A Novel PF-Based Method for Height Reconstruction in Distributed Geosynchronous Repeat-Pass InSAR

With the advantages of high spatial resolution, short repeat-pass cycle, and large observation area in Ka-band distributed geosynchronous (GEO) synthetic aperture radar (SAR) systems, its interferometry (InSAR) measurement can fast retrieve high-resolution and high-accuracy digital elevation model (DEM). However, compared to low-orbit systems, it is more difficult and costly for distributed GEO SAR systems to perform tight formation flying at such a high orbit altitude, and, therefore, atmospheric effects, which bring nonstationary and non-Gaussian phase errors, should be taken into account in its repeat-pass InSAR. To address the problems earlier, a spatial–temporal joint particle filter-based method (ST-PF) for DEM generation by distributed GEO InSAR is proposed in this article. The proposed ST-PF method is validated under several stationary and nonstationary atmospheric conditions through simulation experiments, and high-accuracy and high-resolution DEMs are obtained. Moreover, the ST-PF method can withstand severe nonlinearity in interferometric phases, which results from a relatively small ambiguity height in the Ka-band. It is also tested that the mean square error (MSE) of the retrieved DEM is consistent with the posterior Cramer–Rao bound (pCRB) of the estimation problem, showing the validity and the accuracy of the proposed ST-PF method. With various processing parameters, errors, and scene types tested, the method shows good robustness in different conditions.