Monitoring Surface Deformation and Classifying Activity of Deep-Seated Landslide Sites Using Multi-Temporal InSAR

Through the application of Multi-temporal Interferometric Synthetic Aperture Radar (MTInSAR) for long-term monitoring of surface deformation, the Agency of Rural Development and Soil and Water Conservation (ARGSWC), Ministry of Agriculture, has identified 315 deep-seated landslide sites with protecting targets across Taiwan as of 2024. For this study, imagery from January 2022 to October 2023 from the ALOS-2 satellite, released by the Japan Aerospace Exploration Agency (JAXA), is used for the MTInSAR monitoring. Based on the MTInSAR surface displacement data, activity indices for the deep-seated landslide sites have been developed, including both the arithmetic mean and inverse area-weighted deformations. The k-mean clustering and risk matrix method are then employed to classify and rank the landslide activity. The analysis reveals that approximately 5.8%, 10.7% and 83.6% of the deep-seated landslide sites are classified as high, medium and low activity, respectively. In addition, statistical clustering techniques are applied to group the surface deformation data, which are then compared to slope units derived from the aerial LiDAR Digital Elevation Model (DEM) for the landslide sites. This approach helps to identify active landslide blocks or subzones within the landslide sites.