An Adaptive Statistical Multi-grid DInSAR Technique for Studying Multi-scale Earth Surface Deformation Phenomena

In this study, we show the potential of an adaptive quad-tree-based decomposition method applied to Differential Synthetic Aperture Radar (DInSAR) data. Specifically, the proposed method exploits a multi-resolution scheme for the phase unwrapping of sequences of DInSAR interferograms and allows to produce DInSAR deformation products at different scales of resolution. The selection of the used multi-grid is based on the analysis of the statistical properties of a sequence of interferometric phase, allowing to recognizing major deformation areas where phase unwrapping operations can be performed more efficiently, with a computational improvement and without losing significant information.