GIS-based predictive models for regional-scale landslide susceptibility assessment and risk mapping along road corridors

In this paper, a study aimed to assess the landslide susceptibility at a regional scale for the wide provincial territory of Matera (Basilicata region, southern Italy) and the relative risk along the main road corridors distributed in this area is presented. A heuristic-bivariate statistical predictive model was performed to assess and map the landslide susceptibility in the study area by using a polynomial function of eight predisposing factors, weighted according to their influence on the instability process. The resulting susceptibility map was successively used for assessing the landslide risk along the provincial road network. The importance of these roads, representing the main network connecting the urban centres, derives from the absence of an efficient integrated transportation system through the entire regional territory. The landslide risk was evaluated through a matricial approach, which has allowed to define the risk levels (low, medium and high) along road stretches by overlapping the consequences and hazard maps, by combining their corresponding classes in a matrix and by associating to each combination a risk level. The resulting landslide risk map provides support information for decision-making and for identifying the priorities for the design of appropriate mitigation plans.