A Novel Method for Extracting Mining-Induced Ground Deformation Using InSAR and the Weibull Model

Underground mining of natural resources disrupts the original stress balance of the surrounding rock masses, causing deformation in overlying rock layers and the ground surface. These disturbances may cause various geohazards, such as collapses, landslides, structural damage to buildings and infrastructure, and ecological degradation. Therefore, it is crucial to accurately extract and predict mining-induced ground deformation to assess and prevent mining-related geohazards. Interferometric synthetic aperture radar (InSAR) has been widely applied to monitor mining-induced deformation. However, due to the rapid rates and high spatial gradients of the mining-induced deformation, as well as rapid changes in ground topography, it is difficult to extract accurate and continuous deformation measurements using InSAR. To this end, this study proposed a novel method for extracting mining-induced deformation based on the InSAR and Weibull model.

The core concept behind the proposed method is to link time-interval InSAR-derived deformation using a time-series model, enabling the extraction and prediction of mining-induced deformation. Specifically, the method for connecting the deformation of line-of-sight (LOS) is first established based on the Weibull model. The initial model parameters are then derived using the genetic algorithm-particle swarm optimization (GA-PSO) approach. These parameters are subsequently optimized according to their spatial distribution characteristics. Finally, the trust-region reflective least squares (TRRLS) algorithm is applied to determine the final model parameters, enabling the extraction of mining-induced deformation during the monitoring period. The results indicate that the extracted deformation is accurate and consistent overall, with root mean square errors (RMSE) of approximately 9.8mm and 14.1mm observed for the simulation and field experiments, respectively. Furthermore, leveling data are also used to validate the accuracy of the proposed method, yielding an RMSE of 32.6mm. Additionally, the relationships between the Weibull model parameters, ground subsidence values, and initial subsidence time are analyzed. The effects of various factors—estimation algorithms, number of observations, time intervals, and monitoring errors—on the proposed method are examined. These results suggest that the proposed algorithm can be a practical and cost-effective tool for extracting mining-induced displacements and assessing and mitigating mining-related geohazards.

This work has been supported in part by the National Natural Science Foundation of China under Grant 52474184 and Grant 42474018, in part by China Postdoctoral Science Foundation under Grant 2023T160685 and Grant 2020M671646, in part by Young Elite Scientists Sponsorship Program by CAST under Grant 2023QNRC001-YESS20230599, in part by the National Key R&D Program of China under Grant 2022YFE0102600, in part by supported by the Construction Program of Space-Air-Ground-Well Cooperative Awareness Spatial Information Project under Grant B20046, in part by the China Scholarship Council under Grant 202406420081, in part by the Spanish Agencia Estatal de Investigacion under Grant G2HOTSPOTS (PID2021-122142OB-I00), and in part by the AEI, Ministerio de Ciencia, Innovación y Universidades. Convocatoria Proyectos en Colaboración Público Privada, 2021, under Grant CPP2021-009072 (STONE), and Defsour-PLUS (PDC2022-133304-I00) from the MCIN/AEI/10.13039/501100011033/FEDER, UE with funds from NextGenerationEU/PRTR.