Post-Fire Mapping of Burned Areas Using Sentinel-2 Imagery to Support Ecosystem Restoration Activities

Wildfires are among the largest natural disasters in Europe, causing severe impacts on ecosystems, humans, settlements, and infrastructures. Burned area mapping is essential for planning restoration activities after a wildfire event. In this study, burn severity indices derived from Sentinel-2 datasets were used to map burn severity over Greece and Portugal, representing the first initiative within the planning of restoration activities.

Two megafires that occurred in Thasos, Greece in September 2016 and Médio Tejo, Portugal in August 2017 were considered. The Relativized Burn Ratio (RBR) and the differenced Normalized Burn (dNBR) were calculated for the two megafires and compared against the fire perimeters published in the European Forest Fire Information System (EFFIS). The classification of burn severity was performed using the dNBR classification. RBR and dNBR maps were derived via an automated procedure developed to support the rapid mapping and publishing of the burned areas and facilitate the planning of restoration activities.

The results show that both RBR and dNBR provide greater detail in delineating burned areas within the EFFIS fire perimeters, mapping, on average, unaffected areas corresponding to 27.5% and 20.7% of the total fire area in Thasos and Médio Tejo, respectively. The dNBR classification indicated higher burn severity compared to the RBR classification in both areas. Burn severity was higher in Thasos, where the percentage of area in the moderate-high severity class according to dNBR reached 22.6% of the burned area, compared to 9.9% in Médio Tejo.

RBR and dNBR derived by Sentinel-2 datasets can be used to generate detailed burn severity maps in-time and accelerate the planning of restoration activities. This study was implemented in the framework of the project “IT4EST-Interregional Tech for forEst SusTainability”, co funded by the European Union.