Fire Weather Index (FWI)

The Fire Weather Index (FWI) system is a weather-based framework for estimating fire danger. Developed by the Canadian Forest Service in 1970 and formalised by Van Wagner (1987), it has become the international standard — adopted by EFFIS (European Forest Fire Information System) as the pan-European fire danger rating system since 2007.

The FWI does not predict whether a fire will start. It estimates how a fire would behave if one did, based on current weather conditions and cumulative drying.

Why it matters for fire weather

The FWI system is the primary fire danger index used across Europe and in over 30 countries worldwide. It translates four weather variables — temperature, relative humidity, wind, and precipitation — into a single number that reflects expected fire intensity.

Its strength is that it integrates multiple weather factors over time. Individual variables like temperature or humidity describe the current atmosphere; the FWI captures how those conditions have accumulated to affect fuel moisture across different depths, and what fire behaviour to expect as a result.

How it works

The FWI system consists of six components in two tiers, calculated daily from noon weather observations (Van Wagner, 1987).

Fuel moisture codes

Three codes track how dry different fuel layers are, each responding at a different timescale:

Code	Full name	What it tracks	Response time
FFMC	Fine Fuel Moisture Code	Surface litter and fine fuels (≤ 2 cm)	~16 hours
DMC	Duff Moisture Code	Loosely compacted organic layer (5–10 cm)	~12 days
DC	Drought Code	Deep, compact organic layers (10+ cm)	~52 days

The FFMC responds to today’s weather. The DMC integrates conditions over roughly two weeks. The DC reflects cumulative drought over months — it rises steadily during dry periods and requires sustained, heavy rain to decrease.

Fire behaviour indices

Three indices combine the moisture codes with wind to predict fire behaviour:

Index	Full name	What it predicts	Inputs
ISI	Initial Spread Index	Expected fire spread rate	FFMC + wind speed
BUI	Build-Up Index	Total fuel available for combustion	DMC + DC
FWI	Fire Weather Index	Overall fire intensity	ISI + BUI

The final FWI value combines spread rate (ISI) with fuel availability (BUI) to produce an estimate of fire intensity — specifically, the expected energy output per unit length of fire front.

Key thresholds

The EFFIS danger classes below are the European standard, based on analysis by Vitolo et al. (2020) using 40+ years of ERA5 reanalysis data. They are used across Europe for harmonised fire danger communication.

FWI range	EFFIS class	Context
0–5.2	Very Low	Fires unlikely to sustain spread
5.2–11.2	Low	Fires possible but spread slowly
11.2–21.3	Moderate	Active fire spread possible
21.3–38.0	High	Significant fire behaviour expected
38.0–50.0	Very High	Intense fire behaviour
50.0–70.0	Extreme	Introduced during EU harmonisation
> 70.0	Very Extreme	Introduced June 2021 after successive Mediterranean heat waves

Note

These are pan-European thresholds. The same FWI value means different things in different climates. FWI = 30 is classified as “Very High” in Canada (boreal forest) but only “High” at the EFFIS European level, and sits at the moderate-to-high boundary in Greece (Mediterranean). Regional calibrations exist for this reason — see Going deeper below.

How to read it in Wildflyer

Wildflyer computes FWI from weather station observations and weather model data. You can:

• View current FWI values at weather stations
• See forecasted FWI for the coming days
• Track historical FWI trends for your area
• Compare station-level vs. grid-level FWI data

The FWI components (FFMC, DMC, DC, ISI, BUI) are also available individually in the expert view, allowing you to understand why the FWI is at its current level — whether it’s driven by wind and dry surface fuels (high ISI) or by deep cumulative drought (high BUI).

Going deeper

Regional calibration

The FWI was developed for Canadian boreal forests, where summer Drought Code values typically reach 200–300 and fire seasons last weeks. In the Mediterranean, DC routinely reaches 400–600 during summer, fire seasons span 3–5 months, and the baseline fire danger is inherently higher (Dimitrakopoulos & Bemmerzouk, 2011).

This means:

• Absolute FWI thresholds don’t transfer directly between climates. An FWI of 30 in Canada indicates extreme conditions; in southern Greece it can be a normal summer day.
• The DC component is less discriminating in Mediterranean climates — it routinely saturates at high values and does not differentiate well between dangerous and merely dry conditions.
• ISI and BUI become the driving indices in Mediterranean environments, as they capture the daily variability that matters for assessment.

The percentile approach

To address regional differences, EFFIS increasingly uses a percentile-based calibration (Vitolo et al., 2020):

1. Calculate the historical FWI distribution for each location from 40+ years of data
2. Express today’s FWI as a percentile of that distribution
3. Classify based on percentiles (e.g., > 95th percentile = extreme) rather than absolute values

This automatically adjusts for local climate. What counts as “extreme” in Finland is statistically extreme for Finland, and what counts as “extreme” in Greece is statistically extreme for Greece, even though the absolute FWI values differ substantially.

Limitations

The FWI system has known limitations:

• No fuel type information — it assumes a generic pine forest. Actual fire behaviour depends heavily on vegetation type (maquis, grassland, eucalyptus all behave differently).
• Wind direction not captured — the ISI uses wind speed but not direction, yet wind shifts are among the most dangerous fire weather events.
• Daily resolution — the classic FWI uses noon observations, missing intra-day variability. The FWI 2025 update addresses this with hourly calculations.
• No topography — slope and aspect dramatically affect fire behaviour but are not represented in the FWI.

These limitations are why fire danger assessment should combine FWI with other indices (HDWI, Haines) and direct weather observations.

Sources

• Van Wagner, C.E. (1987). Development and structure of the Canadian Forest Fire Weather Index System. Forestry Technical Report 35, Canadian Forest Service.
• Vitolo, C., Di Giuseppe, F., Krzeminski, B., & San-Miguel-Ayanz, J. (2020). ERA5-based global meteorological wildfire danger maps. Scientific Data, 7: 216.
• Dimitrakopoulos, A.P. & Bemmerzouk, A.M. (2011). Evaluation of the Canadian fire weather index system in an eastern Mediterranean environment. Meteorological Applications, 18(1): 83–93.
• San-Miguel-Ayanz, J. et al. (2023). Forest Fires in Europe, Middle East and North Africa 2022. JRC Technical Reports, European Commission.