This dataset provides global monthly records of burned area and fuel characteristics at ~25 km spatial resolution from 1950–2099. It combines historical observations (1950–2021) with climate-driven projections (2022–2099).
Burned area and fuel characteristics are generated using a machine learning model that integrates climate variables (temperature, precipitation, humidity, wind) from ERA5 (historical) and CMIP6 (projections), together with socioeconomic data including population, land use, and development indicators. This combined approach captures fire weather and fuel drivers, and human influences on fire activity.
Fuel variables include live and dead biomass components for both load and moisture, all modelled to remain consistent with burned area estimates.
Future projections use CMIP6 climate scenarios under Shared Socioeconomic Pathways (SSPs), allowing exploration of wildfire activity across different climate and socioeconomic futures.
The dataset uses a uniform grid, variable set, and calendar structure across the historical and projection periods, enabling direct comparison between periods and scenarios.
Intended applications include wildfire risk assessment, fire regime analysis, carbon cycle modelling, and climate impact studies.
This is version 1.0. Future versions will incorporate model improvements and extended temporal coverage.

ECMWF

Modelled global fuel characteristics and burned area from 1950 to 2099

This dataset provides a record of fuel characteristics at high spatiotemporal resolution: ~9km, daily.
The two main variable groups are fuel load and fuel moisture, both of which are further divided by live/dead and wood/foliage fractions.
The dataset combines state-of-the-art model data (ERA5-Land) with observations from multiple satellites and in-situ variables into a globally complete and consistent dataset. 
The data provides high spatiotemporal resolution data of fuel load, which is essential for modelling wildfire activity, which contributes directly to the derivation of multiple Essential Climate Variables (ECVs).
The data is relevant not only for the wildfire community but also for studying various biogeochemical processes related to land-atmosphere interactions.
The fuel load is initially informed using the static European Space Agency Climate Change Initiative biomass product (ESA-CCI), which provides high-resolution estimates of above ground biomass. Crucially our dataset adds a time evolution of the dataset to provide a daily product with accurate seasonality based on modelled Carbon Dixoide (CO₂) exchange. 
Furthermore, by collecting estimates on biomass allocation and modelling techniques the dataset allocates the biomass into 4 categories of fuel load, which is relevant for wildfires but not currently estimated from the ESA-CCI product. 
Dead fuel moisture content is based on existing modelling principles which have not been scaled up to the global scale nor have they used the high accuracy mapping and output data available within the Integrated Forecast System (IFS) or the  the fifth generation of reanalysis data produced by the European Centre for Medium-Range Weather Forecasts (ERA5). 
The live fuel moisture model was trained on a global insitu dataset which is based on sampling and does not provide a global product, this is the first attempt to achieve such a product using modelling. All variables were validated using independent observations when available.
This is a first version of the dataset which will be updated to provide improved accuracy and expanded through time.

Hours	Number of requests	Number of users	Downloaded data
Last 1H	0	1	0 GB
Last 2H	0	1	0 GB
Last 3H	3	1	6 GB
Last 6H	5	2	6 GB
Last 12H	45	3	269 GB
Last 24H	149	5	1355 GB

	Country	Number of requests	Number of users	Downloaded data
1	NL	12	0	<1 GB
2	AD	0	0	0 GB
3	AL	0	0	0 GB
4	AM	0	0	0 GB
5	AT	0	0	0 GB
6	BA	0	0	0 GB

Live dashboards

Downloads

Completed requests

Dataset popularity

Country

Queue status