Mappatura dei trend di suscettibilità all’erosione post-incendio attraverso una metodologia elaborata in ambiente GIS (Val di Susa, NW Piemonte, Italia)

In autumn 2017, an intense fire season affected Piemonte Region (north-western Italy), burning 9.700 ha after nine major wildfires. The biggest fire occurred in Susa Valley (nearly 4.000 ha), leading to heterogeneous damages to vegetation and soils and therefore to heterogeneous post-fire geomorphic processes in concomitance with intense rainfalls, including widespread sheetflood erosion, floods, hyperconcentrated flows and formation of mud/debris-flows. These processes led to considerable damage to human settlements especially in June 2018 and soon after these events some geomorphological risk-mitigation treatments were planned. Soil erosion and runoff constitute the main triggering factors for geomorphic processes on the slopes and they depend on wildfire effects, mainly referred to as "fire severity", and on the occurrence of intense precipitation. In the current context of changing climate, both wildfires and extreme precipitations are increasing in frequency and post-fire geomorphic processes are becoming a major issue, therefore the study on the topic is crucial in finding new strategies for risk mitigation. This Thesis inserts itself well in this context, as it aims to discuss a model for the long-term evaluation of post-fire geomorphological hazards and related risks. The work was conducted through a GIS-based methodology, analyzing the evolution of fire severity effects (referred to as "burn severity") and soil erosion susceptibility trends during the 4-years period following the fire in Susa Valley, namely from 2018 to 2021. The main phases of the work were: 1. production of multitemporal burn severity maps, based on a selected spectral index performed on multispectral satellite images; 2. assessment of multitemporal erosion models, by adding corrective factors for fire severity to the USA RUSLE model; 3. discussion of the results, downscaling at the watershed scale, to locate the most critical sectors for post-fire erosion; 4. study of the potential effectiveness of a specific post-fire erosion mitigation treatment (contour-felled log erosion barriers), simulating its modifications on soil loss rates over a selected critical area and determining the optimal timing for its application. The elaborated methodology was useful for an expeditious assessment of soil erosion rates in post-fire conditions, indicating a strong relationship with burn severity distribution. The most critical sectors were in the western parts of the forested areas, where high values registered for both burn severity and erosion in 2018 locally persisted over time until 2021, indicating a slow recovery process. The methodology shows some limitations, first of all the lack of field validation and assessment of runoff. However, the results obtained highlight that the knowledge of landscape heterogeneity is crucial in assessing post-fire soil erosion trends in different contexts, allowing to plan efficient fire management strategies.