WP3

Wide-area and site-specific DInSAR processing will be perfomed to generate interferograms at different temporal baselines. Such processing will be carried out by UNIBO and UNIMIB using both commercial software and open-access tools (e.g. SNAP, GMTSAR). The two research units will focusing on wide areas of interest in the Alps and Apennines.

Interferograms will be generated with varying temporal baseline, from 6 days to 1 year. This will allow capturing signals related to different mass movement processes over different timescales. 

Inspection of interferograms will allow selecting the final dataset, also accounting  for spatial baseline, coherence and environmental conditions (e.g. snow cover). The interpreted displacement signals from the interferograms will make up an extensive feature library to train and validate a deep learning model (WP5).

When possible, phase signals will be unwrapped and multi-temporal displacement maps and time series will be extracted and constrained using ground-based data for a better understanding of representative case studies. 

Example of a 12-days interferogram (left) and 1-year interferogram (right) in descending geometry, generated by the software SARscape. Notice how signals change over rock glaciers (thin black lines) and deep-seated landslides (thick black lines). *Credits: UNIMIB.