 87.1%, which was higher than those of SVM (86.0%) and RF (86.7%). The very high susceptibility zones were mainly distributed along the riverbanks in Zigui County, in the mountainous area of southwestern Zigui County, and in central Yiling District. After InSAR deformation information was incorporated, the very high susceptibility zones became more concentrated in slopes with stronger deformation, while some relatively stable areas were adjusted to lower susceptibility levels. These results provide support for landslide monitoring, risk identification, and field investigation in the reservoir area.