RISICO, An Enhanced Forest Fire Danger Rating System: Validation on 2021 Extreme Wildfire Season in Southern Italy

Forest Fire Danger Rating (FFDR) models are widely used in fire management decision-making, from daily operations, to seasonal planning and long-term land management. Early warning of extreme fire danger is necessary for Civil Protection Authorities to mitigate wildfire disasters. Since 2000, the Liguria Region (Italy) and the Italian Civil Protection have conducted independent research programs that led to the development of the FFDR system RISICO [1,2,3,4]. RISICO integrates meteorological observations and forecasts provided by Limited Area Models with vegetation cover and topography as additional inputs to the system. This system is currently adopted at national level in Italy by the Civil Protection Agency (Dipartimento della Protezione Civile). The main use of the information provided by the RISICO system is to support decision-makers at national level in all decisions concerning alert messages to regional authorities. In Liguria, RISICO is used at regional level in order to issue a daily fire hazard bulletin and organize patrolling and monitoring activities. The system has also been employed in Lebanon and Albania since 2011 by the local Civil Defense, in Bolivia since late 2019 by the local Civil Defense and Hydrometeorological Authority, and in Ethiopia since 2021, providing thrice-weekly bulletins at district level. This FFDR system also helped in the framework of the Aristotle “enhanced European Natural Hazard Scientific Partnership” at a pan-European and global scale. RISICO delivers forecasts for Fine Fuel Moisture Content, Rate Of Spread and Fire-line Intensity. Such outputs can be visualized as three-hourly spatial layers, or aggregated temporally (daily averages, percentiles, etc.) and/or spatially (at regional or district level). Static susceptibility maps (computed via Machine Learning Techniques) and phenological information from satellite (Sentinel II) allowed for further improvement of the model. The catastrophic wildfires of Summer 2021 in Southern Italy allowed for a thorough validation of the model.