Fire weather - WorldWeatherAttribution/wwa-wiki GitHub Wiki

There are several indices available to characterise weather-related fire risk, and the most appropriate will depend on the region being studied. We strongly recommend consulting a local fire weather expert to understand the most relevant metric for the study region.

---

Fire weather index (FWI) & daily severity rating (DSR)

In many forested regions, the FWI is a useful indicator of weather-related fire risk, taking into account daily precipitation and noon temperature, relative humidity and wind speed over the preceding days, weeks and months. The FWI is designed to reflect the immediate expected radiative power if ignition were to occur, but is not designed to represent longer-term risks. Instead, when looking at weather-related fire risk over periods of more than a few days, we use the daily severity rating (DSR), a power function of the FWI that provides larger weight to higher values than lower ones, and can be accumulated over weeks or months to reflect the longer-term hazard.

The FWI is ideally computed from observations of temperature, wind speed, humidity and 24-hour accumulated precipitation at local solar noon, but where local noon data is unavailable, a common approach is to linearly interpolate between the nearest available time steps. If necessary (as may be the case for climate models), the daily maximum temperature and daily minimum relative humidity can be used, along with mean wind speed and daily accumulated precipitation.

To calculate the FWI, use the xclim.indices.cffwis_indices python package.
The DSR is calculated from the daily FWI as $0.0272 \times FWI ^{1.77}$.

The FWI is usually used for shorter-duration risk, and annual maximum FWI is typically well modelled by a GEV that shifts with GMST.
Accumulated DSR is often skewed but can usually be transformed to approximate normality by taking logDSR, which shifts with GMST. The resulting absolute changes in logDSR intensity $\Delta_I$ can be converted to relative changes in DSR intensity by taking $100(e^{\Delta_I} - 1)$.

 

---

Hot-dry-windy index (HDWI) & vapour-pressure deficit (VPD)

The Hot-Dry-Windy Index (HDWI; Srock et al., 2018) is a fire weather metric designed for identifying days with weather conditions that make it difficult to control the spread of a wildfire. This index is based on a physical understanding of how fire interacts with the atmosphere, primarily through wind, temperature, and moisture.

Strong winds during a wildfire can cause it to spread rapidly making it more difficult to contain or control. The HDWI incorporates wind speeds, recognizing its critical role in wildfire management. Although shifts in wind direction also influence fire behaviour, they are deemed as less impactful and are therefore excluded from the formulation.

Temperature and moisture jointly influence the evaporative potential of the atmosphere, affecting fuel consumption, fire intensity, and spread. This is captured using the vapour pressure deficit (VPD), which is a measure of evaporative potential of the atmosphere.

• If relative humidity is available, the daily VPD is the maximum of the hourly VPD values $VPD_h$, calculated using hourly temperature $T_h$ (in C) and relative humidity $RH_h$ as

$$VPD_h = \left(1-RH_h\right)\exp{\left(\frac{17.25 T_h}{243.04+T_h}\right)}$$

• If dew point temperature $TD$ is available instead,

$$VPD_h = \exp{\left(\frac{17.25T_h}{243.04+T_h}\right)} - \exp{\left(\frac{17.25TD_h}{243.04+TD_h}\right)}$$

• The HDWI is then calculated as $U_{max} \times VPD_{max}$ where $U_{max}$ is the daily maximum of the hourly sustained wind speeds, and $VPD_{max}$ is the daily maximum of the hourly vapour pressure deficit.

HDWI was used to characterise the fire weather conditions in the WWA rapid study on the wildfires that impacted Chile in February 2024.