Key points are not available for this paper at this time.
Italy faces significant vulnerability to landslides, necessitating reliable forecasting models for effective property and population protection. These models must not only guarantee high accuracy but also facilitate easy integration into early warning systems for civil protection.Physically based landslide forecasting models meticulously replicate the triggering mechanism of shallow landslides. These models employ numerous input parameters interconnected through complex mathematical relationships to assess the probability of landslide occurrences. Despite their precision, these techniques encounter challenges in spatializing geotechnical and hydrogeological parameters across extensive areas, restricting their application to slope-scale assessments. Additionally, the output of these models, presented as probability maps, lacks immediate utility for civil protection purposes, where a risk definition would be more operationally advantageous.This study aims to address this gap by analyzing the optimal criterion for spatializing input data of physical models for regional-scale application. The goal is to develop a procedure that transforms model outcomes into readily usable risk scenarios. The study focuses on the Metropolitan City of Florence, leveraging a richly populated database of geotechnical and hydrogeological parameters. The selected model, HIRESSS (High-Resolution Slope Stability Simulator), simulates events occurring from January to March 2016, encompassing eight reported landslide events.Through p-value analysis derived from statistical hypothesis testing, the study explores two criteria for parameterizing geotechnical and hydrological variables: a lithological criterion and one based on pedological-landscape units. This dual approach aims to consider both the lithological origin of soils and the impact of surface erosive processes on the spatial variability of input parameters. The study employs an innovative GIS-based procedure, integrating field surveys and morphometric parameters, to connect landslide probability maps with vulnerability and elements at risk, ultimately determining a risk scenario for the catchment area of the Cesto stream (southeast of Florence).The analysis highlights the mixed criterion as the most supported spatialization approach, incorporating lithological factors for cohesion and friction angle and pedological-landscape criteria for hydraulic conductivity, soil unit weight, and porosity. Back-analysis validation reaffirms the model's high predictive capability with the adopted mixed-criterial parametrization. The results align with our understanding of landslide triggering mechanisms, particularly sensitive to cohesion and slope gradient.The study concludes with a GIS-based risk analysis, providing impact scenarios for identified exposed elements. This final product proves instrumental for both prevention and emergency management. Once calibrated, the developed procedure holds potential for automation and replication in other study areas, offering a scalable solution for landslide risk assessment and mitigation.
Morreale et al. (Fri,) studied this question.