InSAR monitoring of large gradient deformation in coalfield and phase unwrapping research

Small Baseline Subset Interferometric Synthetic Aperture Radar (SBAS-InSAR) is widely used for deformation monitoring in mining regions. However, phase unwrapping errors often limit its effectiveness in areas with large-gradient displacements. We propose a novel approach integrating spatio-temporal modeling to address this limitation and improve phase unwrapping. The Simplified Rectangular Dislocation Model (SRDM) simplifies the phase recovery process by reducing model parameters and improving efficiency in large-gradient deformation zones. Subsequently, the Power Exponential Knothe Model (PEKM) is applied to invert the deformation time series and restore the deformation phase from the time dimension. We validated this method using Sentinel-1 and Radarsat-2 data in the Majialiang Coalfield, China. The results demonstrate that the approach achieved 5% (Sentinel-1) and 14% (Radarsat-2) phase range extensions, elevating maximum detectable phase values from 120/100 rad to 160 rad for respective datasets, while the proportions of pixel points with restored phase exceeding 8 rad (− 0.035 m) were 17.4% and 32.1%. The proposed method outperformed conventional techniques in resolving phase discontinuities, demonstrating strong adaptability. This approach significantly enhances subsidence monitoring accuracy and supports sustainable mining operations in coal-dependent regions.