Statistical Evaluation of the Impact of Wildfires on Forest Habitats using Earth Observation data and Machine Learning

Wildfires constitute a threat to both environment and human activities. Furthermore, they represent one of the major disturbances to biodiversity and its natural capital. The accurate identification and post-fire monitoring of burned areas in forest habitats is therefore important for research and institutions to implement effective restoration policies. ISPRA, as the Italian National Environmental Agency, has developed an environmental surveillance and classification system based on Earth Observation data, aimed at the assessment and post-fire monitoring of forest habitat affected by wildfires. Wildfires’ burned areas are identified and classified in a geo-spatial database in order to allow a real time monitoring of the impact of fires on forest habitat, as well as the statistical analysis at different spatial scales. This will hopefully support policies dedicated to the ecological recovery of damaged ecosystems over time.