Work Package I

Recursive and robust displacement estimation from SAR image time series

The main objective of this work package (WP) is to develop recursive and robust multi-temporal InSAR approaches for displacement estimation from SAR image time series. The interferometric phase estimation is formulated as a covariance matrix estimation problem. This work package is based on two Ph.D. theses. In the Ph.D. thesis of Hoa Viet Phan Vu (10/2020–12/2023), we first proposed a new phase linking approach based on the Maximum Likelihood Estimator (MLE), and then extended this work to a general covariance fitting framework integrating SAR data modeling (complex circular Gaussian, generalized Gaussian), regularization (low rank, shrinkage, banding) on the covariance matrix structure and optimization algorithm (Majorization-Minimization, Riemannian gradient descent). In the Ph.D. thesis of Dana El Hajjar (12/2022– 12/2025), sequential algorithms are being developed, based on MLE-Phase Linking and covariance fitting, to efficiently integrate new images (either one by one or block by block).