Andrea Tamburini, Tele-Rilevamento Europa TRE S.r.l. (Italy)
Chiara Giannico, Tele-Rilevamento Europa TRE S.r.l. (Italy)
Fabrizio Novali, Tele-Rilevamento Europa TRE S.r.l. (Italy)
Davide Carlo Guido Martelli, IMAGEO S.r.l. (Italy)
Massimo Broccolato, Regione Autonoma Valle d'Aosta (Italy)
Thanks to the availability of satellite data archives covering more than one decade, Permanent Scatterer SAR Interferometry (PSInSAR®) represents nowadays one of the most powerful techniques capable of retrieving surface displacements. Mapping landslide distribution at regional scale is traditionally based on geomorphological analysis, both from aerial-photo interpretation and field surveys. Nevertheless, where displacement rate is very low (millimeters to centimeters per year), assessing the activity of a landslide is generally difficult or even impossible without the help of long-term displacement data. For these reasons, thanks to its capability to detect small displacements over long periods and large areas, PSInSAR® analysis can be considered complementary to conventional geological and geomorphological studies in performing landslides inventories at regional scale.
The integration between PSInSAR® analysis results and traditional ground based monitoring data allows for interpreting the behavior of landslides over long periods and helps in understanding the impact of flood events at regional scale on their evolution. The example of the flood which affected the northwestern Po basin and particularly the Valle d'Aosta Region in October 2000 will be presented. The reactivation of some major landslides during and immediately after the event spurred the Regional Government to install ground based monitoring networks, which started acquiring data since the beginning of 2001. Nevertheless, the lack of data before the meteorological event made it impossible to understand the consequences of the reactivation on the evolution of the studied phenomena. Tanks to the availability of displacement records provided by the PSInSAR® analysis, referred to a period of ten years just before the October 2000 event, it was possible to integrate the knowledge of the studied phenomena and link the displacement data series before and after the reactivation.