Wildfire risk assessment and management: insights, challenges and future perspectives for Sardinia, Italy

The large and severe wildfires that affected Mediterranean ecosystems over the past decades have led to extensive research and application of decision support tools for operational purposes. Like the rest of the Mediterranean region, Sardinia (Italy) has experienced substantial long- and short-term variation in fire regime and behavior. Overall, the island has seen marked reduction in area burned since the 1980s, with a sharp transition in the mid-1990s. In recent years Sardinia experiences on average about 2,500 fires annually that burn about 17,000 ha. Despite the downward trend of fire occurrence and area burned, localized areas and seasons of high wildfire risk still persist. Furthermore, recurring large wildfires still cause substantial financial losses, accidents and fatalities, and costs related to large wildfire suppression usually cover the most of the wildfire budgets. For example, about 7% of fires are responsible for 85% of the overall area burned, with the largest fires concentrated in a few extreme-weather days. As part of efforts to improve risk estimation and prevention efforts, we have calibrated and implemented a number of simulation models and applied them to better understand spatiotemporal patterns in fire behavior and risk. The models and tools are also being used to evaluate fuel treatment and other mitigation strategies. A key part of this work is the application of a compact fire simulation algorithm (minimum travel time, MTT) that makes computationally feasible to simulate thousands of fires and generate burn probability and intensity maps over large areas. The fire simulation outputs provide a way to study wildfire topology on complex landscapes under variable weather, and analyze uncertainty associated with wildfire events in terms of timing, location, intensity, and duration. We also coupled wildfire simulation models with geospatial interfaces to streamline preparation of inputs and post-process outputs. These latter tools also include spatial optimization software to design landscape fuel treatment projects. The result of the work is a substantial increase in landscape scale information to support wildland fire mitigation efforts. In addition, wildfire simulation outputs can provide data and information for a range of wildfire-related issues including hydrological impacts, habitat conservation, carbon offsets, WUI protection, and preservation of biodiversity. The models are also being used at regional scales to supporting large-scale climate and land-use change analyses as well as wildfire policy, management and mitigation issues. This work was partially founded by the Sardinia Region under the Regional Law n. 7/2007 (EXTREME Project - "Development of methodologies and support systems for fire risk assessment in extreme weather conditions") and by the Italia-Francia Marittimo Programme (Proterina 2 Project).