Establishing Time-Series 2D Surface Deformation to Investigate the Characteristics and Mechanisms of Land Subsidence in the Central Taiwan

Over the past 50 years, the Choushui River Fluvial Plan (CRFP) has been plagued by land subsidence caused by excessive groundwater extraction in the central Taiwan. While many geodetic techniques have successfully monitored surface deformation in this area, high spatiotemporal resolution data on vertical surface deformation remains insufficient. Thanks to the high observation frequency and moderate spatial resolution of Sentinel-1 satellite series, combined with well-developed multi-temporal InSAR (MTI) analysis techniques, this limitation has gradually been addressed. This study applied Persistent Scatterer InSAR (PSI) to analyze Sentinel-1 satellite data from 2016 to 2021, obtaining LOS displacement information from two orbits in the CRFP. Temporally, the study reduced disturbances in the LOS time series and constructed synchronized LOS observation data for both tracks at the same time intervals. Spatially, a 200-meter averaging grid was constructed to resolve PSI points mismatch issues from both orbits. The results were ultimately resolved into two-dimensional (E-W and U-D) time-series displacement components. By performing k-means clustering on the time-series vertical displacement data, the land subsidence characteristics of the study area were categorized into four groups: severe subsidence, moderate subsidence, mild subsidence, and normal condition. These clustering results can aid governmental agencies in drafting groundwater usage regulations. In the future, this study will integrate borehole data and groundwater level information to infer hydrogeological parameters and explore the spatial variability of groundwater volumes and geological materials.