A method for GB-InSAR temporal analysis considering the atmospheric correlation in time series

摘要

GB-InSAR, with high time-spatial resolution and high accuracy, shows great potential in landslide monitoring. However, the accuracy of GB-InSAR is usually reduced by the atmospheric disturbance and temporal decorrelation. PS-InSAR technology can solve those problems well and get accurate deformation information, so it has been widely adopted in space-borne SAR. For the atmospheric correction, PS-InSAR assumes that the atmospheric phase is only strongly correlated in the space domain. But for GB-InSAR, atmospheric phase shows correlation in both the time and space domains, because of the short time interval. Therefore, the calculation of linear velocity will be affected by the atmospheric disturbance when the PS-InSAR technology is applied to the GB-SAR data. To solve this problem, a PS-InSAR strategy for GB-SAR data considering the atmospheric disturbance in the time domain is proposed. The proposed method uses the differential interferograms interfered by nearby SLCs to ensure the high coherence of interferometric phase and reduces the impact of atmospheric disturbance. The coherence is used for PS selection besides the amplitude deviation index, which can increase the density of PS points. Furthermore, a method for atmospheric correction based on the wrapped phase is presented. Thus, the linear velocity is computed based on the interferogram without most atmospheric disturbance. The validation using the data of the open pit in Malanzhuang, Hebei, China, shows that the proposed method can get a deformation monitoring accuracy of sub-millimeter.