Extreme flood significantly altered urban ground deformation patterns: multi-source remote sensing evidence from the 2020 Chongqing flood

Rapid urbanization and extreme weather events exert significant pressure on urban geological environments, yet the impact of extreme floods on ground deformation, particularly the interplay between underground engineering, building density, and surface response, remains unclear. Using Enhanced Small Baseline Subset (E-SBAS) InSAR technology and machine learning methods, this study analysed the impact of the 2020 Chongqing flood on ground deformation patterns. Results shifted from uniform subsidence (−5.19 ± 8.66 mm/year) pre-flood to significant spatial differentiation (−0.16 ± 125.46 mm/year) post-flood. Areas within 250 m of subway lines shifted from subsidence (−25 to −5 mm/year) to uplift (0 to 15 mm/year), with high-risk zones (35%) exceeding the study area’s average (15%). Machine learning revealed land cover type (18.25%) and subway proximity (14.45%) to be key factors affecting flood susceptibility. These findings emphasize the importance of coordinated surface and underground development in urban disaster prevention.