Abstract
Ground subsidence poses a significant threat to the structural stability and operational safety of railway systems, highlighting the critical need for effective monitoring throughout the construction lifecycle. Subsidence monitoring data along railway alignments can be leveraged to analyze subsidence characteristics across different sections and to track time-series deformation at specific locations. This study proposes a comprehensive framework for the spatio-temporal analysis of subsidence along railway lines, grounded in the Small Baseline Subset Interferometric Synthetic Aperture Radar (SBAS-InSAR) technique. To this end, 39 Sentinel-1 A SAR images were processed to derive subsidence data spanning the pre-construction, construction, and post-construction phases of the Lai-Rong Railway, with results validated against ground leveling measurements. Hotspot analysis was employed to identify subsidence clustering zones, while time-series deformation, average cumulative subsidence, and correlation coefficients were utilized to examine spatio-temporal patterns at targeted locations. The findings demonstrate that SBAS-InSAR achieves millimeter-level accuracy in subsidence monitoring, revealing an uneven spatial distribution along the railway. Notably, the DK30 ~ DK50 section exhibited substantially higher subsidence rates, primarily attributed to large-scale crustal deformation. Construction activities, such as ground excavation, leveling, and load introduction, emerged as primary drivers of subsidence, particularly in surface fill soils. To mitigate long-term subsidence risks, thorough compaction of fill layers is recommended, along with increased monitoring frequency in high-risk sections to enhance safety. This study offers valuable insights into the application of SBAS-InSAR for ground subsidence monitoring across the full construction cycle of large-scale railway projects.