Factsheet 10: Fire weather

Many Australian landscapes are fire prone and have been severely damaged by bushfires in the recent past. Understanding how a changing climate could affect the weather conditions conducive to bushfires is essential for building preparedness.

In Australia, the McArthur Forest Fire Danger Index (FFDI) (McArthur, 1967) is frequently used to quantify the influence of weather on fire risk, with the Australian Bureau of Meteorology (BoM) providing short-term forecasts of FFDI for use by fire authorities for several decades up until the end of 2022 (Sauvage et al., 2024). McArthur defined FFDI to assist foresters in relating weather conditions to the expected fire behaviour. The FFDI uses observations of temperature, relative humidity and wind speed with an estimate of the fuel state to predict fire risk. The fuel state is determined by the 'drought factor' which depends on the daily rainfall and the period of time since the last rain.

Note, unlike other methods such as the Australian Fire Danger Rating System (AFDRS), the McArthur FFDI only assesses the atmospheric or climatic component of wildfire risk. This allows for projections of fire weather while avoiding issues associated with estimating future fuel state under changing land uses and management practices, and also allows for continuity with historical weather data to inform hazard and risk assessments.

The Queensland Future Climate Science Program has used state-of-the-art high-resolution downscaled climate simulations from both CMIP5 and CMIP6 global climate models to assess the impact of climate change on fire weather.

The Queensland Future Climate Dashboard and Regional Explorer offer several options for exploring projections of future fire weather, including:

  • high-resolution interactive maps and plots of future changes in FFDI across Queensland for different regions, emission scenarios, seasons and time slices
  • table summaries that provide a quick and easy way to access summary FFDI projections at a regional level
  • customisable timeseries charts of modelled FFDI anomalies for the period 1981-2100.

FFDI calculation

The Queensland Government calculated the McArthur FFDI using the following equation:

FFDI = 2e(0.0338 T + 0.0234 W-0.0345 RH + 0.987 ln(DF)-0.45)

where the daily maximum temperature at 2m (T), mid-afternoon (3 pm) relative humidity at 2m (RH), and mid-afternoon (3 pm) 10 m wind speed (W) were derived using bias-corrected downscaled climate simulations from the Queensland Future Climate Science Program (see Trancoso et al. 2022 for an example of estimating FFDI using a similar approach).

The drought factor (DF) was based on an estimate of soil moisture deficit (MD) calculated using the Keetch-Byram Drought Index (KBDI), where the KBDI is estimated from daily rainfall and the daily maximum temperature. For a detailed description of the drought factor as well as assessment of the performance of different drought factors see Holgate et al. (2017).

Note that the McArthur FFDI only assesses the atmospheric or climatic component of wildfire risk and does not account for landscape factors such as vegetation characteristics, fuel availability and potential sources of fire ignition.

Seven categories for FFDI were selected to represent a range of fire danger classifications, with indices based on the annual count of days within each of the categories:

  1. Severe fire risk days: Seasonal count of days where FFDI ≥ 50
  2. Very high fire risk days: Seasonal count of days where 24 ≤ FFDI &lt 50
  3. High fire risk days: Seasonal count of days where 12 ≤ FFDI &lt 24
  4. Moderate fire risk days: Seasonal count of days where 5 ≤ FFDI &lt 12
  5. Low fire risk days: Seasonal count of days where 0 ≤ FFDI &lt 5
  6. 95th Percentile fire risk days: Seasonal count of days with an FFDI greater or equal to the 95th percentile for the reference period (1995-2014)
  7. 99.7th Percentile fire risk days: Seasonal count of days with an FFDI greater or equal to the 99.726th percentile for the reference period (1995-2014).

FFDI counts were calculated for each of the downscaled regional climate models for the different seasons (annual, summer, autumn, winter, spring, wet and dry), and several emissions scenarios (RCP4.5 and RCP8.5 for the CMIP5 modelling). A multi-model average of the downscaled models was also calculated.

Results on the Queensland Future Climate website are shown as change relative to a reference period (1986-2005 for the CMIP5 modelling).

Fire weather on Queensland Future Climate

The main tools for viewing climate projections data on the Queensland Future Climate website are the Queensland Future Climate Dashboard and Regional Explorer. Please refer to the user guide for detailed information on how to access and interpret information available from these resources.

Fire weather information is initially available for CMIP5 models only, but CMIP6 information will be released in the near future.

Examples of the FFDI output available on the Queensland Future Climate Dashboard and Regional Explorer are shown below in Figures 1 and 2.

Fire weather figure
Figure 1: Sample map and plots from the CMIP5 Queensland Future Climate Dashboard showing changes in very high fire risk days in the Whitsunday Regional Council area by 2070 for a high emissions scenario (RCP8.5). The graphs on the right of the Dashboard display the seasonal variation in the chosen metric and the change over four time periods to 2090. The map also allows the download of fire weather layers as shapefiles for easy overlay with other layers (e.g. vegetation type, towns or building locations).
Fire weather figure
Figure 2: Sample timeseries plot from the CMIP5 Regional Explorer showing projected changes in very high fire risk days in the Whitsunday Regional Council area for the two emissions scenarios (RCP4.5 and RCP8.5). The timeseries plots allow users to compare projections for the different emissions scenarios and view results for individual models at annual time steps.

References

Holgate, C.M., van Dijk, A.I.J.M., Cary, G.J., Yebra, M., 2017. Using alternative soil moisture estimates in the McArthur Forest Fire Danger Index. Int. J. Wildland Fire 26, 806. https://doi.org/10.1071/WF16217

Sauvage, S., Fox-Hughes, P., Matthews, S., Kenny, B.J., Hollis, J.J., Grootemaat, S., Runcie, J.W., Holmes, A., Harris, R.M.B., Love, P.T., Williamson, G., 2024. Australian Fire Danger Rating System Research Prototype: a climatology†. Int. J. Wildland Fire 33. https://doi.org/10.1071/WF23144

Trancoso, R., Syktus, J., Salazar, A., Thatcher, M., Toombs, N., Wong, K.K.-H., Meijaard, E., Sheil, D., McAlpine, C.A., 2022. Converting tropical forests to agriculture increases fire risk by fourfold. Environ. Res. Lett. 17, 104019. https://doi.org/10.1088/1748-9326/ac8f5c

Last updated: 21 February 2025
Creative Commons Attribution 4.0 International (CC BY 4.0)