ASSESSING SHALLOW LANDSLIDE SUSCEPTIBILITY OF VEGETATED HILLSLOPES THROUGH A PHYSICALLY-BASED SPATIALLY-DISTRIBUTED MODEL

Slope instabilities are a serious threat to human activities, settlements, and safety worldwide. Among the different types of slope movement, shallow landslides are the most common phenomena and are often associated to other soil instabilities and to various channel processes (i.e. sediment transport, woody debris). Vegetation and in particular forests is an effective and well-known tool in preventing and mitigating hydrogeological hazards, mainly through the effects of the reinforcement exerted by the root systems. Root reinforcement, then, is a factor that should be included in hazard estimation and the resulting maps that represent a fundamental tool for planning and managing the hydrogeological hazards. Accordingly, in the last two decades, a wide number of different methods and approaches have been proposed to produce landslide hazard maps, with particular reference to Physically-Based Spatially-Distribute Models, PBSDM. However, including root reinforcement is still a challenge for the scientific community due to the huge spatial and temporal variability and the difficulties in incorporating into slope stability analysis. The main gaps to be filled can be summarized as follows:  the knowledge on the spatial distribution of the soil reinforcement due to the root systems have to be improved and linked to the stand forest characteristics;  a 3-D probabilistic PBSDM of hillslope failure able to include in a comprehensive but simple manner the presence of the forest vegetation have to be developed;  the use of information at coarse spatial resolution, which introduces an additional source of uncertainty has to be properly managed. This study gives a brief review of the role of forests against natural hazards and on the state of the art concerning the implementation of root reinforcement into stability models. Thereafter, it attempts to fill such gaps improving the knowledge about modelling and quantifying the effects of vegetation on slope stability. The main outcome is the development of a 3-D probabilistic PBSDM of hillslope failure, based on geotechnical sound hypothesis and stochastic approach through the Monte Carlo Simulation (MCS) analysis. Such a model is able to manage the uncertainty of model parameters and is a reliable way to deal with the problem of a lack, or a poor knowledge of terrain characteristics over large study areas. In addition, it allows evaluating the effects of silvicultural operations, to estimate the woody material, recruitable from the hillslopes in small mountainous catchments, and to quantify the additional soil reinforcement provided by some cultivations such as the grapevine. Moreover, a series of field experiments on the rooted-soil under compression is presented in order to investigate the hydro-mechanical process that occurs during the triggering mechanisms of shallow landslides. Finally, the proposed modelling framework will allow:  to assess the probability of hillslopes failure considering the characteristic of the vegetation and to provide more reliable shallow landslide hazard maps at catchment scale;  to improve the efficiency of prevention and protection due to vegetation, and particularly to the forests, against natural hazards evaluating different land management strategies;  to support the planning of eventual forest interventions or soil-bio engineering works identifying the areas affected by high landslide susceptibility.