The wildfire risk estimates the danger that a vegetation fire reaches the site. The local fire weather (Fire Weather Index, ThinkHazard! levels) counts only if there is natural vegetation around the site that can carry the fire: without it the class is A. The euro come from the vegetation fires seen nearby by the MODIS satellites, not from fires that start inside the site.
Indicator. Yearly maximum Fire Weather Index (FWI), dimensionless (Canadian system), at return periods of 2, 10 and 30 years, and the natural vegetation around the site.
Data. ThinkHazard! wildfire layer (CSIRO for GFDRR, World Bank, CC BY 4.0): yearly maximum FWI at 2, 10 and 30 years, from a GEV distribution on the daily NASA GFWED data 1981-2016, 0.5 degree cells. Land cover ESA WorldCover 2021, 10 m. The days of high, very high and extreme FWI of the CMIP6 scenarios (Copernicus CDS, SSP1-2.6 / SSP2-4.5 / SSP3-7.0 / SSP5-8.5) are still computed and shown.
Calculation.
(1) Published ThinkHazard! categories (Leonard et al. 2017, Table 3): high if the FWI is above 30 at 2 years, medium if above 20 at 10 years, low if above 15 at 30 years, otherwise very low.
(2) Fuel: the FWI "does not consider the amount of fuel present" (same report). The site is exposed if it is in the wildland-urban interface of Schug et al. 2023: at least 50% natural vegetation within 500 m (woods, shrubs, grassland, wetlands, moss; cropland counts half, an xNatura choice), or less than 2.4 km, the distance embers travel, from a natural area of at least 5 km² with more than 75% vegetation. A site that is not exposed is A whatever the weather: no vegetation can bring a fire to it. A fire that starts inside the site, from its own processes, is not this risk.
Thresholds → level A–F.
• F: 2-year FWI of at least 50 (EFFIS "extreme" class)
• E: 2-year FWI above 30 (ThinkHazard! high)
• D: 10-year FWI above 20 and 2-year FWI from 21.3 (EFFIS "high" class)
• C: 10-year FWI above 20 (ThinkHazard! medium)
• B: 30-year FWI above 15 (ThinkHazard! low)
• A: otherwise, or site outside the wildland-urban interface.
The four categories are published; splitting medium and high into two letters uses the European EFFIS FWI classes given in the same report (San Miguel-Ayanz et al.): the six-letter scale is not in a single source. The class is the same in every scenario and year: the layer describes the 1981-2016 climate and no published rule turns the models' FWI days into its return periods.
From physical to impact. The physical level (A–F) is combined with the production process's sensitivity through the risk matrix (6 physical levels by 5 sensitivity bands) → impact level. The euro amount of this risk is the one described in the "Economic damage" paragraph (Version 2026): the damage factor per class of Version 2025 is no longer used for this risk.
Economic damage. Expected annual loss in euro, as a share of the asset value. Basis: "published source".
• CLIMADA wildfire impact function (Lüthi et al. 2021): share of the value destroyed f = i³ / (1 + i³), with i = max(brightness − 295 K, 0) / (409.4 K − 295 K), calibrated on 4 km cells.
• Fires: the MODIS active fires 2001-2024 of NASA FIRMS whose point falls in the 4 km square centred on the site, with confidence of at least 30.
• Vegetation fires only, as type 0 of the MODIS guide (Giglio et al. 2021): static hot spots, detected on 16 or more days in a year, and those on urban land (MODIS MCD12Q1) are left out. A furnace or a stack of the site or of nearby industry does not count.
• Each year counts its brightest fire; the annual loss is the mean of f over the 24 years (0 in years without fire). Outside the wildland-urban interface the loss is 0.
• Today's climate in every scenario and year; expected annual loss only, no 100-year loss from 24 years of data.
Stated limits: the function is calibrated on the total value of 4 km cells against reported disasters (mostly the United States and Australia, typical error of a factor of about 20, RMSF 22.6) and is applied to a single site; small fires that MODIS does not see are not counted.
Sources.
• [Leonard et al. 2017, ThinkHazard! Wildfire Hazard Layer, Technical Basis and Demonstration (CSIRO, GFDRR)](Leonard et al. 2017, ThinkHazard! Wildfire Hazard Layer, Technical Basis and Demonstration (CSIRO, GFDRR))
• [Global wildfire hazard (World Bank, CSIRO, GFDRR)](Global wildfire hazard (World Bank, CSIRO, GFDRR))
• [Comprehensive Monitoring of Wildfires in Europe: The European Forest Fire Information System](Comprehensive Monitoring of Wildfires in Europe: The European Forest Fire Information System)
• [Schug et al. 2023, The global wildland-urban interface](Schug et al. 2023, The global wildland-urban interface)
• [Zanaga et al. 2022, ESA WorldCover 10 m 2021 v200](Zanaga et al. 2022, ESA WorldCover 10 m 2021 v200)
• [Lüthi et al. 2021, Globally consistent assessment of economic impacts of wildfires in CLIMADA v2.2](Lüthi et al. 2021, Globally consistent assessment of economic impacts of wildfires in CLIMADA v2.2)
• [Giglio et al. 2021, MODIS Collection 6 and 6.1 Active Fire Product User's Guide](Giglio et al. 2021, MODIS Collection 6 and 6.1 Active Fire Product User's Guide)
• [NASA FIRMS, MODIS active fires](NASA FIRMS, MODIS active fires)
• [MODIS MCD12Q1 v061 land cover (NASA LP DAAC)](MODIS MCD12Q1 v061 land cover (NASA LP DAAC))
• [Development and structure of the Canadian Forest Fire Weather Index System. Canadian Forestry Service](Development and structure of the Canadian Forest Fire Weather Index System. Canadian Forestry Service)
• xclim.
Indicator. Yearly maximum Fire Weather Index (FWI), dimensionless (Canadian system), at return periods of 2, 10 and 30 years, and the natural vegetation around the site.
Data. ThinkHazard! wildfire layer (CSIRO for GFDRR, World Bank, CC BY 4.0): yearly maximum FWI at 2, 10 and 30 years, from a GEV distribution on the daily NASA GFWED data 1981-2016, 0.5 degree cells. Land cover ESA WorldCover 2021, 10 m. The days of high, very high and extreme FWI of the CMIP6 scenarios (Copernicus CDS, SSP1-2.6 / SSP2-4.5 / SSP3-7.0 / SSP5-8.5) are still computed and shown.
Calculation.
(1) Published ThinkHazard! categories (Leonard et al. 2017, Table 3): high if the FWI is above 30 at 2 years, medium if above 20 at 10 years, low if above 15 at 30 years, otherwise very low.
(2) Fuel: the FWI "does not consider the amount of fuel present" (same report). The site is exposed if it is in the wildland-urban interface of Schug et al. 2023: at least 50% natural vegetation within 500 m (woods, shrubs, grassland, wetlands, moss; cropland counts half, an xNatura choice), or less than 2.4 km, the distance embers travel, from a natural area of at least 5 km² with more than 75% vegetation. A site that is not exposed is A whatever the weather: no vegetation can bring a fire to it. A fire that starts inside the site, from its own processes, is not this risk.
Thresholds → level A–F.
• F: 2-year FWI of at least 50 (EFFIS "extreme" class)
• E: 2-year FWI above 30 (ThinkHazard! high)
• D: 10-year FWI above 20 and 2-year FWI from 21.3 (EFFIS "high" class)
• C: 10-year FWI above 20 (ThinkHazard! medium)
• B: 30-year FWI above 15 (ThinkHazard! low)
• A: otherwise, or site outside the wildland-urban interface.
The four categories are published; splitting medium and high into two letters uses the European EFFIS FWI classes given in the same report (San Miguel-Ayanz et al.): the six-letter scale is not in a single source. The class is the same in every scenario and year: the layer describes the 1981-2016 climate and no published rule turns the models' FWI days into its return periods.
From physical to impact. The physical level (A–F) is combined with the production process's sensitivity through the risk matrix (6 physical levels by 5 sensitivity bands) → impact level. The euro amount of this risk is the one described in the "Economic damage" paragraph (Version 2026): the damage factor per class of Version 2025 is no longer used for this risk.
Economic damage. Expected annual loss in euro, as a share of the asset value. Basis: "published source".
• CLIMADA wildfire impact function (Lüthi et al. 2021): share of the value destroyed f = i³ / (1 + i³), with i = max(brightness − 295 K, 0) / (409.4 K − 295 K), calibrated on 4 km cells.
• Fires: the MODIS active fires 2001-2024 of NASA FIRMS whose point falls in the 4 km square centred on the site, with confidence of at least 30.
• Vegetation fires only, as type 0 of the MODIS guide (Giglio et al. 2021): static hot spots, detected on 16 or more days in a year, and those on urban land (MODIS MCD12Q1) are left out. A furnace or a stack of the site or of nearby industry does not count.
• Each year counts its brightest fire; the annual loss is the mean of f over the 24 years (0 in years without fire). Outside the wildland-urban interface the loss is 0.
• Today's climate in every scenario and year; expected annual loss only, no 100-year loss from 24 years of data.
Stated limits: the function is calibrated on the total value of 4 km cells against reported disasters (mostly the United States and Australia, typical error of a factor of about 20, RMSF 22.6) and is applied to a single site; small fires that MODIS does not see are not counted.
Sources.
• [Leonard et al. 2017, ThinkHazard! Wildfire Hazard Layer, Technical Basis and Demonstration (CSIRO, GFDRR)](Leonard et al. 2017, ThinkHazard! Wildfire Hazard Layer, Technical Basis and Demonstration (CSIRO, GFDRR))
• [Global wildfire hazard (World Bank, CSIRO, GFDRR)](Global wildfire hazard (World Bank, CSIRO, GFDRR))
• [Comprehensive Monitoring of Wildfires in Europe: The European Forest Fire Information System](Comprehensive Monitoring of Wildfires in Europe: The European Forest Fire Information System)
• [Schug et al. 2023, The global wildland-urban interface](Schug et al. 2023, The global wildland-urban interface)
• [Zanaga et al. 2022, ESA WorldCover 10 m 2021 v200](Zanaga et al. 2022, ESA WorldCover 10 m 2021 v200)
• [Lüthi et al. 2021, Globally consistent assessment of economic impacts of wildfires in CLIMADA v2.2](Lüthi et al. 2021, Globally consistent assessment of economic impacts of wildfires in CLIMADA v2.2)
• [Giglio et al. 2021, MODIS Collection 6 and 6.1 Active Fire Product User's Guide](Giglio et al. 2021, MODIS Collection 6 and 6.1 Active Fire Product User's Guide)
• [NASA FIRMS, MODIS active fires](NASA FIRMS, MODIS active fires)
• [MODIS MCD12Q1 v061 land cover (NASA LP DAAC)](MODIS MCD12Q1 v061 land cover (NASA LP DAAC))
• [Development and structure of the Canadian Forest Fire Weather Index System. Canadian Forestry Service](Development and structure of the Canadian Forest Fire Weather Index System. Canadian Forestry Service)
• xclim.