Robust satellite techniques for early detection of hydrological stress in cultivated areas: the case of Basilicata (Italy) region

The amount and continuity of water resources has a crucial role on vegetation growth and is every day more affected by climatological changes. From this point of view, continuous monitoring of vegetation conditions can help not just to timely identify climatological forcing but also to adjust water management practices in cultivated areas. To this aim global observations from satellite Earth Observing systems - and particularly (since the 80s) from LANDSAT missions and (since 2015) from Sentinel-2 constellation which significantly improved temporal (up to 5 days) and spatial (up to 10 meters) resolutions – play a major role. To timely identify those variations possibly related to medium-long term climatological forcing and to distinguish them from the occasional, short-term, ones however affecting cultivated areas management, long-term analyses of satellite observations are required. In this paper the general change detection approach, named Robust Satellite Techniques (RST) is applied which identifies “anomalous” spatial-temporal patterns on the base of a preliminary computation of its behavior (in terms of expected value and variability) as historically observed at each location in similar (same period of the year, same time of observation, etc.) observational conditions. To this aim long-term time-series of LANDSAT (for investigating possible climatological forcing) and Sentinel-2 (for short-term analyses) satellite data have been analyzed to investigate water stress conditions in cultivated areas of Basilicata Region. Preliminarily results achieved in the framework of the OD4SA (On Demand Services For Smart Agriculture) Project will be presented.