Kelley et al. (2025) State of Wildfires 2024–2025

Identification

• Journal: Earth system science data
• Year: 2025
• Date: 2025-10-15
• Authors: Douglas I. Kelley, Chantelle Burton, Francesca Di Giuseppe, Matthew W. Jones, Maria Lucia Ferreira Barbosa, Esther Brambleby, Joe McNorton, Zhongwei Liu, Alexander S. Bradley, Katie Blackford, Eleanor Burke, Andrew Ciavarella, Enza Di Tomaso, Jonathan Eden, Igor José Malfetoni Ferreira, Lukas Fiedler, Andrew James Hartley, Theodore R. Keeping, Seppe Lampe, Anna Lombardi, Guilherme Mataveli, Yuquan Qu, Patrícia S. Silva, Fiona R. Spuler, Carmen B. Steinmann, Miguel Ángel Torres‐Vázquez, Renata Moura da Veiga, Dave van Wees, Jakob Benjamin Wessel, Emily M. Wright, Bibiana Bilbao, Mathieu Bourbonnais, Cong Gao, Carlos M. Di Bella, Kebonye Dintwe, Victoria M. Donovan, Sarah Harris, Elena A. Kukavskaya, Aya Brigitte N’Dri, Cristina Santín, Galia Selaya, Johan Sjöström, John T. Abatzoglou, Niels Andela, Rachel Carmenta, Emilio Chuvieco, Louis Giglio, Douglas S. Hamilton, Stijn Hantson, Sarah Meier, Mark Parrington, Mojtaba Sadegh, Jesús San-Miguel-Ayanz, Fernando Sedano, Marco Turco, Guido R. van der Werf, Sander Veraverbeke, Liana O. Anderson, Hamish Clarke, Paulo M. Fernandes, Crystal A. Kolden
• DOI: 10.5194/essd-17-5377-2025

Research Groups

• Water and Climate Science, UK Centre for Ecology and Hydrology, Wallingford, UK
• Hadley Centre, Met Office, Exeter, UK
• European Centre for Medium-Range Weather Forecasts (ECMWF), Reading and Bonn, Germany
• Tyndall Centre for Climate Change Research, School of Environmental Sciences, University of East Anglia, Norwich, UK
• National Centre for Earth Observation, University of Leicester, Leicester, UK
• Department of Life Sciences, Imperial College London, London, UK
• Centre for Agroecology, Water and Resilience, Coventry University, Coventry, UK
• Earth Observation and Geoinformatics, National Institute for Space Research (INPE), São José dos Campos, Brazil
• Institute of Oceanography, Center for Earth System Research and Sustainability, University of Hamburg, Hamburg, Germany
• International Max Planck Research School on Earth System Modelling, Max Planck Institute for Meteorology, Hamburg, Germany
• Centre for Environmental Policy, Imperial College London, London, UK
• Leverhulme Centre for Wildfires, Environment and Society, Imperial College London, London, UK
• Water and Climate, Vrije Universiteit Brussel, Brussels, Belgium
• Faculty of Science, Vrije Universiteit Amsterdam, Amsterdam, the Netherlands
• School of the Environment, Yale University, New Haven, USA
• Department of Meteorology, University of Reading, Reading, UK
• The Alan Turing Institute, London, UK
• Institute for Environmental Decisions, Department of Environmental Systems Science, ETH Zurich, Zurich, Switzerland
• Federal Office of Meteorology and Climatology MeteoSwiss, Zurich Airport, Switzerland
• Environmental Remote Sensing Research Group, Department of Geology, Geography and Environment, Universidad de Alcalá, Alcalá de Henares, Spain
• Laboratory for Environmental Satellite Applications (LASA), Department of Meteorology, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil
• BeZero Carbon Ltd, London, UK
• Department of Mathematics and Statistics, University of Exeter, Exeter, UK
• Biology, Departamento de Estudios Ambientales, Universidad Simón Bolívar, Caracas, Venezuela
• UMR Art-Dev 5281, Université Paul Valéry Montpellier, Montpellier, France
• Earth, Environmental and Geographic Sciences, University of British Columbia – Okanagan, Kelowna, Canada
• Department of Geography, University of Hong Kong, Hong Kong SAR, China
• Departamento de Métodos Cuantitativos y Sistemas de Información, University of Buenos Aires, CABA, Argentina
• IFEVA (Agricultural Physiology and Ecology Research Institute), CABA, Argentina
• Department of Environmental Science, University of Botswana, Gaborone, Botswana
• West Florida Research and Education Center, School of Forest, Fisheries, and Geomatics Sciences, University of Florida, Milton, USA
• Fire Risk, Research and Community Preparedness, Country Fire Authority, Burwood East, Australia
• Laboratory of Experimental and Applied Ecology, V.N. Sukachev Institute of Forest Siberian Branch of the Russian Academy of Sciences, Krasnoyarsk, Russian Federation
• Department of Natural Sciences, Nangui Abrogoua University, Abidjan, Côte d’Ivoire
• Research Institute of Biodiversity (IMIB), CSIC-University of Oviedo-Principality of Asturias, Mieres, Spain
• Biosciences, Swansea University, Swansea, UK
• Research and action, ECOSCONSULT, Santa Cruz de la Sierra, Bolivia
• Research, Fundacion Innova, Santa Cruz de la Sierra, Bolivia
• Fire and Safety, RISE Research institutes of Sweden, Borås, Sweden
• School of Engineering, University of California, Merced, USA
• Tyndall Centre for Climate Change Research, School of Global Development, University of East Anglia, Norwich, UK
• Department of Civil Engineering, Boise State University, Boise, USA
• Disaster Risk Management Unit (E.1), Directorate E (Space, Security, and Migration), European Commission Joint Research Centre, Brussels, Belgium
• Regional Atmospheric Modelling (MAR) Group, Regional Campus of International Excellence Campus Mare Nostrum (CEIR), Department of Geography, University of Hong Kong, Hong Kong SAR, China
• Meteorology and Air Quality, Wageningen University and Research, Wageningen, the Netherlands
• Cemaden/MCTI, São José dos Campos, Brazil
• FLARE Wildfire Research, School of Agriculture, Food and Ecosystem Sciences, University of Melbourne, Parkville, Australia
• ForestWISE – Collaborative Laboratory for Integrated Forest and Fire Management, Centre for the Research and Technology of Agro-Environmental and Biological Sciences, Universidade de Trás-os-Montes e Alto Douro, Vila Real, Portugal
• Wildfire Resilience Center, School of Engineering, University of California, Merced, USA

Short Summary

This report systematically tracks global and regional fire activity for the 2024–2025 fire season, analyzing the causes of prominent extreme wildfire events and projecting their likelihood under future climate scenarios. It found that global fire-related carbon emissions totaled 2.2 Pg C (9% above average) despite below-average global burned area, driven by extreme seasons in South America and Canada, with climate change significantly increasing the likelihood of these events.

Objective

• To regionally identify extreme individual wildfires or extreme wildfire seasons during March 2024–February 2025 and contextualize them within recent trends.
• To shortlist four "focal" extreme events with notable societal or environmental impacts for dedicated analysis.
• To globally assess the impacts of extreme wildfire events on population exposure, physical assets, carbon projects, and air quality.
• To diagnose the contributions of weather, fuel dryness, fuel load, ignitions, and suppression to the occurrence of each focal event.
• To assess the capacity of operational predictive systems to forecast the scale of fire occurrence for each focal event.
• To attribute each focal event to anthropogenic influences by testing the role of climate change and socioeconomic factors (e.g., land use, land-use change, human ignitions).
• To provide an outlook for the probability of extreme events in the early months of the 2025–2026 fire season.
• To project future changes in the probability of each focal event under future climate scenarios.

Study Configuration

• Spatial Scale: Global, continental, biome, ecoregional, national, and sub-national (level 1 administrative regions). Data resolutions range from 500 m (MODIS BA, Global Fire Atlas) to 0.1° (GFAS) and 0.25° (GFED4.1s) for observations, and 9 km (ECMWF forecasts) to 0.5° (ConFLAME, ISIMIP) for models.
• Temporal Scale: The 2024–2025 fire season (March 2024–February 2025) is the primary focus. Historical analyses span from 2002 or 2003 to 2025. Attribution studies use periods like 1960–2013 (HadGEM3-A), 2003–2019 (ISIMIP3a/FireMIP), and 1901–1917 (early industrial counterfactual). Forecasts cover short- to medium-range (1–15 days) and subseasonal to seasonal (up to 6 months ahead). Multi-decadal projections extend from 2010 to 2100.

Methodology and Data

• Models used:
  • Fire Weather Index (FWI): Canadian Forest Service metric for fire danger.
  • Probability of Fire (PoF) / Sparky fire model: Data-driven fire prediction system (XGBoost methodology) incorporating meteorological variables, fuel load/moisture, and ignition events.
  • Controlar Fogo Local Analise pela Máxima Entropia (ConFLAME): Probabilistic Bayesian framework for burned area (BA) causality analysis and attribution, optimized regionally.
  • Hadley Centre Large Ensemble (HadGEM3-A): Climate model for attributing changes in extreme fire weather to anthropogenic climate forcing.
  • Canadian Earth System Model (CanESM5): Used for attributing changes in the frequency of extreme fire weather to total climate forcing.
  • Intersectoral Impacts Model Intercomparison Project (ISIMIP3a/3b): Framework for attributing and projecting BA changes due to total climate forcing and socioeconomic factors, using multiple bias-corrected Global Climate Models (GCMs) (GFDL-ESM4, IPSL-CM6A-LR, MPI-ESM1-2-HR, MRI-ESM2-0, UKESM1-0-LL).
  • Joint UK Land Environment Simulator Earth System model (JULES-ES): Dynamic vegetation model used to simulate vegetation structure and fuel availability for future projections.
  • Global Fire Assimilation System (GFAS): Estimates fire carbon emissions by assimilating fire radiative power (FRP) observations.
  • Global Fire Emissions Database (GFED4.1s): Estimates fire carbon emissions based on burned area data and biomass productivity/fuel consumption models.
  • CLIMADA: Global risk assessment platform for estimating population and physical asset exposure to wildfires.
  • Integrated Forecasting System extended with modules of atmospheric composition (IFS-COMPO): Global model framework used by Copernicus Atmosphere Monitoring Service (CAMS) to simulate PM2.5 concentrations.
  • Random Forest (RF) algorithm: Statistical climate-fire model for forecasting BA anomalies.
• Data sources:
  • Satellite Observations:
    • NASA MODIS Burned Area Product (MCD64A1, Collection 6.1): Daily BA at 500 m resolution.
    • NASA MODIS Active Fire Products (MOD14A1, MYD14A1): Daily Fire Radiative Power (FRP) at 1 km resolution.
    • Global Fire Atlas (Andela et al., 2019b): Individual fire size, duration, speed, and direction from MODIS BA.
    • ESA Climate Change Initiative FireCCIS311: Burned area from Sentinel-3 SYN reflectance and VIIRS active fires at 300 m resolution.
    • NASA VIIRS Burned Area Product (VNP64A1 v002): Burned area from VIIRS imagery at 750 m resolution.
  • Reanalysis Data:
    • Copernicus Climate Change Service ERA5 reanalysis: FWI dataset.
    • ERA5-Land: Meteorological data for PoF model training and 12-month Standardized Precipitation Evapotranspiration Index (SPEI).
  • Socioeconomic and Impact Data:
    • Gridded Population of the World (GPW) / LitPop: Global population distribution and physical asset density.
    • Emergency Events Database (EM-DAT): Historical asset damages from wildfire events.
    • BeZero Carbon Ltd.: Project boundaries for forestry carbon offset projects in the Voluntary Carbon Market (VCM).
    • National/regional fire records, land and fire management agencies reports, disaster management reports, news reports, social media (for expert panel input).

Main Results

• Global Overview (2024–2025 Fire Season):
  • Total burned area (BA) was 3.7 × 10⁶ km², 9% below the average since 2002, ranking 16th lowest.
  • Fire-related carbon (C) emissions totaled 2.2 Pg C, 9% above average and the sixth highest on record since 2003, indicating more intense fires in carbon-rich ecosystems despite lower overall BA.
• Regional Extremes:
  • South America: Experienced an unprecedented fire season, setting a new record for C emissions (263 Tg C, 84% above average) and 120,000 km² BA (35% above average). Bolivia, Brazil, and Venezuela saw record-breaking anomalies, with Bolivia's BA 169% above average and C emissions 383% above average. Fires were characterized by extremely large, fast-spreading, and intense events.
  • North America: Second most severe fire year on record, with 194 Tg C emissions (112% above average) and 31,000 km² BA (35% above average). Canada had its second consecutive extreme fire year (46,000 km² burned, 282 Tg C emitted).
  • Africa: Overall BA was 12% below average, but the Congo Basin experienced record fire activity and C emissions, contributing to a 150% increase in primary forest loss in 2024 vs. 2023.
  • Asia & Europe: Generally experienced below-average fire seasons, but notable regional extremes occurred in Nepal (over 100 fatalities), northern India (severe haze, PM2.5 13x WHO standard), Iran (worst season since 2002), and parts of Southeast Europe (e.g., Portugal, Serbia, North Macedonia).
• Focal Event Analysis (2024–2025):
  • Northeast Amazonia (January–March 2024): Record forest BA (332% above average), severe impacts on Indigenous communities, air/water quality degradation. Extreme fire weather was 30–70 times more likely due to anthropogenic climate forcing, increasing BA by approximately 4.3 times.
  • Pantanal–Chiquitano (August–September 2024): Record BA (nearly triple average) and C emissions (6 times average). PM2.5 concentrations reached 903.2 µg m⁻³ (60 times WHO daily standard). Extreme fire weather was 4–5 times more likely due to anthropogenic climate forcing, increasing BA by 34.5 times. Socioeconomic factors also showed a very strong role, with an amplification factor exceeding 100.
  • Southern California (January 2025): Catastrophic losses from Palisades and Eaton fires: 31 fatalities, 11,750 homes destroyed, USD 140 billion in damages (USD 20–75 billion insured losses). PM2.5 levels peaked at 483 µg m⁻³. Regional BA was 25 times greater due to anthropogenic climate change, with an 89% likelihood of increased burning.
  • Congo Basin (July–August 2024): Highest recorded fire activity (28% above average), contributing to a 150% increase in primary forest loss. Extreme fire weather was 3–8 times more likely due to anthropogenic climate forcing, increasing BA by 2.69 times.
• Impact Assessments:
  • Population Exposure: Approximately 100 million people exposed globally, with highest numbers in India and Democratic Republic of the Congo (15 million each). Only 20,046 people were formally displaced (0.02% of exposed).
  • Physical Asset Exposure: Estimated USD 215 billion in physical assets exposed globally. Top countries: India (USD 44 billion), United States (USD 26 billion), China (USD 17 billion). Direct losses recorded in EM-DAT were USD 57 billion, with Southern California fires alone causing USD 53 billion.
  • Carbon Project Exposure: 18% (169 of 927) of Voluntary Carbon Market forestry projects experienced fire in 2024, a record since 2001, with 1.6% of project areas burned on average. 72% of projects experienced above-average drought.
• Predictability:
  • Weather was the dominant driver (40–70% explainability) for focal events, with fuel availability/dryness increasing in importance for severe fires (up to 40%). Human ignitions were consistently present but not a primary cause of severity (10–20%).
  • Forecasting challenges include human factors (suppression, land use) not fully captured by models, slow build-up of fuel dryness (Pantanal-Chiquitano), and high-resolution atmospheric phenomena (Southern California).
• Future Outlook (by 2100 under SSP370 - medium-high emissions):
  • Northeast Amazonia: Likelihood of 2024-scale regional BA events projected to increase by up to 57%.
  • Pantanal–Chiquitano: Likelihood of 2024-scale regional BA events projected to increase by up to 34%.
  • Congo Basin: Likelihood of 2024-scale regional BA events projected to increase by up to 50%.
  • Southern California: Future trajectory of extreme fire likelihood remains highly uncertain, with models suggesting potential decline due to CO2 fertilization effects on vegetation, but this is sensitive to vegetation model representation.
  • Strong mitigation (SSP126) can limit frequency increases to below 15% in all three tropical regions.

Contributions

This report significantly advances wildfire science by: - Introducing a new analysis of fire intensity and evaluating the dependence of extreme event identification on multiple burned area (BA) observation sources (MODIS, FireCCIS311, VIIRS VNP64A1). - Formally integrating regional expert knowledge through expert panels for event identification and characterization. - Presenting an entirely new suite of impact assessments, including population exposure, physical asset exposure, carbon project exposure, and air quality degradation. - Expanding predictability analysis to include seasonal burned area forecasts, complementing existing fire danger forecasts. - Developing a novel approach for directly attributing extreme regional BA totals and sub-regional BA extremes to specific 2024–2025 focal events using near-real-time counterfactuals and aggregating probabilities across space, representing a step-change from previous reports. - Extending forward-looking capabilities with future projections of the Fire Weather Index (FWI) at different global warming levels (1.5–4.0 °C). - Bridging the gap between event-focused real-time attribution and global process-based fire models, providing a more comprehensive and robust understanding of human influence on extreme fire activity through both climate change and socioeconomic factors. - Systematically evaluating model performance across diverse regions and highlighting limitations, particularly in representing human activities and fine-scale processes.

Funding

• UK Natural Environment Research Council (NERC) LTSM2 TerraFIRMA project and NC-International programme (grant no. NE/X006247/1)
• UK Department for Science (DSIT) Met Office Hadley Centre Climate Programme
• DSIT Innovation and Technology International Science Partnerships Fund (ISPF; UK Met Office Climate Science for Service Partnership (CSSP) Brazil)
• European Commission service contract (no. 942604) issued by the Joint Research Center
• UK NERC (NE/V01417X/1)
• UK NERC ARIES Doctoral Training Partnership (grant no. NE/S007334/1)
• FAPESP (grant nos. 2019/25701-8 and 2023/03206-0)
• Westpac Scholars Trust via a Westpac Research Fellowship
• Australian Research Council via an Industry Fellowship with the Victorian Department of Energy, Environment, and Climate Action, the Victorian Country Fire Authority and Natural Hazards Research Australia (grant no. IM240100046)
• UK Engineering and Physical Sciences Research Council (EPSRC; 2696930)
• São Paulo Research Foundation (FAPESP; 2021/07660-2 and 2020/16457-3)
• Brazilian National Council for Scientific and Technological Development (CNPq; 409531/2021-9 and 314473/2020-3)
• Spanish Ministry of Science and Innovation (MCIN; MCIN/AEI/10.13039/501100011033; ONFIRE PID2021-123193OB-I00)
• European Regional Development Fund (ERDF; project “A way of making Europe”)
• Copernicus Atmosphere Monitoring Service (operated by ECMWF on behalf of the European Commission)
• Spanish Ministry of Science, Innovation and Universities through the Ramón y Cajal (grant no. RYC2019-027115-I)
• Portuguese Foundation for Science and Technology (FCT; UID/04033/2025 and LA/P/0126/2020)
• Brazilian National Council for Scientific and Technological Development (CNPq; #443285/2023-3)
• Brazilian Institute of Environment and Renewable Natural Resources (IBAMA)/Federal University of Rio de Janeiro (UFRJ; #968711)
• Dragon Capital Chair on Biodiversity Economics
• European Research Council Consolidator Grant (grant no. 101000987)
• PRODIGY project by the BMFTR (grant no. 01LC2324)
• Fulbright Amazonia programme (sponsored by the U.S. Department of State and the Fulbright Commission in Brazil)
• Swiss Innovation Agency Innosuisse (grant no. 120.464 IP-SBM)
• State Assignment Project (grant no. FWES-2024-0040)

Citation

@article{Kelley2025State,
  author = {Kelley, Douglas I. and Burton, Chantelle and Giuseppe, Francesca Di and Jones, Matthew W. and Barbosa, Maria Lucia Ferreira and Brambleby, Esther and McNorton, Joe and Liu, Zhongwei and Bradley, Alexander S. and Blackford, Katie and Burke, Eleanor and Ciavarella, Andrew and Tomaso, Enza Di and Eden, Jonathan and Ferreira, Igor José Malfetoni and Fiedler, Lukas and Hartley, Andrew James and Keeping, Theodore R. and Lampe, Seppe and Lombardi, Anna and Mataveli, Guilherme and Qu, Yuquan and Silva, Patrícia S. and Spuler, Fiona R. and Steinmann, Carmen B. and Torres‐Vázquez, Miguel Ángel and Veiga, Renata Moura da and Wees, Dave van and Wessel, Jakob Benjamin and Wright, Emily M. and Bilbao, Bibiana and Bourbonnais, Mathieu and Gao, Cong and Bella, Carlos M. Di and Dintwe, Kebonye and Donovan, Victoria M. and Harris, Sarah and Kukavskaya, Elena A. and N’Dri, Aya Brigitte and Santín, Cristina and Selaya, Galia and Sjöström, Johan and Abatzoglou, John T. and Andela, Niels and Carmenta, Rachel and Chuvieco, Emilio and Giglio, Louis and Hamilton, Douglas S. and Hantson, Stijn and Meier, Sarah and Parrington, Mark and Sadegh, Mojtaba and San-Miguel-Ayanz, Jesús and Sedano, Fernando and Turco, Marco and Werf, Guido R. van der and Veraverbeke, Sander and Anderson, Liana O. and Clarke, Hamish and Fernandes, Paulo M. and Kolden, Crystal A.},
  title = {State of Wildfires 2024–2025},
  journal = {Earth system science data},
  year = {2025},
  doi = {10.5194/essd-17-5377-2025},
  url = {https://doi.org/10.5194/essd-17-5377-2025}
}