FireMAFS was led by Prof Martin Wooster (Kings College, London) as part of QUEST Theme 3 (Quantifying and Understanding the Earth System) project. This dataset collection contains the MODIS Land Cover Type product multiple classification schemes, which describe land cover properties derived from observations spanning a year’s input of Terra and Aqua data. The data are stored in a 10 arc minute grid. Fire was the most important disturbance agent worldwide in terms of area and variety of biomass affected, a major mechanism by which carbon is transferred from the land to the atmosphere, and a globally significant source of aerosols and many trace gas species. Despite such clear coupling between fire, climate, and vegetation, fire was not modelled as an interactive component of the climate/earth systems models of full complexity or intermediate complexity, that are used to model terrestrial ecosystem processes principally for simulating CO2 exchanges. The objective of FireMAFS was to resolve these limitations by developing a robust method to forecast fire activity (fire 'danger' indices, ignition probabilities, burnt area, fire intensity etc), via a process-based model of fire-vegetation interactions, tested, improved, and constrained. This used a state-of-the-art EO data products and driven by seasonal weather forecasts issued with many months lead-time. Much of the activity of FireMAFS was shaped by the research and technical priorities of QUESTESM (earth system model). Key activities included the progressive development of the JULES-ED and SPITFIRE submodels. Fire is now very well represented in QESM (Quest Earth System Model), making progress towards a modelling capability for fire risk forecasting in the context of global change.

FireMAFS was led by Prof Martin Wooster (Kings College, London) as part of QUEST Theme 3 (Quantifying and Understanding the Earth System) project. The objective of FireMAFS was to resolve limitations of fire modelling by developing a robust method to forecast fire activity (fire 'danger' indices, ignition probabilities, burnt area, fire intensity etc), via a process-based model of fire-vegetation interactions, tested, improved, and constrained. This used a state-of-the-art EO data products and driven by seasonal weather forecasts issued with many months lead-time. This dataset contains the MODIS Land Cover Type product multiple classification schemes, which describe land cover properties derived from observations spanning a year’s input of Terra and Aqua data. The data are stored in a 10 arc minute grid.

This dataset encompasses data produced in the study ''Seasonal Arctic sea ice forecasting with probabilistic deep learning'', published in Nature Communications. The study introduces a new Arctic sea ice forecasting AI system, IceNet, which predicts monthly-averaged sea ice probability (SIP; probability of sea ice concentration > 15%) up to 6 months ahead at 25 km resolution. The study demonstrated IceNet''s superior seasonal forecasting skill over a state-of-the-art physics-based sea ice forecasting system, ECMWF SEAS5, and a statistical benchmark. This dataset includes three types of data from the study. Firstly, IceNet''s SIP forecasts from 2012/1 - 2020/9. Secondly, the 25 neural network files underlying the IceNet model. Thirdly, CSV files of results from the study. The codebase associated with this work includes a script to download this dataset and reproduce all the paper''s figures. This dataset is supported by Wave 1 of The UKRI Strategic Priorities Fund under the EPSRC Grant EP/T001569/1, particularly the "AI for Science" theme within that grant and The Alan Turing Institute. The dataset is also supported by the NERC ACSIS project (grant NE/N018028/1).