An Adaptive High-Accuracy Terrain Phase Error Compensation Algorithm in GNSS-Based InBSAR via Multisatellite Collaborative Observations

The Global Navigation Satellite System (GNSS)-based Interferometric Bistatic Synthetic Aperture Radar (InBSAR) systems can retrieve 3-D deformation fields. However, the terrain phase error significantly deteriorates their accuracy. In this article, an adaptive high-accuracy terrain phase error compensation algorithm is proposed. First, based on the inherent simultaneous availability of multiple satellite signals, a multiangle SAR image association algorithm is proposed to obtain the observations of the same point from different angles, and image shift under different bistatic configurations is also eliminated here. Second, incorporating a nonrobust data extraction and discarding mechanism, an improved estimator is proposed, along with modified loss and weight functions, to robustly estimate terrain errors of persistent scatterers. Third, since position dilution of precision (PDOP) can quantify the impact of multistatic configurations on 3-D deformation retrieval accuracy, a filter that considers both the PDOP and distances between the persistent scatterers named PDOP-Distance-weighted spatial filter is proposed to maximize the utilization of multisatellite collaborative observations and obtain the terrain errors of whole scene. The experimental data of five BeiDou satellites verify the proposed algorithm. The deformation measurements and terrain error estimations from GNSS-based InBSAR are compared with classical differential GNSS and autonomous aerial vehicles (AAV)-mounted light detection and ranging (LiDAR), both showing good consistency. These results prove the reliability of GNSS-InBSAR systems and indicate their great potential for deformation monitoring.