Abstract
The South-to-North Water Diversion Middle Route Project (SNWD-MRP) traverses two major agricultural production areas and populous regions in northern China, from Henan to Beijing-Tianjin-Hebei (BTH). It is of profound significance in alleviating water resource scarcity and improving the local ecological environment in these areas. This paper introduces a wide-area subsidence monitoring technology that combines the FS-InSAR method with ground measurement data correction. Based on this, a correlation analysis is conducted using measured groundwater level data along the route, exploring the intrinsic link between groundwater level fluctuations and uneven subsidence dynamics. Finally, a machine learning algorithm is employed to quantify the impact of groundwater changes on ground subsidence. The results indicate a high correlation between subsidence and groundwater level fluctuations (R: 0.6 ~ 0.9). The machine learning training results show that changes in groundwater storage are the main factor affecting subsidence, accounting for 75.1%. Further analyzing the reasons, groundwater over-exploitation triggers a drop in the groundwater level, which leads to a decrease in the pore pressure and an increase in the effective stress in the sand layer, and secondary consolidation and subsidence in the clay layer due to the drainage lag, whereas SNWD-MRP contributes to the rebound of the groundwater level through the reduction of groundwater extraction and ecological water recharge, which reduces the pressure release from the aquifer, and leads to the stabilization of the compression process of the strata. This study indicates that groundwater losses remain the primary controlling factor of subsidence. After the SNWD's initiation, subsidence in the Beijing and Tianjin areas has noticeably slowed and stabilized. Overall, this major water project has effectively mitigated uneven subsidence, underscoring its positive role in regional ground stabilization and environmental improvement.