A Robust High-Frequency InSAR Framework for Mining Subsidence Monitoring via Meteorological a Priori Optimization and MultiTrack Fusion

Abstract

Surface subsidence induced by mineral extraction poses severe threats to infrastructure and ecological security in mining regions, thereby necessitating high-frequency, precise, and dynamic monitoring. To address the challenges of decorrelation and lagged dynamic capture inherently associated with conventional time-series InSAR in mining areas, this study proposes a robust, high-frequency monitoring framework that integrates meteorological a priori optimization with multitrack spatiotemporal fusion. By employing meteorological priors, SAR acquisitions contaminated by heavy precipitation are actively excluded to intercept the propagation of phase errors. Subsequently, high-density, continuous deformation time series are reconstructed via multitrack geometric projection and temporal fusion. Application in the Xinxiang mining area, Henan Province, demonstrates that compared to ground leveling truth, the proposed method reduces the root mean square error (RMSE) of the annual deformation rate to 5.5 mm/year, marking a progressive error reduction of 35.3% over the conventional full-dataset fusion baseline (8.5 mm/year) and an additional 22.5% accuracy leap over geometry-weighted multitrack fusion approaches (7.1 mm/year). Furthermore, the accuracy of cumulative time-series deformation is improved by 51% (with the temporal axis densified up to a 5-day resolution). This framework accurately reconstructs the nonlinear collapse dynamics at the margins of subsidence basins and elucidates the hydrological response mechanisms underlying data voids. Ultimately, it provides reliable data support for mining subsidence monitoring and disaster early warning under complex climatic conditions.