Integrating sediment connectivity into the assessment of landslides susceptibility for road network

Landslides provoke significant direct and indirect economic losses to infrastructures, in particular along road networks. Thus, it is fundamental identifying the route sectors that could be affected by landslides, in order to reduce the risk level for the population and the economic cost of road damaging. Moreover, several researches conducted in different contexts stressed that the exposure of road networks to slope instabilities could increase because of ongoing climate change and as a consequence of growing economy in several countries. For these reasons, the present work aims to develop and test a data-driven model, based on Genetic Algorithm Method (GAM), for the identification of the sectors road network sectors that are susceptible to be affected by landslides triggered upstream the infrastructure. This work quantifies, also, the impact of sediment connectivity on the susceptibility evaluation in the case studies. The study area corresponds to the north-eastern area of Oltrepò Pavese (northern Italy), a zone very prone to shallow landslides causing severe damages to the road networks. This work shows that the effectiveness of the model in the identification of the most susceptible routes increases including sediment connectivity in the predisposing factors. This parameter, indeed, characterizes runout and the travel distance of a slope instability, improving the ability in identifying the road sectors hit by landslides. The modeled susceptible roads are, then, mapped correctly by the methodology, furnishing an important tool for land use planning and for implementing tools able to reduce the risk for the infrastructures.