Atmospheric perturbation detection

Person in charge : Cécile Lasserre, LGIT-UJF

Nicolas Meger, LISTIC-UdS

Separating the displacement signal from atmospheric perturbations in SAR interferometric data is a rather difficult task, particularly when only a few differential interferograms are available. Though various approaches have been developed so far, this problem still remains an open question. In this Work-package, we will focus on two main directions:

1. Expert corrections: An essential task is to implement "expert" atmospheric corrections in the PS (WP3) or NSBAS (WP4) techniques. These corrections will rely on a systematic analysis of the phase-topography dependency and a better characterization of the atmospheric conditions at the time of SAR data acquisitions.
2. Data mining: We also propose to develop a different approach, based on data mining techniques, for identifying ground deformation zones. Such an approach is well suited for handling the huge amount of data gathered in times series of interferograms over large areas.

This study is carried out on the Lake Mead subsidence through a collaboration between LISTIC-UdS where Nicolas Meger has been starting to develop specific methods for knowledge extraction and change detection in remote sensing image time series and LG-ENS where Marie Pierre Doin and Cécile Lasserre who have a great experience in gathering and analysing large interferogram series to measure small deformations. This Work-package benefits from the support of a PhD student (Andreea Julea) supervised by Nicolas Méger and Philippe Bolon in the LISTIC-UdS team, and a post-doctorant (Stéphane Guillaso) in the LG-ENS team.