Author: Ahmad Merii
What this shows
A relative, drought-aware wildfire susceptibility index for Norfolk, computed per year over the June to July window, 2022 to 2026. Higher values mean conditions more conducive to ignition and spread. It is an index, not a probability or a forecast.
How it is built
Multi-criteria decision analysis by weighted overlay. Several environmental layers are each normalised to 0 to 1, then combined as a weighted sum. The components are:
- Fuel: vegetation quantity from Sentinel-2 (greenness / NDVI), June–July median.
- Moisture: vegetation and surface moisture (NDWI / NDMI); drier surfaces raise the score.
- Terrain: slope and aspect from SRTM; steeper, sun-facing slopes score higher.
- Fire weather: temperature, wind, and humidity from ERA5-Land over the window.
- Human ignition: proximity to roads and settlements from OSM / population layers.
Why it is drought-aware
The moisture and fire-weather terms respond to each year's own conditions, so dry summers push the index up where fuel is present. That is what makes the five years comparable as a series rather than five unrelated maps.
Processing and export
Per-year composites are built in Google Earth Engine (JavaScript), cloud-masked, reduced to a June–July median, normalised, and combined by the weighted sum. Outputs are exported to Google Drive as GeoTIFF in EPSG:27700, then converted with GDAL to Cloud-Optimised GeoTIFF in EPSG:4326, masked to the Norfolk coastline. True-colour Sentinel-2 (June–July median) is exported alongside.
How it is drawn here
The index is stretched over one fixed window (0.20 to 0.80) with the same ramp every year, so years compare directly. Change view maps right minus left over ±0.15; differences under 0.03 are hidden so real increases and decreases stand out.
Caveats
- Relative, not absolute — a susceptibility index, not a fire probability.
- MCDA weights are a modelling choice; different weights shift the pattern.
- A single June–July window can miss events outside it; cloud gaps can leave artefacts.
- Rasters are downsampled (~2400 px) for the web; use the source exports for analysis.
- Not for operational fire response. Validate against recorded incidents before drawing conclusions.
Data sourcing and attribution
- Modified Copernicus Sentinel data 2022, 2023, 2024, 2025, 2026, Sentinel-2 (ESA). Contains modified Copernicus Sentinel data; not officially endorsed by the EU or ESA.
- Generated using Copernicus Climate Change Service information 2022, 2023, 2024, 2025, 2026, ERA5-Land. Neither the European Commission nor ECMWF is responsible for any use that may be made of the Copernicus information or data it contains.
- SRTM elevation data courtesy of NASA / USGS (public domain).
- Contains information from OpenStreetMap, OpenStreetMap contributors, made available under the Open Database License (ODbL).
- Basemap CARTO, OpenStreetMap contributors (shown on the map).