Fire Danger Forecast

In 2007, after a test phase of 5 years during which different national fire danger indices were implemented in EFFIS, the EFFIS network adopted the Canadian Forest Fire Weather Index (FWI) System as the method to assess the fire danger level in a harmonized way throughout Europe. However, given the different climatic conditions in Europe, EFFIS publishes two indicators that provide information on the local/temporal variability of the FWI compared to a historical series of approximately 30 years. These indicators are the ranking, which provides percentiles of occurrence of the values, and the anomaly, computed as a standard deviation from the 30-year historical mean values. These indices are available in the pull down menu of the fire danger section.  

Currently, the fire danger forecast module of EFFIS provides access to fire danger indices using numerical weather forecast from two deterministic models i.e. ECMWF (8 km) and MeteoFrance (10 km), and one probabilistic model, the ECMWF Probabilistic model, at 18 km spatial resolution.  You can select the model in the pull-down menu of the EFFIS "current situation viewer."  


ECMWF and MeteoFrance deterministic models

FWI is computed from the ECMWF model (8 km), which provides 1 to 9 days forecasts, and from the MeteoFrance model (10 km), which provides up to 3 days forecasts.  The Fire Weather Index is mapped in 6 classes (very low, low, medium, high, very high and extreme). The fire danger classes are the same for all countries and maps show a harmonized picture of the spatial distribution of fire danger level throughout Europe, Middle East and North Africa.  The values for the FWI and its sub-components are provided below.


A "Very Extreme" Fire Danger Class was introduced in June 2021 to provide discrimination about the level of fire danger in extensive areas that were initially classified at "Extreme" Fire Danger in the Mediterranean region during the summer months. The "Very Extreme" class include areas with FWI values above 70.

In 2019, due to the interest of countries to compare the performance of the FWI with other relevant fire danger indices, the Australian McArthur Forest Fire Danger Index (MARK-5), the Keetch-Byram Drought Index (KBDI) and the National Fire Danger Rating System (NFDRS) have been made available through the EFFIS Fire Danger Forecast module. Information on these indices, which are all computed from numerical weather predictions of the ECMWF deterministic model, is available at the links provided in this paragraph. 


ECMWF Probabilistic model

The ECMWF Probabilistic model provides four fire danger indices: (1) FWI Extreme Forecast Index ( FWI EFI), (2) FWI Shift of Tails (FWI SOT), (3) Fine Fuel Moisture Content  Extreme Forecast Index (FFMC EFI) and (4) Fine Fuel Moisture Content Shift of Tails (FFMC SOT).  

The EFI is calculated by computing the difference between the climate model of the last 20 years for a given day and an 11-model ensamble forecast for that day, using 5 start dates in different years.  This larger window brings the benefit of better defining the climatology tails - i.e. the more extreme values - which is important for a consistent computations of EFI and SOT.  The Shift of Tails index complements the EFI by providing information about how extreme an event could potentially be. Specifically, SOT for FWI compares the tails of the distributions for the climate model and the ensemble using the 90th and 99th (upper tail) percentiles. Positive SOT values indicate that at least 10% of the ensemble is forecasting an extreme event above what observed in the climate. A high value of SOT shows how extreme the top 10% ENS results are.   computed at ECMWF is available at this link."> Additional information on EFI and SOT computed at ECMWF is available at this link.

The maps of forecasted fire danger level can be consulted through the web mapping interface of EFFIS and are also emailed daily to the users.

Additional information on the map viewer tool downloading the User Guide to the EFFIS Current Situation.