Landslides are an increasingly significant concern in a world of increasing climate volatility, and there have been a number of efforts to improve the predictive technology around using different methods to improve landslide monitoring techniques.

Landslides often happen without clear warning. Consequences are catastrophic in terms of human losses. Governments are therefore interested in collaborating with researchers to detect landslides and mitigate their effects. So far, methodological investigations have been largely focused on using labor-intensive preprocessing techniques on landslides or risky territories.

The proposal aims to create a system to classify landslides, already occurred, and capable of detecting future landslides.

A research work is in progress and initial results have been published and presented at the recent IGARSS Conference held in Yokohama [1].

Next step in the project development is to use besides Sentinel-2, also Sentinel-1 data better suited for monitoring land movements. They would allow also to overcome the problem related to cloud coverage, proved to be a significant problem for Sentinel-2 data collection. Moreover, a system trained on SAR data or on a combination of SAR and optical data, might produce highly accurate results. The problem is that Sentinel-1 data have resulted not available on the period of interest, because of the USA shut down. For future works Sentinel-1 data will be provided by the ESA centre for Earth observation (ESRIN) in Frascati (Rome). Another aspect to explore is the possibility to use the software for the classification of different types of past landslides. Landslide classification is in fact primarily based on different types of movement (falls, topples, slides, spreads, flows) and material (rock, soil, mud and debris). Different sizes of landslides will also be taken into account. The project may include visits in situ.

The results of the research will be disseminated in conferences and/or international journals.

The student/s will be jointly supervised by Prof. Stefania Sica (https://www.unisannio.it/user/631/contatti), with great expertise in geotechnical engineering, by Prof. Silvia Ullo (https://www.unisannio.it/user/622/contatti), working on satellite data analysis since a long time, and by Dr. Maria Pia Del Rosso and Dr. Alessandro Sebastianelli, graduated at University of Sannio, working at phi-lab of ESA-Esrin on machine learning applied to earth observation and remote sensing.