Editor’s summary

Anthropogenic climate change has made wildfires bigger, hotter, and more common. Jones et al. used a machine learning approach to break down the “why” and “where” of the observed increases. The authors identified different forest ecoregions, grouped them into 12 global forest pyromes, and described their differing sensitivities to climate, humans, and vegetation. Their analysis shows how forest fire carbon emissions have increased in extratropical pyromes, where climate is the major control, overtaking emissions from the tropical pyromes, where human influence is most important. It also illustrates the increasing vulnerability of forests to fire disturbance under climate change. —Jesse Smith

Structured Abstract

INTRODUCTION

Forest fires are a natural disturbance mechanism made more likely by climate change, with major impacts on global forest ecosystems and carbon © storage. Recent trends show a worrying increase in forest fire activity, particularly in extratropical regions. This study aims to disentangle the factors driving the recent increases in fire activity by analyzing global forest fire extent and emissions and their relationship with climatic, human, and vegetation controls. Using machine learning, we grouped global forest ecoregions into 12 distinct pyromes in which forest fire extent depends on similar sets of controls.

RATIONALE

Understanding the drivers of fires in distinct pyromes is essential for developing targeted strategies to predict and manage fire risks. By grouping forest ecoregions into pyromes with distinct fire controls, we aimed to better understand the regional variations in fire dynamics and their sensitivity to climate change. This approach allows us to isolate the effects of climate change from other influencing factors such as land use and vegetation productivity.

RESULTS

Our analysis revealed that extratropical forest fire emissions have increased substantially under climate change. Fire emissions in one extratropical pyrome spanning boreal forests in Eurasia and North America nearly tripled between 2001 and 2023. This increase was linked to a rise in fire-favorable weather conditions, reduced soil moisture, and increased vegetation productivity. By contrast, tropical pyromes showed a decline in fire emissions linked to reduced deforestation fires in moist tropical forests and increased fragmentation of dry tropical forests with agriculture and other land uses. Overall, forest fire C emissions increased by 60% globally during the study period, with the most substantial contributions coming from extratropical regions. The increase in extratropical fire activity highlights the strong influence of climatic factors compared with human activities, which play a more dominant role in tropical regions. The increases in forest fire C emissions were explained both by changes in fire extent and by changes in fire severity (measured in terms of the C emitted per unit area burned by fire). In the extratropical forest pyromes, we observed major increases in fire severity alongside expansion of areas affected by fire. This finding shows that the intensity and severity of fires is increasing in extratropical forests, which is consistent with fires affecting drier, more flammable stocks of vegetation fuels as the climate warms and as droughts become more frequent.

CONCLUSION

The steep trend toward greater extratropical forest fire emissions is a warning of the growing vulnerability of forest C stocks to climate change. This poses a major challenge for global targets to tackle climate change, with fire reducing the capacity of forests to act as C sinks. Effective forest management and policies aimed at reducing greenhouse gas emissions are essential to mitigate these risks. Our study underscores the importance of considering regional distinctions in the controls on fire when developing strategies to manage fire and protect forest ecosystems. Proactive measures such as monitoring changes in vegetation and productivity can guide the prioritization of areas for forest management in the extratropics. In tropical pyromes, reducing ignitions during extreme fire-favorable weather and preventing forest fragmentation should protect forests and enhance C retention. In regions with substantial fire suppression history, shifting focus to managed, ecologically beneficial fires may prevent C sink-to-source conversion. Addressing the primary causes of climate change, particularly fossil fuel emissions, is central to minimizing future risks of forest fires globally and securing resilient forests for the future. In addition, our work supports growing calls for more comprehensive reporting of forest fire emissions to the United Nations as part of national reporting of anthropogenic C fluxes. The present norm of counting forest fire emissions fluxes as natural, on both managed and unmanaged land, is increasingly at odds with the observed growth in fire emission fluxes tied to anthropogenic climate change. This contributes to emerging gaps between the anthropogenic C budgets that are officially reported to the United Nations and the budgets constructed based on models and observations of terrestrial C stocks or atmospheric concentrations of CO2. Finally, we highlight the potential for major overestimation of C storage (and therefore C credits) by reafforestation schemes in extratropical forests if the growing risk of fire disturbance is not appropriately factored into accreditation protocols.