What actually burned in the coastal wildfires, and where?
Burned-area boundaries and severity for the October 2020 fires (14,164 ha) and the July 2025 fires in rural Latakia (10,454 ha), computed from before-and-after images at 20 m, with burned area per sub-district and nearest village, plus 13,361 fire detections 2020–2024 in the western governorates.
In short
July 2025: 10,454 ha burned, 3,164 at high severity.
October 2020: 14,164 ha burned, 3,153 at high severity.
Most-burned sub-districts and villages are in the tables.
Who it is for
Agriculture and forestry faculties and directorates, reforestation and rehabilitation organisations, municipalities and journalists.
Method: what we add to the raw data
- Sentinel-2 images before and after each event were collected; the normalised burn ratio (NBR) was computed from near- and short-wave infrared with cloud and shadow masked, taking the median per period.
- The difference (dNBR) measures burn severity: 0.27–0.44 moderate-low, 0.44–0.66 moderate-high, above 0.66 high.
- Only trees, shrubland and grassland were kept, because harvesting fields also changes the index and looks like fire.
- Patches under one hectare were removed, and each patch linked to its sub-district and nearest village.
- Daily fire detections for 2020–2024 in the western governorates were added, excluding gas flares.
Product details
| What you get | 3 spatial layers in GeoPackage, 4 GeoTIFF rasters, 4 CSV tables, lyrx and qml symbology for ArcGIS Pro and QGIS, a PNG map at 220 dpi, and an Arabic and English guide |
|---|---|
| Formats | GPKG, GeoTIFF, CSV, lyrx, qml, PNG |
| Coordinate system | WGS 84 (EPSG:4326); engineering rasters in UTM 37N (EPSG:32637) |
| Licence | Commercial use allowed for you and the organisation you buy for; no resale or redistribution |
| Delivery | A private download link valid for 24 hours, within 6 hours of payment confirmation |
| Package size | ZIP 55.4 MB |
| Personal data | The product contains no personal data |
Spatial layers
burned_areas— polygons, 49,476: Burned areas by severitysubdistricts_burned— polygons, 55: Sub-districts with burned area per eventfire_hotspots_2020_2024— points, 13,361: Fire detections
Raster files (GeoTIFF)
D081_dnbr_2020_10.tif— 3,628 × 7,587, 20 m: dNBR — October 2020 firesD081_dnbr_2025_07.tif— 2,928 × 2,693, 20 m: dNBR — July 2025 firesD081_severity_2020_10.tif— 3,628 × 7,587, 20 m: Severity classes — October 2020 firesD081_severity_2025_07.tif— 2,928 × 2,693, 20 m: Severity classes — July 2025 fires
Tables
D081_burned_by_subdistrict.csv— 55 rows: Burned per sub-districtD081_burned_by_village.csv— 360 rows: Burned per villageD081_hotspots_by_subdistrict_year.csv— 116 rows: Hotspots per sub-district and yearD081_events_summary.csv— 2 rows: Event summary
Limitations
- The 0.27 threshold is conservative: light burning (0.1–0.27) is outside the computed areas but visible in the dNBR cells.
- Fires in fields and orchards are excluded because the analysis is limited to natural cover.
- Fire detections for 2025 are not yet published in the annual files, so points run to 2024.
FAQ
- How long does delivery take?
- Up to 6 hours after payment is confirmed.
- How long does the download link last?
- 24 hours, and it is renewed free on request.
- Can I get a refund?
- Digital products are not refundable once downloaded — unless the file is defective or does not match the description, in which case it is replaced or fully refunded.
- What does the licence allow?
- Commercial use is allowed for you and the organisation you buy for; resale or redistribution is not.