The Paris Agreement aims to limit global warming to 1.5 °C to avoid the worst impacts of climate change. However, existing carbon budget estimations largely neglect the crucial interplay between fire, vegetation, and carbon cycling, which are essential for understanding future ecosystem resilience. This research addresses this gap by employing a coupled fire-vegetation model to analyze regional and global impacts of warming on fire regimes and their feedback on carbon storage. The study investigates whether the 1.5 °C target is achievable without causing substantial ecosystem shifts considering the altered fire dynamics. The significance of this work lies in its comprehensive evaluation of the fire feedback mechanism, providing a more accurate assessment of the remaining carbon budget and the potential for exceeding critical thresholds in ecosystem stability. The inclusion of fire significantly alters the predictions of land carbon sinks, necessitating a reevaluation of the 1.5 °C target and strategies for climate change mitigation and adaptation. The increasing frequency and intensity of extreme weather events globally, at even the current warming level of 1.26 °C, underscore the urgency of this investigation. Existing models often lack adequate representation of fire processes and their impact, underestimating the severity of ecological and carbon cycle transformations.

Previous research indicates that fire regime and biome shifts are accelerating due to climate change, projecting increased transformations with rising temperatures. These transitions result from complex interactions involving climate, land use, and fire, impacting ecosystems through vegetation mortality, hydrological cycle alterations, and greenhouse gas emissions. Existing studies on fire risk have shown increased danger at and beyond 1.5 °C of warming; however, the impact on land carbon sinks, considering fire feedbacks, has not been fully explored. While existing carbon budget estimates from the IPCC exist, they have limitations in accurately accounting for fire's impact, highlighting the need for a more precise quantification of fire's influence on carbon sinks.

This study leverages the fire-enabled land-surface model JULES-INFERNO, which incorporates nitrogen limitation, dynamic vegetation, and fire processes. The model's performance was validated using benchmarking metrics, assessing its ability to simulate fire, tree cover, and carbon uptake. To account for climate and land-use uncertainties, four different Earth System Models (ESMs) were used, generating simulations with and without fire. The ESMs provided bias-corrected General Circulation Model (GCM) data through the ISIMIP framework, minimizing uncertainties in climate-system response to emissions. The analysis focused on projected burnt area changes, trends in tree cover, and changes in Net Biome Productivity (NBP) to assess the impacts of fire at various global warming levels (GWLs). The model evaluated multiple metrics against observational data to ensure accuracy, including spatial patterns and trends in present-day burnt area, vegetation carbon, and tree cover, ensuring the model is suitable for assessing future carbon and tree cover changes. A key methodological step involved defining an 'equivalent impact of tree cover change,' comparing the temperature at which tree cover reaches a similar level with and without fire, clarifying the influence of fire in accelerating or delaying ecosystem transitions. Statistical tests, including a Wilcoxon signed-rank test, were employed to determine when fire first significantly impacts NBP.

The study projects an increase in burnt area in most Global Fire Emissions Database (GFED) regions at 1.5 °C and 2.0 °C above pre-industrial levels, with significant increases in Europe and boreal North America being notable exceptions. Africa showed a projected decrease in burnt area. Analysis shows that at 1.5 °C GWL, climate change and CO2 fertilization lead to higher tree cover in high-latitude regions and the Congo, but a decrease in tropical South America and Asia. The inclusion of fire accelerates ecosystem change, with equivalent impacts on tree cover occurring at lower temperatures in many regions than in models without fire. This indicates that climate change’s impacts are happening earlier than previously thought. Globally, net biome productivity (NBP) is projected to increase with temperature due to high-latitude warming and CO2 fertilization, increasing the tree line and tree cover. However, fire significantly reduces NBP in most regions, except for a few where the decrease in burnt area leads to a small increase at lower GWLs. In some regions such as TENA, NHSA, and SHSA, fire can shift the system from a net carbon sink to a net carbon source, highlighting the increased risk of exceeding tipping points. A significant impact of fire on global NBP is projected at 0.8–1.34 °C above pre-industrial temperatures (central estimate 1.07 °C). Critically, the study quantifies the reduction in the remaining carbon budget due to fire, estimating a reduction of 25 Gt CO₂ for 1.5 °C and 64 Gt CO₂ for 2.0 °C above pre-industrial levels, representing 4-6% reduction in both cases. This indicates that limiting warming below 1.5 °C is even more critical than previously estimated.

The findings underscore the crucial role of fire as a feedback mechanism in the climate system, influencing both the timing and magnitude of ecosystem transitions. The study demonstrates that the previously assumed safe-unsafe boundary at 1.5 °C GWL is not as clear-cut when considering the interactive effects of fire and vegetation dynamics. The substantial reduction in the remaining carbon budget highlights the increased urgency in implementing mitigation strategies to avoid exceeding the 1.5 °C target. The regional variations in the impacts of fire emphasize the importance of localized adaptation strategies, considering the uneven distribution of regional temperature changes. The study's findings necessitate a reassessment of global temperature targets and a shift towards more comprehensive climate models that fully integrate fire feedbacks. The accelerated timing of ecosystem changes has significant implications for biodiversity and ecosystem services, requiring proactive conservation efforts.

This research significantly advances our understanding of the critical role of fire in shaping future ecosystems under climate change. The key finding is that fire weakens land carbon sinks much earlier than previously understood, at temperatures below the 1.5 °C target. This necessitates more ambitious emission reduction targets and adaptation strategies. Future research should expand the analysis using a wider range of fire models and investigate the impact of projected changes in lightning on fire occurrence. The results highlight the need for comprehensive modelling incorporating fire feedback mechanisms and adaptive management strategies for sustainable ecosystem resilience.

The study primarily uses one fire and land-surface model (JULES-INFERNO) within the ISIMIP framework, limiting the breadth of model representations. While using four bias-corrected ESMs reduces uncertainties, inter-model variability remains, influencing the precise estimates of the carbon budget reduction. Assumptions about land use and fire are based on global trends, potentially differing from regional specificities. The analysis uses global mean warming levels, which may not fully capture regional temperature variations and extreme events, potentially underrepresenting some local impacts.