ABSTRACT
Landslide susceptibility is the likelihood of a landslide occurrence in an area, assessed on the basis of local terrain conditions and represents the degree to which an area can be affected by future slope failure. Landslide Susceptibility Maps (LSM) are essential for land-use strategies as tools to support land management decision making. The production of a LSM can be obtained by using different methods, often through Geographic Information Systems (GIS). In this work, available data (i.e. geological and land cover maps, DEM-derived products and the pre-existing landslide inventory) were used to produce the landslide susceptibility map of the Messina Province (Italy), after the emergency occurred in 2009–2010. A simple implementation of Random Forest developed in Matlab was used to produce the LSM by performing multivariate classification. Results highlight most of the Province is susceptible to landslides, especially the Nebrodi Mountains, where detailed analysis have been planned to perform mitigation actions.
