Joint Probability Integral Method and TCPInSAR for Monitoring Mining Time-Series Deformation

Abstract

Because of the high vegetation coverage, fast deformation in certain mine areas, some SAR interferograms are seriously incoherent. When using time-series synthetic aperture radar interferometry (InSAR) to monitor the surface movement basin of the mining area, there may be a certain period of missing deformation information, making the obtained surface time-series deformation incomplete. To this end, this paper proposes a way of using the results predicted by probability integral method (PIM) to replacing the monitoring results that cannot be obtained because of the seriously incoherent SAR interferograms; then, the monitoring results of the high-coherence SAR interferograms and the results predicted by PIM are used by the improved temporarily coherent point SAR interferometry (TCPInSAR) to invert the deformation, thereby obtaining a complete mining time-series deformation. The TCPInSAR using a linear model does not reflect the complex deformation characteristics of the mining area. So this paper focus on the characteristics of deformation of study area, the original linear model is changed to a polynomial model, which improves the applicability of TCPInSAR to monitoring mine deformation. Comparison between the experimental results and levelling shows that the root mean square error (RMSE) and the maximum deviation (MD) of the results obtained by combining the PIM with the improved TCPInSAR are 14.2 mm and 43.0 mm, respectively. Compared with the results obtained by combining the PIM with the TCPInSAR (RMSE = 16.2 mm, MD = 57.5 mm) and the results of using only the TCPInSAR (RMSE = 26.5 mm, MD = 88.4 mm), the monitoring accuracy is increased by 12.3% and 46.4%, respectively.