The SAFERS “EO-based fire delineation and burnt area” intelligent service will establish an on-demand service able to delineate burned areas, fire fronts using Copernicus Sentinel-2 imagery, when available, and advanced artificial intelligence algorithms.
In detail, an AI module will be developed to delineate burned area boundaries and fire fronts based on Sentinel-2. Multiple machine and deep learning models will be used for comparison, and the best performing models will be operationalized to be triggered by other SAFERS detection services.
This service enables end-users to request a mapping activity by specifying the spatial-temporal references i.e. an area of interest and a time interval. Moreover, the service will produce maps according to the satellite data availability, including burned areas with a severity index and fire fronts.
Wildfire management phase: detection and response; adaptation and restoration.

Example of inputs and outputs of the mapping module: (a1) a Sentinel-2 tile (a2) severity estimation (a3) estimated delineation map
Wildfire management phase: detection and response; adaptation and restoration.

Example of inputs and outputs of the mapping module: (a1) a Sentinel-2 tile (a2) severity estimation (a3) estimated delineation map