Automated Time-Series InSAR Monitoring and Failure Prediction of Active Landslides: A Case Study of the Xinmo Landslide, China

Landslides, a prevalent type of geological hazard, are often sudden and highly destructive, posing serious threats to human life and infrastructure, particularly in mountainous regions with complex geology and active tectonics. To overcome the limitations of conventional monitoring methods—such as low timeliness, limited spatial coverage, and manual intervention—this study presents an automated preprocessing workflow for time-series Interferometric Synthetic Aperture Radar (InSAR) analysis based on Sentinel-1A data. The workflow is implemented within a Graphics Processing Unit (GPU) acceleration processing framework and includes automated data preparation, interferogram generation, and sequential deformation analysis. Building on this foundation, a landslide detection approach is developed by integrating deformation rate thresholding, time-series consistency analysis, and spatial clustering using the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm. Time-series classification and inflection point detection are then applied to characterize deformation evolution patterns. To support early warning applications, an Inverse Velocity Method (IVM) is used to estimate potential failure time in regions exhibiting accelerating displacement. The framework is validated using the June 24, 2017 Xinmo landslide in Mao County, Sichuan Province, where pre-failure deformation reached approximately 80 mm and predicted failure time was 2-3 days ahead of the actual event. The results indicate that the proposed framework enables reliable deformation monitoring, trend classification, and failure prediction under complex geomorphological conditions. This study provides a scalable technical solution for automated landslide detection and early warning using remote sensing.