Multi-source data-based quantitative risk analysis of road networks to slow-moving landslides

The paper addresses the quantitative risk analysis for a state road crossing an area of southern Italian Apennines diffusively affected by slow-moving landslides. In this area, Palaeozoic gneissic rocks suffering from intense weathering processes, which produce complex and deep weathering profiles, are present, and this determines severe predisposing conditions to deep-seated slow-moving landslides. Although not directly threatening human lives, for years these slope instabilities have been causing damage and temporary traffic interruptions or limitations to many road sections. To pursue a sustainable landslide risk management, a method that fully exploits multi-disciplinary data consisting of geological-geomorphological features, geotechnical characterization of geomaterials, both conventional (i.e. GPS and inclinometer) and remote sensing (i.e. MT-DInSAR) displacement measurements, in-situ and virtual (i.e. Google Street View images) surveys, and probabilistic tools (i.e. fragility and vulnerability curves) is implemented. As a novelty, such a rich dataset allows overcoming some limitations of the (few) previous studies in the scientific literature on the analysis of the risk posed by slow-moving landslides to roads by exploiting i) the multi-temporal recording of the road damage to catch the response of the infrastructure (i.e. both the road pavement and the side retaining structures) with time, and ii) the association of the cumulative landslide-induced displacements with the corresponding damage in order to feed empirical forecasting tools for consequence analyses. The thorough knowledge of the slow-moving landslides and their interaction with the exposed roads are implemented within the proposed method to assess the direct economic losses in terms of repair costs, should no countermeasures or mitigation works be implemented in due time. Considering that the studied area resembles very typical conditions of inner roads in hilly and mountain areas of southern Italy, the method can represent a valuable tool for decision makers to prioritize money allocation for risk adaptation and mitigation actions for roads in similar geo-environmental contexts. • A multi-source data-based method is proposed for landslide risk analysis of roads. • Multi-temporal damage data depict the road response with time. • Landslide-induced displacement and damage data are merged for consequence forecast. • Direct economic losses (road repair costs) are assessed for a fixed time scenario. • The method allows decision makers to prioritize risk mitigation actions for roads.