Wildfire Risk Assessment Using Earth Observation Data: A Case Study of the Eastern Carpathians at the Slovak-Ukrainian Frontier

Wildfires, as a global phenomenon, are one of the primary sources of environmental and social disturbances. To prevent and mitigate the damaging effects of fires in Europe, a great concern is focused on fire risk assessment at the Pan-European scale. The research proposes a conceptual approach for fire risk assessment using multiscale multi-temporal Earth observation data. The elaborated methodology integrates time series of components associated with the fire process, and risk analysis is performed according to hazard functions. The fuel load was restored by multivariate regression, with the 1 km FirEUrisk fuel types map and leaf area index obtained from 10 m Sentinel-2. Meteorological conditions were restored from soil moisture modelling based on 10 m radar Sentinel-1, optical Sentinel-2, and ALOS DEM data, as well as 1 km MODIS land surface temperature. The proposed approach was tested on natural and semi-natural mountain landscapes of the Eastern Carpathians at the Slovak-Ukrainian frontier. The resulting map of fire risk was referred to four gradations: high, existential, potential, and no risk. The result was validated using the FIRMS NASA fire data archive.