Data Description

This page describes the Wildfire FWI variables, units, conventions, spatial and temporal characteristics, and file formats.

Available variables

The output of the Wildfire FWI application consists of indicators derived from the Canadian Forest Fire Weather Index (FWI) system (Van Wagner, 1987). These indicators describe fire weather conditions based on atmospheric variables.

Note

For additional information, see the Wildfire FWI entry in the Destination Earth Data Portfolio.

FWI indicators

The Wildfire FWI application computes the Fire Weather Index (FWI) and related indices on a daily basis using meteorological input data.

The FWI system consists of several components:

  • Fuel moisture codes:

    • Fine Fuel Moisture Code (FFMC): describes the moisture content of surface fuels

    • Duff Moisture Code (DMC): represents moisture in medium-depth organic layers

    • Drought Code (DC): reflects long-term dryness in deep soil layers

  • Fire behaviour indices:

    • Initial Spread Index (ISI): estimates potential fire spread rate

    • Buildup Index (BUI): indicates available fuel

  • Composite index:

    • Fire Weather Index (FWI): represents potential fire intensity

These indices provide a consistent framework for assessing wildfire danger based on atmospheric conditions.

Inputs, assumptions, and interpretation

FWI is a meteorological fire danger indicator. It uses atmospheric inputs to estimate fuel dryness and potential fire intensity, but it does not directly represent burned area, ignition probability, fire suppression, land management, or exposure. Users should therefore interpret FWI as a measure of fire-conducive weather rather than realised wildfire risk.

The required meteorological inputs are temperature, humidity, wind speed, and 24-hour accumulated precipitation. FWI is a threshold-based indicator, and biases in these input variables can significantly bias the FWI calculation. The transformation from meteorological inputs to FWI and related fire danger classes is nonlinear, so biases in individual input variables do not translate directly or proportionally into FWI biases.

Because precipitation, humidity, and wind can vary strongly over short distances, differences in model resolution and regridding can also affect local FWI values and the number of days assigned to higher fire danger classes.

The FWI system was originally developed for Canadian forest conditions and has been adapted for global application. It does not include explicit information on vegetation, fuel type, fuel continuity, snow or ice cover, ignition sources, or human activity. As a result, high FWI values may occur in areas where the meteorological conditions are fire-conducive but where actual burning is unlikely or physically unrealistic, such as sparsely vegetated deserts, ice margins, or parts of Antarctica. For applications requiring realised fire risk, FWI should be combined with local or sector-specific information on vegetation, fuels, land cover, ignition, and exposure.

Further scientific context is available from the EFFIS fire danger forecast technical background and the Copernicus Fire Weather Index dataset description.

Wildfire FWI output

Variable

Short name

Unit

Dimensions

Description

Fire Weather Index

FWI

dimensionless

time, lat, lon

Indicator of potential fire intensity based on combined fire weather conditions.

Additional derived products

In addition to daily FWI values, the application may provide aggregated statistics such as:

  • Percentiles (e.g. median, 95th, 99th)

  • Threshold exceedances (e.g. number of high fire danger days per year)

  • Domain-averaged time series

These products support long-term assessment of wildfire risk and variability.

Spatial and temporal coverage

Wildfire FWI output is provided globally on a regular latitude/longitude grid at the native resolution of the Climate DT simulations (kilometre-scale).

The temporal resolution is daily, consistent with the FWI calculation.

The temporal coverage follows the underlying climate simulations, including both historical and projection periods.

File formats

The output is provided in NetCDF format.

  • Each file contains gridded FWI-related variables

  • Coordinates include latitude, longitude, and time

  • Files follow standard conventions for climate data interoperability

References

Van Wagner, C. E. (1987). Development and structure of the Canadian Forest Fire Weather Index System. Forestry Technical Report 35. Chalk River, Ontario: Canadian Forest Service, Petawawa National Forestry Institute. 37 pp. Available from the Natural Resources Canada Open Science and Technology Repository.