Rockfall susceptibility along communication routes and in urban areas in Italy based on physical and statistical models

Landslide susceptibility, the likelihood of landslide occurrence in different locations of a specific area, can be obtained with many different methods. Here, we aim at comparing two state-of-the art susceptibility zonations obtained within two such methods, and their predictions at the location of different infrastructure in the whole of Italy. The study considers rockfall landslide type, and susceptibility zonations obtained by a grid-based three-dimensional physically based models and one slope-unit based statistically based model. We first compare the two susceptibility zonations on the whole of Italy, reducing model dependencies by aggregating results of the physical model on the same slope unit zonation as the statistical method. Next, we consider the predictions of both maps restricted to urban areas, railways, and road network. The study shows that the predictions of the two maps are drastically different. We discuss the reasons for such differences, and interpretation of the two versions of rockfalls susceptibility map.