Sentinel-2 Remote Sensing for Forest Fire Mapping and Climate Resilience: Operational Solutions for Mediterranean Ecosystems in Northern Algeria
Forest fires are an accelerating, climate-driven threat to Mediterranean ecosystems along the Silk Road, and Northern Algeria is experiencing increasingly severe episodes. This paper presents an operationally validated methodology for mapping burned areas from free Sentinel-2 imagery (Copernicus), combining multi-temporal preprocessing, NDVI, NBR and the differenced Normalized Burn Ratio (dNBR) with a four-class fire severity classification. Applied to a coastal forest zone in the wilaya of El Tarf, it mapped 3,815 ha of burned area, 64.2% of which was of high or very high severity. Validation against 120 ground control points yielded an overall accuracy of 87.4% (Cohen's Kappa = 0.81). The optimal detection window is 10–30 days after the fire, with vegetation regeneration over 4–6 months. Sentinel-1 SAR integration, machine-learning automation and institutional transfer to civil protection services are outlined as routes to operational scaling and transboundary early warning across the Silk Road.
Authors
- Seyfallah Bouraoui (ORCID: https://orcid.org/0000-0002-9317-8208)
Institutions
- University of Sciences and Technology Houari Boumediene (DZ)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-15
- DOI
- https://doi.org/10.5281/zenodo.24262917
- Primary Topic
- Fire effects on ecosystems
- Type
- article
- Field-Weighted Citation Impact
- 0.00