Early Identification of Potential Dangers of Loess Landslide Based on Multi-Source and Time Series InSAR

Landslide disasters are frequent and widespread in mountainous areas of China, especially those potential disasters and dangers at high altitudes, where traditional technologies are less effective in identification and monitoring. Interferometric Synthetic Aperture Radar (InSAR) technology, as a ground observation technique based on a wide-area surface, can rapidly acquire minor and slow ground deformations over large areas, offering innate advantages over point monitoring techniques and playing a significant role in the identification of landslide risks. This study focuses on the Yecheng area of Xinjiang, utilizing 10 scenes of ALOS-2 data and 98 scenes of Sentinel-1 data. Based on the SBAS-InSAR method, identification and monitoring of geological hazards and potential landslide risks were conducted. By interpreting the deformation results in conjunction with optical remote sensing images, a remote sensing interpretation standard was established, revealing 22 potential landslide sites with deformation characteristics. Field verification confirmed 20 of these sites, achieving an identification accuracy rate of 91%. Detailed analysis of the time series deformation and causes at two typical risk sites based on deformation characteristics and field verification results showed a general trend of slow creep, with the potential for accelerated deformation in the event of rainfall or snowmelt. The results indicate that multi-source InSAR technology effectively identifies potential landslide risks in the Yecheng area, providing reliable data support for subsequent landslide disaster prevention and control measures.