Forest fire size amplifies postfire land surface warming

Large forest fires have become more frequent and larger in recent decades due to climate warming, fuel management, and socio-demographic changes, with mean fire size doubling to tripling in several regions. Larger fires have been associated with higher burn severity and greater combustion, implying higher CO2 emissions per unit burnt area and stronger biogeochemical climate impacts. Yet, biophysical (surface energy balance) impacts constitute roughly half of fires’ near-surface climate effects, and it is unclear whether larger fires amplify or offset biogeophysical warming. This study focuses on northern temperate and boreal forests (40°N–70°N), where pronounced postfire surface warming has been reported, fire size and burned area are projected to increase, fire regimes vary by forest type, and wildfires are largely climate-driven. The research question is whether and how increasing fire size amplifies postfire land surface warming through biogeophysical processes, how long these effects persist, and how forest type modulates the response. The authors combine satellite-based datasets (2003–2016) of wildfire patches, land surface radiometric temperature, albedo, evapotranspiration, incoming shortwave radiation, and indicators of fire behaviour (duration, spread rate, FRP) and severity (LAI change, forest mortality) to quantify the postfire energy balance response as a function of fire size over 1–14 years after fire.

Prior work shows increased extreme fire weather linked to atmospheric warming and humidity changes and documented enlarging fire sizes in North America and elsewhere. Larger fires tend to exhibit higher burn severity, lower postfire vegetation greenness, and greater surface organic carbon combustion. Biogeochemical effects of CO2 emissions from fires account for about half of near-surface climate impacts, with the remainder from biophysical changes in albedo, ET, and energy fluxes. Pronounced postfire warming has been observed in northern forests, with species composition influencing climate feedbacks, and albedo-driven cooling observed in some Siberian larch systems. Forest fragmentation and spatial extent modulate deforestation-induced surface warming, suggesting that disturbance patch size may influence energy flux partitioning. The literature also links fire behaviour (intensity, duration, spread) to drought and fuel moisture, and highlights management strategies (e.g., increasing broadleaf share) to mitigate fire risk and warming in boreal systems.

Study domain: Northern Hemisphere temperate and boreal forests between 40°N and 70°N. Period: 2003–2016. Datasets: global wildfire patch data (Global Fire Atlas, with validations against higher-quality regional patch datasets for western Canada and Alaska), MODIS-derived surface radiometric temperature (Ts), surface albedo (α), ecosystem evapotranspiration (ET), incoming shortwave radiation, outgoing longwave radiation, and ancillary products including fire radiative power (FRP), leaf area index (LAI), and forest loss/mortality. Spatial analysis: 2° grid resolution; analyses restricted to grid cells with at least 10 fire patches. Temporal windows: seasonal (summer JJA; winter DJF) and up to 14 years postfire. Statistical models: For each grid cell, the authors fit linear regressions of the form y = a + β × log10(fire size), where y is the postfire change (Δ) in variables relative to prefire/reference conditions: ΔTs (summer), Δα (summer/winter), ΔET (summer), ΔLAI (one year postfire), and forest mortality (percentage). Field significance is assessed using false discovery rate correction (FDR = 0.10). A domain-wide regression was also fit pooling all eligible patches to derive a single β value per year since fire. Energy balance assessment: Components of radiative (shortwave absorption via albedo change; longwave emission) and non-radiative (latent, sensible, ground heat flux) processes were examined; latent heat inferred from ET; sensible+ground flux estimated via energy conservation (residual), acknowledging accumulated uncertainties. Seasonality: Separate analyses for summer and winter to contrast radiative vs non-radiative dominance, with annual effects synthesized and tracked over 1–14 years. Forest type modulation: Analyses stratified by forest types—evergreen needleleaf (ENF), deciduous needleleaf (DNF), mixed forest (MF), and deciduous broadleaf (DBF)—and alternatively by the share of broadleaf content. Multiple linear regressions included time since fire, log10(fire size), forest type, and interactions to estimate amplification slopes β across variables (βLAI, βET, βα, βT) by forest type. Fire behaviour links: Relationships between fire size and fire duration, spread rate, and FRP were quantified to infer intensity and severity scaling; ΔLAI and forest mortality changes with size were used as severity proxies. Sensitivity/robustness: Compared GFA with higher-quality regional fire patch datasets; found GFA underestimates fire size but size–warming amplification robust or stronger in regional datasets. Additional comparison to clear-cut harvest patches in Canada to distinguish direct size effects from fire-specific intensity/albedo effects.

• Postfire summer surface warming scales positively with fire size across northern temperate and boreal forests. Domain-wide amplification slope βΔT = 0.44 ± 0.01 K (log10[km²])⁻¹. Each doubling of fire size increases postfire summer land surface temperature by 0.13 ± 0.01 K, about 22% of the regional mean warming (0.60 K) one year after fire. - Both the warming (ΔT) and its amplification with fire size (βΔT) persist for up to 14 years postfire, though they decay over time. - Spatial patterns: Strongest βΔT in North America and eastern boreal Asia dominated by ENF and DNF; weaker or insignificant βΔT in regions dominated by DBF and MF. - Energy balance drivers (summer): Larger fires produce greater decreases in surface albedo (Δα more negative) and ET (ΔET more negative), increasing shortwave absorption and reducing latent cooling; outgoing longwave radiation increases with size due to warmer surfaces. Despite enhanced radiative cooling (net radiative loss), non-radiative flux changes (reduced latent, increased sensible/ground heat) dominate, yielding net surface warming. - Winter responses: Opposite-sign changes dominate (higher albedo and radiative cooling lead to ΔT < 0), with weaker dependence on size and primarily in North America; ET/latent heat changes are minimal in winter. - Annual evolution: For the first ~4 years, annual ΔTs is dominated by summer warming with size dependence; after ~4 years, winter cooling dominates annually, but radiative cooling persists up to 14 years. - Direct size effect: Larger contiguous burn patches reduce landscape heterogeneity and surface roughness relative to many small patches, lowering turbulent dissipation and enhancing warming; corroborated by comparisons with clear-cut harvest patches. The slope of warming vs patch size is 0.16–0.21 K (log10[km²])⁻¹ larger for fires than for harvests, indicating additional fire-specific intensity/albedo effects. - Fire behaviour effect: Fire size correlates positively with fire duration, spread rate, and FRP, indicating higher intensity. ΔLAI decreases and forest mortality increase with size, indicating higher severity. These co-varying behaviour effects contribute to size-dependent warming. - Forest type modulation: Fire vulnerability and size amplification vary by forest type: ENF shows the largest ΔLAI decrease and warming; followed by DNF, then MF; DBF shows minimal ΔLAI change and negligible summer warming. Corresponding amplification slopes βLAI, βET, βα, and βT follow ENF > DNF > MF > DBF. Results are consistent when using continuous broadleaf share and persist up to a decade. - Regional extension: Amplified postfire summer warming with size also found in parts of the continental USA and Australia, suggesting broader relevance beyond northern forests. - Neighboring effects: Enhanced warming following large fires also affects nearby unburnt forests, contributing to regional warming and potentially promoting drought and heatwaves. - Implications: Larger fires increase direct CO2 emissions (combustion) and legacy emissions (dead-wood decomposition); in boreal regions, enhanced soil warming promotes permafrost thaw and additional soil carbon loss, strengthening positive climate–fire feedbacks. Aerosol-induced radiative cooling adds complexity to the net climate effect.

The study demonstrates that increasing fire size systematically amplifies postfire land surface warming in summer via biogeophysical pathways. This addresses the open question of whether larger fires exacerbate or mitigate biophysical warming, showing that size magnifies non-radiative warming (reduced ET and turbulent dissipation due to lower roughness in homogeneous large patches) despite enhanced radiative cooling. The relationships between fire size and fire behaviour (duration, spread rate, FRP) indicate that larger fires tend to be more intense and severe, further decreasing albedo through charring and increasing vegetation mortality, compounding warming effects. Forest composition strongly modulates both vulnerability and amplification: conifer-dominated (ENF, DNF) systems exhibit the strongest size-dependent warming, whereas broadleaf-dominated systems show little to no amplification. Seasonality is key: summer responses dominate early postfire years, while winter radiative cooling grows over time; nevertheless, enhanced summer warming persists up to 14 years. The findings imply that as climate warming increases the occurrence and size of fires, surface energy balance changes may reinforce regional warming, reduce cloud cover, and increase drought and heatwave frequency, escalating fire risk and altering postfire regeneration trajectories (including potential regeneration failure and shifts in species composition). At larger scales, greater CO2 emissions and permafrost-driven carbon losses amplify global climate feedbacks, although aerosol-induced cooling partially offsets warming, making the net effect context dependent.

This work provides satellite-based evidence that forest fire size amplifies postfire summer land surface warming across northern temperate and boreal forests and that this amplification persists for over a decade. The amplification arises from both direct patch-size effects on surface roughness and turbulent fluxes and from co-varying fire behaviour (intensity and severity) that enhance albedo reduction and vegetation loss. Forest type strongly modulates vulnerability and amplification, with broadleaf presence dampening size-dependent warming. Management implications include prioritizing climate-smart forestry that reduces the climate risks of large fires, for example by increasing the share of broadleaf species where appropriate and avoiding active pyrophytes, using broadleaf strips as firebreaks, and integrating species selection into regeneration planning. Future research should refine estimates with improved fire patch datasets, quantify net climate impacts considering aerosol effects, test management strategies across different fire regimes (including light taiga surface-fire systems), and assess socio-ecological trade-offs where broadleaf expansion may increase spring fire risks or affect permafrost stability.

• Fire patch data (Global Fire Atlas) have documented omission errors and underestimated fire sizes in high latitudes; however, comparisons with higher-quality regional datasets showed the size–warming amplification is robust and even stronger regionally. - Energy budget diagnostics rely partly on residual calculations for sensible+ground heat, aggregating uncertainties from other fluxes. - Winter size effects and cooling are regionally variable and more evident in North America, reflecting snow/albedo dynamics; generalization to all high-latitude regions may be limited. - The study uses observational regressions; while mechanisms are physically consistent, causality among co-varying fire size, intensity, and severity cannot be fully disentangled. - Management recommendations may not generalize to all fire regimes (e.g., light taiga surface-fire systems) or socio-ecological settings where broadleaf expansion could elevate spring fire risk or affect permafrost-related ground insulation; careful local planning is required. - The analysis period (2003–2016) and 2° spatial resolution may miss fine-scale heterogeneity and very long-term recovery trajectories beyond 14 years.