Advanced Time-Series InSAR Deformation Estimation Integrating Multi-Reference Points and Sparse Model Optimization

Time-Series Interferometric Synthetic Aperture Radar (TSInSAR) is a microwave remote sensing technology capable of large-scale, high-precision, and periodic deformation monitoring under all-weather conditions, widely utilized in geological hazard detection. Focusing on the challenges of landslide deformation monitoring, specifically, significant residual errors in mountainous regions and low computational efficiency due to spatially sparse deformation signals. this study proposes an improved TSInSAR deformation estimation method integrating multi-reference point constraints and sparse model optimization. The method addresses residual errors through multi-reference point adjustment and leverages the sparsity of deformation signals by reformulating the inversion problem into a sparse optimization framework via a block matrix strategy, which is efficiently solved using the proximal gradient method. Experimental validation with LuTan-1 (LT-1) SAR data over the Outang landslide in Chongqing, China, demonstrates that the proposed approach achieves a 33% improvement in monitoring accuracy to conventional methods, confirming its reliability for practical landslide risk assessment and deformation mechanism analysis.