A global fuel characteristic model and dataset for wildfire prediction Joe R. McNorton and Francesca Di Giuseppe present a mid-complexity model combining data-driven and analytical methods to predict fuel characteristics for wildfire prediction. The model integrates data from meteorological variables, land surface models, and satellite observations to provide real-time forecasts and climate records. The dataset, with high spatiotemporal resolution (approximately 9 km, daily), is the first of its kind and will be regularly updated. The model partitions fuel load and moisture into live and dead fuels, including both wood and foliage components. An example is a 10-year dataset that is well correlated with independent data and largely explains observed fire activity globally. While dead fuel moisture correlates highest with fire activity, live fuel moisture and load are shown to potentially enhance prediction skill. The use of observation data to inform a dynamical model is a crucial first step toward disentangling the contributing factors of fuel and weather to understand fire evolution globally. The model estimates fuel load and moisture content using satellite-derived above-ground biomass (AGB) data and modelled net ecosystem exchange (NEE). AGB is estimated using the ESA Biomass Climate Change Initiative (ESA-CCI) version 3 dataset and NEE data generated by the ECLand land surface model. The model further subdivides AGB into foliage and wood mass, considering both alive and dead components. The model also estimates live and dead fuel moisture content using a semi-empirical model based on key variables determined using a random forest regression. The model is validated against various regional datasets and shows reasonable agreement with observed total fuel load and moisture content. The model is validated against independent datasets for different regions, including Oregon, Canada, and Africa. The results show that the model provides accurate estimates of fuel load and moisture content, with reasonable agreement with observed values. The model is also validated against case studies, including wildfires in Bolivia, Canada, Portugal, and Australia, showing that the model can detect elevated fire risk based on fuel moisture content. The model provides a foundation for diagnosing landscape fire disturbance in an operational weather forecasting system and to generate a long-term record of these essential climate variables. The model output could be used for modelling other land surface processes, such as trace gas fluxes or biochemical soil properties, with the future aim of achieving full dynamical coupling of fires between the land surface and atmosphere.A global fuel characteristic model and dataset for wildfire prediction Joe R. McNorton and Francesca Di Giuseppe present a mid-complexity model combining data-driven and analytical methods to predict fuel characteristics for wildfire prediction. The model integrates data from meteorological variables, land surface models, and satellite observations to provide real-time forecasts and climate records. The dataset, with high spatiotemporal resolution (approximately 9 km, daily), is the first of its kind and will be regularly updated. The model partitions fuel load and moisture into live and dead fuels, including both wood and foliage components. An example is a 10-year dataset that is well correlated with independent data and largely explains observed fire activity globally. While dead fuel moisture correlates highest with fire activity, live fuel moisture and load are shown to potentially enhance prediction skill. The use of observation data to inform a dynamical model is a crucial first step toward disentangling the contributing factors of fuel and weather to understand fire evolution globally. The model estimates fuel load and moisture content using satellite-derived above-ground biomass (AGB) data and modelled net ecosystem exchange (NEE). AGB is estimated using the ESA Biomass Climate Change Initiative (ESA-CCI) version 3 dataset and NEE data generated by the ECLand land surface model. The model further subdivides AGB into foliage and wood mass, considering both alive and dead components. The model also estimates live and dead fuel moisture content using a semi-empirical model based on key variables determined using a random forest regression. The model is validated against various regional datasets and shows reasonable agreement with observed total fuel load and moisture content. The model is validated against independent datasets for different regions, including Oregon, Canada, and Africa. The results show that the model provides accurate estimates of fuel load and moisture content, with reasonable agreement with observed values. The model is also validated against case studies, including wildfires in Bolivia, Canada, Portugal, and Australia, showing that the model can detect elevated fire risk based on fuel moisture content. The model provides a foundation for diagnosing landscape fire disturbance in an operational weather forecasting system and to generate a long-term record of these essential climate variables. The model output could be used for modelling other land surface processes, such as trace gas fluxes or biochemical soil properties, with the future aim of achieving full dynamical coupling of fires between the land surface and atmosphere.