Seasonally Optimized Interferometric Network Stacking-InSAR for Subsidence Detection of Mined-Out Zones in Alpine Terrain: A Case Study of the Songhu Iron Mine, Xinjiang, China

Human-induced mined-out areas are widely distributed across alpine regions, making reliable deformation detection particularly challenging. However, widespread low coherence and seasonal variations in snow cover cause severe spatiotemporal decorrelation and phase-unwrapping errors, substantially limiting the applicability of conventional InSAR techniques in these regions. To address these limitations, this study proposes an SCA-constrained seasonal partitioning Stacking-InSAR method based on the relationship between snow cover and interferometric coherence. The proposed method was applied to the Songhu Iron Mine in Xinjiang, China. Snow cover area (SCA) was first extracted from high-resolution Planet optical imagery and subsequently used to partition the interferometric network by season, thereby selecting high-quality summer interferometric pairs. This screening procedure reduced the interferometric network from 590 pairs to 168 high-quality pairs. The subsequent Stacking-InSAR results clearly delineated a spatially coherent surface subsidence zone within the mined-out area. The maximum line-of-sight (LOS) subsidence rate derived from the Stacking-InSAR analysis reached −132.31 mm/year. Overall, the proposed optical SCA-constrained seasonal partitioning strategy enabled Stacking-InSAR to identify a distinct subsidence funnel in an alpine region characterized by low coherence while maintaining high interferometric quality. The proposed method provides valuable technical and data support for deformation monitoring and geological hazard early warning in complex alpine environments.