Cascading Impacts of Wildfire Emissions on Air Quality, Human Health, and Climate Change Based on Literature Review

Abstract

Wildfires are a major source of greenhouse gases (GHGs), particulate matter (PM), and atmospheric pollutants, exerting widespread impacts on air quality, human health, and global climate. To address knowledge gaps, this study conducts a literature review of GHG emissions from wildfires across diverse ecosystems and fire regimes. The analysis quantifies emission magnitudes and compositions, evaluates their influence on regional and global climate processes, and synthesizes trends and methodological advances. Results show that the burned area is the main determinant of total emissions, with CO₂ as a robust predictor for estimated CO and CH₄, reflecting coupled emission behavior under varying combustion conditions. The Modified Combustion Efficiency (MCE) demonstrates a stronger predictive capacity for the CO/CO₂ ratio than for CH₄/CO₂, suggesting that CO/CO₂ can be predicted from MCE. Complete combustion dominates most fire events, while incomplete combustion increases the release of CO, CH₄, N₂O, and PM, contributing to tropospheric ozone formation and enhanced radiative forcing. Exposure to PM_2.5 and ozone remains a major health concern in fire-affected regions. This review provides a quantitative synthesis linking combustion efficiency and GHG co-variability, offering insights to refine emission modeling and guide climate mitigation strategies.

1. Introduction

1.1. The Global Challenge of Increasing Wildfire Occurrence

Wildfires have emerged as one of the most pressing environmental challenges of the twenty-first century, with profound consequences for ecosystems, human societies, and the global climate system. Liu et al. [1] reported that wildfire events are responsible for the global loss of approximately 15% of forest cover while contributing to a mean surface temperature increase of 0.15 K. Similarly, Lund et al. [2] warned that continued global warming may rapidly push boreal forest regions into increasingly unfamiliar fire–weather regimes, with substantial increases in fire risk projected. Supporting this, Williams and Abatzoglou [3] demonstrated robust positive relationships between fire activity and aridity across multiple timescales, ranging from interannual to millennial. Regional projections reinforced these concerns; for example, Rogers et al. [4] estimated significant increases in both the burned area (76–310%) and burn severity (29–41%) in the U.S. Pacific Northwest by the end of the twenty-first century.

The increasing incidence of forest fires is not limited to a particular region but has been documented worldwide, highlighting their impact in accelerating climate change. In central Greece, Angra and Sapountzaki [5] observed dramatic increases in fire risk since the beginning of the century. Corrêa [6] projected that Mediterranean regions may experience up to 40 additional days of high fire danger relative to the late twentieth century. In Siberia and North America’s boreal regions, unprecedented-scale and intense fire events release vast amounts of carbon each year. Lund et al. [2] estimated that Canada may experience a doubling of days with moderate-or-higher Fire Weather Index (FWI) under a 1 to 4 °C global warming scenario, consistent with earlier findings by Gillett et al. [7], who provided evidence that anthropogenic climate change has significantly influenced wildfire extent in Canada in recent decades. Furthermore, van der Werf et al. [8] identified substantial interannual variability in fire-related emissions, particularly in boreal Asia, South America, and Indonesia. Beyond their climatic impacts, the massive emissions from wildfires also alter atmospheric chemistry and air quality, linking ecological disturbances directly to human health risks.

1.2. Atmospheric Pollution and Public Health Crises

From an atmospheric perspective, wildfires represent a dominant source of pollutants, often surpassing anthropogenic emissions during active burning periods. Aguilera et al. [9] emphasized that wildfire smoke significantly deteriorates air quality levels and constitutes a critical driver of public health crises. Evidence suggests that wildfires account for at least 55% of surface-level PM_2.5 and contribute substantially to toxic particulates such as mercury and lead [10]. Elevated concentrations of PM_2.5, ozone (O₃), carbon monoxide (CO), and other hazardous gases have been documented globally during fire episodes. In Southeast Asia, biomass burning has been shown to increase surface O₃ concentrations by up to 40 ppb, with increasing PM_2.5 of between 40–120 μg/m³ (60–90%) of the daily average in the emission source regions [11]. Similarly, in Spain’s Doñana National Park, extreme wildfire events resulted in ozone spikes exceeding 480 ppbv (951 μg/m³), far above the international health standards [12].

The health impacts of wildfire smoke are increasingly evident, spanning both short- and long-term effects. Short-term exposures have been associated with acute respiratory distress, exacerbations of asthma, pneumonia, and bronchitis, as well as cardiovascular complications including heart failure [13,14]. For instance, high PM_2.5 concentrations during the Wallow Fire were linked to significant increases in emergency department visits for respiratory and cardiovascular conditions in Albuquerque, with elderly populations (65 years+) experiencing heightened vulnerability [15]. Long-term exposure contributes to the development of chronic obstructive pulmonary disease (COPD), lung cancer, hypertension, and premature mortality [16,17,18,19]. Notably, populations in fire-prone regions exposed to PM_2.5 levels exceeding 8 μg/m³ face significantly elevated cardiopulmonary disease and mortality risks. Although these atmospheric and health consequences are highly documented, current research remains fragmented, with ecological, atmospheric, and socio-economic perspectives often examined in isolation rather than through integrative approaches.

1.3. Research Gaps

Research on wildfire emissions and their impacts has been extensive; however, studies are limited across ecological, atmospheric, and socio-economic disciplines, with limited interdisciplinary synthesis. Contemporary trends in forest fire research emphasize the important integration of aerospace and earth observations, precise simulations, and interdisciplinary approaches to improve fire management and policymaking [20,21]. Methodological inconsistencies in biomass burning emission estimates driven by variations in fire classification, combustion completeness, emission factor, and satellite data resolution remain a significant source of uncertainty [22,23]. These discrepancies significantly affect the reliability of emission inventories and limit their utility for assessing environmental and climatic impacts [24,25]. Cross-ecosystem comparisons are scarce, with only a few studies systematically analyzing emission dynamics across tropical, temperate, Mediterranean, and boreal forests, which limits the potential for global generalization. Most studies emphasize immediate fire impacts, such as those during El Niño events, overlooking long-term carbon cycle effects and recovery processes [26], and changes in aerosol properties after forest fire [27]. Although descriptive reviews are available [28], such as examining research trends on fires based on the number of publications, comprehensive meta-analyses that quantitatively integrate various datasets to assess the magnitude, composition, and climate impact of fire emissions. These research gaps underscore the importance of conducting a comprehensive literature review to integrate previous findings and clarify the conceptual linkages within this field. Despite substantial progress, three key gaps remain: (1) inconsistencies in biomass burning emission estimates due to variable methods and data sources, (2) a lack of comparative analyses across tropical, boreal, peat, and Mediterranean ecosystems, and (3) limited integration of health and climate perspectives within fire emission studies.

1.4. Study Objectives and Novelty

The objectives of this review are threefold: (1) to quantify the magnitude and composition of GHG emissions (CO₂, CH₄, N₂O); (2) to assess their influence on regional and global climate dynamics, including temperature rise, radiative forcing, and feedback mechanisms; and (3) to synthesize trends, methodologies, and knowledge gaps in previous studies. The novelty of this work lies in its integrative and analytical approaches. Unlike previous narrative reviews, it systematically organizes and evaluates published literature to identify key patterns, contradictions, and research gaps. This approach yields a more reliable understanding of emission variability, atmospheric pollutants, and their ecological and health impacts.

Beyond methodological contributions, the study provides practical scientific and policy insights. By highlighting areas of consensus, critical gaps, and methodological inconsistencies, thereby guiding future investigations toward standardized measurement and broader spatial-temporal coverage. For policymakers, it identifies priority ecosystems and fire types whereby targeted interventions could deliver significant climate and health benefits, aligning with global mitigation targets.

This study strengthens the science–policy interface by bridging ecological processes with climate governance. It advances understanding of wildfire dynamics as a driver of atmospheric change while informing multiscale strategies for emission reduction, fire management, and climate adaptation. Collectively, the study offers a robust knowledge base and research compilations for more resilient and sustainable responses to wildfire impacts on air quality, health and climate change.

2. Methods

The literature review followed a structured approach adapted from established review frameworks to ensure transparency, completeness, and clarity in reporting, thereby supporting evidence-based understanding of the topic [29]. The review process involved four main stages—identification, screening, eligibility, and inclusion—to select and evaluate relevant studies (Figure 1). This structured literature review provides an integrative basis for examining wildfire emissions and their associated impacts [30].

2.1. Literature Search

The literature search included peer-reviewed journal articles in Scopus and Web of Science databases. Although numerous bibliographic databases and metrics have been reported over the past decade, Web of Science and Scopus remain the most comprehensive and authoritative sources for publication metadata and impact indicators [31]. A comprehensive list of keywords and synonyms was developed to capture all relevant studies. The final search string was constructed as follows: (“forest fire” OR “wildfire” OR “biomass burning”) AND (“climate change” OR “global warming”) AND (“greenhouse gas” OR “Emission”) AND (“model” OR “satellite” OR “remote sensing”). In addition, snowball search techniques, both backwards and forward, were applied during the first database identification.

2.2. Selection Criteria

The selection process for this literature review followed structured and transparent procedures conducted in several stages. Records were initially retrieved from two major databases, Scopus and Web of Science, with an additional 10 publications manually identified, resulting in a total of 1829 records (N). After removing duplicates, 1505 records remained for screening. The screening process was performed in several steps. First, studies published outside the period of 2010–2025 were excluded. Next, titles and abstracts were screened, narrowing the dataset to 726 potentially relevant articles. Non-English publications and non-original research formats such as book chapters, review papers, editorials, notes, and conference proceedings were also excluded. Further eligibility assessment removed articles that only discussed wildfires without reporting emissions, those in which emissions were unrelated to wildfires, and studies lacking direct relevance to the main research questions of this review. After these exclusions, 34 articles were retained for detailed analysis. Figure 1 summarizes the results of the literature search and screening process employed in this review.

2.3. Analytical Framework

The analytical framework for this study integrates the interrelated dimensions of wildfire dynamics, emission characteristics, methodological approaches, and their subsequent impacts on climate and human health, as illustrated in Figure 2. Wildfire types are categorized into four primary ecosystems, tropical, temperate, boreal, and Mediterranean, representing distinct biophysical conditions and combustion behaviors. Each fire regime contributes to emission profile variations, encompassing GHGs, air pollutants, particulate matter, and aerosols. These emissions constitute the central pathway linking wildfires to broader atmospheric and ecological processes.

The framework further emphasizes methodological diversity, incorporating in situ field measurements, laboratory analyses, remote sensing applications, and modeling techniques. These approaches provide complementary perspectives, enabling the quantification and characterization of emissions across spatial and temporal scales. The resulting emission data feed into two critical domains of impact. First, wildfire emissions affect the atmosphere and climate by contributing to ozone layer depletion, intensifying radiative forcing and surface temperature change, and altering cloud formation and precipitation patterns, thereby exacerbating global climate change. Second, in human health, as inhalation of air pollutants and fine particulates is directly associated with adverse respiratory outcomes and an increasing burden of cardiovascular diseases, leading to long-term morbidity and mortality. Combined, this framework provides a comprehensive analytical observation through which the complex interactions between wildfires, atmospheric processes, and human health can be systematically evaluated.

Following the literature selection, quantitative data were systematically compiled from each study, focusing on the extent of burned areas and the magnitude of associated emissions. Statistical regression analyses were then conducted to elucidate the relationships among key variables, particularly the correlation between fire-affected area and emission output.

3. Results and Discussion

3.1. Overview of Selected Literature

Most studies rely on Moderate Resolution Imaging Spectroradiometer (MODIS) satellite-derived burned area and active fire products, analyzed through remote sensing techniques and Geographic Information System (GIS) frameworks. As summarized in Appendix A (

Table A1. Summary of selected studies.

Reference	Region	Methodology & Data	Key Findings	Main Impact	Primary Uncertainty Sources	Dominant Forest Type/Land Cover	Recommendations
Wu et al. [39]	China	Combined MCD64A1 and MCD14ML to improve detection of small fires.	Crop residue burning was the largest source. A clear trend of increasing emissions from forest/shrubland fires (2003–2015).	Vital for modeling severe haze events in populated regions of China, linking fire activity to human exposure to extreme PM2.5.	Under-detection of small fires by MCD64A1.	Agricultural lands, Forest, Shrubland	Blend active fire and burned area products to create a more complete fire inventory.
Qiu et al. [36]	China	High-res inventory using MODIS (MCD64A1) BA & active fires (MCD14ML) with detailed land cover.	Crop residue burning was dominant (>50% of emissions). Peaks in June and October linked to agricultural cycles.	Critical for crafting targeted air pollution policies in China, directly linking emissions to specific agricultural practices.	Detection of small, agricultural fires.	Agricultural lands	Integrate data from multiple satellites and ground reports to improve detection in agricultural regions.
Shi et al. [41]	China	Used MCD64A1 BA, MODIS FRP, and active fire data (MCD14ML) for NEC.	Average annual emissions (2001–2017): e.g., CO2 (30,816.8 Gg), PM2.5 (225.3 Gg). Forest fires in Inner Mongolia and Heilongjiang explained variations.	Provides a high-resolution, multi-year inventory crucial for understanding regional carbon cycling and air pollution control.	Fuel load assessment and combustion efficiency in northeastern forests.	Temperate forest, Grasslands	Improve spatial resolution of biomass data and validate emission factors for regional applications.
Yin et al. [45]	China	Used MODIS Fire Radiative Energy (FRE) data and GlobeLand30 land cover.	Average annual emissions (2003–2017): e.g., CO2 (91.4 Tg), PM2.5 (0.51 Tg). Forest fires were the primary source (45% of CO2).	Demonstrates the significant role of biomass burning and provides an alternative FRE-based methodology for emission estimation.	FRE-to-emissions conversion factors.	Temperate Forest, Grasslands	Develop and validate region-specific conversion coefficients for different biomes.
Zhang et al. [47]	China	Used MODIS fire and BA products with NPP (MOD17A3) data.	Southwest forests showed net carbon loss, while Northeast forests showed resilience (increased NPP post-fire).	Reveals divergent regional responses to fire, highlighting the importance of a net carbon budget perspective.	Carbon cycle dynamics post-fire and NPP estimation accuracy.	Mixed forests (SW), Temperate forests (NE)	Integrate carbon emission and sequestration assessments for a net carbon budget perspective.
Song et al. [38]	China	Used MODIS BA data (2003–2015) and a bottom-up approach.	Annual average emissions: ~130 Tg CO2. Highest emissions in Yunnan, Sichuan, Inner Mongolia. Peaks in spring and autumn.	Provides a comprehensive national-scale assessment essential for understanding contribution to regional air pollution.	Emission factors for Chinese forests and fire detection accuracy.	Mixed forest	Develop a national-specific emission factor database and improve fire monitoring capabilities.
Ye et al. [48]	China	Combined Sentinel-2 and field data to quantify burning and suppression emissions.	16.5% of total GHG emissions came from suppression activities (10,498.30 t CO2e). Helicopter transport was a major contributor.	Introduces a novel, holistic perspective on the carbon cost of fire management, suggesting suppression efforts have a non-negligible climate impact.	Emissions from emergency response infrastructure are currently overlooked.	Evergreen forest	Develop a comprehensive accounting framework that includes both fire and suppression emissions.
Wu et al. [46]	China	Used MODIS (MCD64A1 and MCD14ML) data to track moving high emissions from biomass burning (2003–2014).	Emissions from heating season decreased while corn harvest season emissions increased.	Reveals shifting patterns of biomass burning in China linked to changes in agricultural practices and energy use.	Tracking the movement of emission hotspots over time.	Agricultural lands, Forests	Develop dynamic emission inventory methods that can capture spatial-temporal changes in burning patterns.
Li et al. [49]	China	Used FireCCI51 BA and a 1km aboveground biomass dataset for temperate forests.	Estimated multi-year emissions for Heilongjiang. Forest fires occurred mainly in autumn (62.78%) and spring (36.24%).	Provides a detailed inventory essential for atmospheric transport models and supports pollution control strategies in temperate regions.	Biomass fuel loading estimates and combustion efficiency in temperate forests.	Temperate forest	Improve spatial resolution of fuel load data and validate consumption models for temperate regions.
Wang et al. [50]	China	Land cover MODIS (MCD12Q1 v006), Burned Area data (MCD64A1 v006).	Forest fire emissions peaked in spring and winter.	Impact on local air quality and the global climate	Forest fire emission uncertainties stem from burned area, biomass density, burning efficiency, and emission factors	Mixed forest, Broad-leaf forest, Needle-leaf forest	Development of high-resolution regional forest fire emission inventories
Yang and Jiang [51]	China	GABAM dataset for burned area by the Chinese Academy of Science.	CO2 dominated emissions (2.25 × 104 Gg, 92.5%), followed by CO (1.13 × 103 Gg), PM10 (200.5 Gg), and PM2.5 (140.3 Gg)	Influence of fire on the local environment and policy on China’s air pollution control	This method only includes forests, shrublands, and grasslands, potentially underestimating total fire emissions	Evergreen forest	New large-scale CE methods are needed to reduce fire emission uncertainties.
Kayet et al. [56]	India	1 × 1 km gridded inventory using MODIS MCD45A1 BA and NRSC land cover.	Estimated annual average emissions for Karnataka (2000–2022), e.g., SO2 (6.67 Gg), NOx (9.48 Gg), NH3 (9.80 Gg), CO (670.12 Gg), OC (59.78 Gg), BC (5.09 Gg).	Provides a high-resolution inventory crucial for local air quality assessment, health impact studies, and targeted mitigation strategies.	Spatial allocation of emissions and fuel load estimates.	Tropical forests, Agricultural lands	Use numerical simulation models integrating climate parameters and high-resolution temporal data.
Reddy et al. [57]	India	Mapped burned area using high-resolution (56m) Resourcesat-2 AWiFS data. Long-term trends from MODIS active fires	17% of national CO2 emissions originated from Protected Areas (PAs). Dry deciduous forests contributed the most.	Sounds an alarm for biodiversity conservation, showing PAs are highly vulnerable to fires, leading to significant carbon losses.	Biomass data and combustion completeness for diverse Indian forests.	Dry Deciduous forest, Protected areas	Use high-resolution satellites (e.g., AWiFS) to monitor ecologically sensitive areas.
Saranya et al. [58]	India	Forest fire analysis AWiFS and LISSIII datasets	The mean annual carbon emission rate was 1.26 Tg CO2/yr	Rising CO2 levels negatively affect human health	-	semi-evergreen	Strategic plan control forest fires
Bartowitz et al. [55]	United States	Fire severity and area burned from the Monitoring Trends in Burn Severity database	Harvesting mature trees to prevent fire increases emissions instead of reducing them	Site-specific forest management that balances short-term protection with long-term carbon preservation is vital for mitigating climate change	Uncertainties for contemporary forest fire emissions	Temperate and Mediterranean forests	Prescribed burns reduce fire risk
Larkin et al. [34]	United States	Comparative analysis of four inventories (GFED, FINN, NEI+, EPA GHG).	Inventories varied by a factor of 10 (e.g., CO2e). NEI+ showed highest totals; GFED the lowest. Disagreement in seasonal peaks.	Highlights critical uncertainties in US emission reporting, affecting air quality management, climate policy, and carbon accounting.	Fire area detection (small/prescribed fires), fuel loading databases, modeling of deep organic soil combustion.	Boreal, Temperate, Grasslands	Standardize detection methods and improve dynamic fuel mapping.
Carvalho et al. [44]	Portugal	Used CHIMERE model with future area burned projections under IPCC SRES A2.	Projected increases in fires will lead to higher O3 and PM10 concentrations, potentially offsetting gains from emission controls.	Provides a forward-looking perspective that climate-induced fire increases may severely hamper future air quality improvements.	Projecting future fire activity and its interaction with changing atmospheric chemistry.	Mediterranean forest	Integrate dynamic fire-emission modules within climate-air quality modeling frameworks.
Martins et al. [63]	Portugal	Burnt area from the National Forest Fires Inventory	summer condition contribution to the higher measured PM10 values	Forest fire impacts on PM10 and ozone	-	Mediterranean forest	Forest fire emissions should be included in summer air quality models
Permadi & Oanh [40]	Indonesia	MODIS MCD45A1 BA & GlobCover land cover. Region-specific EFs and GWP.	Peatland and forest fires contributed 85–90% of emissions. BC was the 3rd most important warming agent (12–21% of forcing).	Reveals the disproportionate climate impact of peat fires, emphasizing the need to include short-lived climate pollutants (SLCPs) in mitigation.	Peat combustion depth and emissions factors.	Tropical peatland, Secondary forest	Develop specialized algorithms for detecting smoldering peat fires and improve peat-specific EFs.
Junpen et al. [37]	Thailand	Utilized MODIS active fire data and conducted prescribed burning experiments for EFs.	27,817 fire hotspots detected (2005–2009), peaking in March. Total burned area: 159,309 ha.	Confirms forest fires as a major source of atmospheric pollutants in Thailand, providing critical data for regional air quality management.	Emission factors and detection of small fires.	Tropical forest	Use country-specific emission factors and improve detection of understory fires.
Bondur et al. [42]	Russia	Long-term (2001–2023) analysis using MODIS (MCD64A1)BA & land cover (MCD12Q1).	Record high burned area in 2021 (~91,800 km2). A rising FRP trend linked to climate change. The Far East is a disproportionate hotspot.	Quantifies massive carbon emissions from Russian wildfires, underscoring their growing impact on the global carbon cycle.	Scaling emissions from boreal forests with high fuel loads.	Boreal forest (Taiga)	Develop and employ biome-specific emission factors and fuel consumption models for boreal regions.
Baldassarre et al. [43]	Turkey	Evaluated geostationary (SEVIRI) vs. polar-orbiting (MODIS) FRP. WRF/CMAQ simulations.	15 min data captured diurnal cycle and peak intensity missed by MODIS. LSA SAF emissions showed superior plume agreement.	Demonstrates the paramount importance of high-temporal-resolution data for accurate air quality forecasting and public health warnings.	Algorithm choice for FRP retrieval (WF_ABBA vs. LSA SAF). Temporal and vertical allocation of emissions.	Mediterranean forest	Prioritize high-temporal-resolution geostationary data for emission modeling.
Bhujel et al. [52]	Nepal	burnt-area product (MCD45A1), and Field survey	Annually, over 3158 ha of forest burns, emitting ~1108 t C (≈4066 t CO2, 2581 t CO, 1474 t CH4)	-	-	Tropical forest, Hill sall forest, Riverine forest	Forest management should combine fire lines, conservation ponds, and community capacity building
Bertolin et al. [64]	Argentina	Field study and laboratories. Burn area from Fire Program, Subsecretaría de Bosques de la Provincia de Chubut	C losses from fires were 104.6, 90.7, and 94.7 Mg C/ha across the three sites	Negative carbon balance for all three locations due to no carbon sequestration.	-	Mediterranean forests	Future work should ensure continued carbon capture, reduce disturbance losses, manage old-growth forests, and conserve forest diversity and connectivity
Scarpa et al. [65]	Italy	Fire activity data from Italian Forest Service and Satellite data for land use	Italy’s average GHG and particulate emissions were 2621 Gg/yr, ranging from 772 Gg in 2013 to 7020 Gg in 2007.	Essential for air quality management, mitigating wildfire impacts, and guiding prescribed fire decisions	Uncertainties in emission estimates	Mediterranean forests	Thoroughly assess the model and compare with field data.
Teixeira et al. [59]	Brazil	Temporal and spatial analysis of fire spot and biomass burning. Fire Inventory (FINN) model version 1.5, from the National Center for Atmospheric Research (NCAR)	Most fire events occurred in natural forests (37%), with croplands/pastures (29%) and grasslands (19%) representing the next most affected land types.	Provide information for tackling climate and health issues related to air quality	Uncertain class, which can be either agricultural or pasture areas.	Tropical and subtropical Atlantic forest	Future studies should integrate land use, human activities, and meteorology to better understand fire drivers and emissions
Bilgiç et al. [66]	Greece	Sentinel-2 imagery calculated normalized burn rate difference index (dNBR), while CORINE land cover data found burned area land cover	The largest burned areas (~50,000 ha) occurred in western Türkiye and central Greece	FINN and GFED databases mostly underestimated emissions	uncertainties in the fuel load and combustion completeness parameters	Mediterranean forests	Sentinel-2–based method for emission calculations.
Korísteková et al. [53]	Slovakia	Used a Tiered approach to estimate GHG emissions from forest fires.	The share of GHG emissions from forest fires is less than 1% of national totals. TIER 1 underestimated compared to TIER 2/3.	Highlights the small but non-negligible contribution of forest fires to national GHG budgets, emphasizing the need for accurate accounting.	Biomass available for burning and combustion factors.	Temperate mountain forest	Use more complex methods (TIER 3) for GHG emission determination, especially for larger fires.
Bar et al. [54]	Himalaya	Used MODIS MCD64A1, land cover, and biomass data over 20 years.	Eastern Himalayas (India) were the largest emission source (20.37 Tg CO2). Emissions showed high interannual variability.	Addresses a critical data gap, linking fire emissions to glacial melting (Via BC deposition) and threatening water security.	Complex terrain causing satellite detection errors; poorly constrained mountain forest biomass.	Mountain forests	Apply terrain corrections to satellite data and develop region-specific biomass maps for mountain ecosystems.
Shi & Yamaguchi [35]	Southeast Asia	Bottom-up inventory using MODIS burned area (MCD64A1) and active fire products (MOD14/MYD14).	Quantified significant annual emissions (55,388 Gg CO, 817,809 Gg CO2). Major peak in Jan-Mar (dry season).	Provides a foundational emission dataset for modeling transboundary haze pollution and its impacts on regional air quality and public health.	Relies on accuracy of MODIS burned area detection and regional average EFs.	Tropical forest, Peatlands, Agriculture	Incorporate higher-resolution data and region-specific EFs to reduce uncertainty, especially for peatlands.
Aditi et al. [60]	South Asia	Used Suomi-NPP VIIRS and MCD12Q1 land cover to develop an inventory.	Estimated annual emissions: 91.58 Tg CO2, 0.60 Tg PM2.5. Major emissions from forest fires in the HKH and Central Highlands.	Establishes forest fire as a major sector of GHG and aerosol emissions in South Asia, essential for regional climate models.	Emission factors for diverse South Asian forests.	Tropical, Subtropical, Temperate forest	Utilize VIIRS data for improved fire detection and develop region-specific emission inventories.
Chang & Song [61]	Tropical Asia	Compared two satellite BA products (L3JRC and MCD45A1) to estimate emissions.	Indonesia and India were largest contributors. MCD45A1 generally yielded lower estimates than L3JRC. Two emission peaks were identified: Feb-Mar (forest fires) and Aug-Oct (peat fires)	Highlights the sensitivity of regional emission budgets to input data selection, affecting haze prediction accuracy.	Choice of satellite burned area product (L3JRC vs. MCD45A1).	Tropical forest, Peatlands	Fuse multiple burned area products to create a consensus dataset with uncertainty bounds.
Shi et al. [62]	Tropical continents (Americas, Africa, Asia)	MODIS MCD64A1 burned area product Fire Radiative Power (FRP) and Fire Radiative Energy (FRE)	Average annual CO2 emissions: 6083.69 Tg/yr Major contributors: woody savanna/shrubland (52%), savanna/grassland (27%), forest (17%), cropland (3%), peatland (1%) Africa is the largest emitter (62%), followed by Asia (20%) and the Americas (18%) Peak emissions during August–September	Identifies spatial-temporal emission dynamics by land type and continent	Burned area detection limitations (small fires missed by MODIS) Uncertainty in AGB (~50%) and CE (20–30%) Emission factor variability by biome FRE conversion ratio error (~10–31%) Incomplete detection of small agricultural fires	Woody savanna/shrubland (Africa) Savanna/grassland (Americas) Forest (Asia) Minor: peatland (SE Asia), cropland (India, SE Asia)	Integrate small fire detection (using VIIRS or higher-resolution sensors) Improve AGB and CE estimation using ground validation Link emissions inventory with air quality and health models
Huang et al. [67]	Alaska	Conducted In Situ aircraft measurements during interior Alaska fires.	CO2 and CH4 concentrations were significantly higher near flaming fronts. BC deposition enhanced local radiative forcing.	Provides rare empirical data on immediate GHG and aerosol concentrations within plumes, crucial for validating models.	Spatial heterogeneity of plume composition and radiative effects.	Boreal forest	Conduct more in situ measurements to validate model simulations and improve RF calculations.
Lamarque et al. [32]	Global	Historical reconstruction (1850–2000) of anthropogenic and BB emissions.	Model simulations indicated underestimation of CO concentrations, particularly in the Northern Hemisphere.	Provides a foundational historical emission dataset for modeling long-term atmospheric composition changes.	Historical data accuracy and emission factor consistency over time.	Global	Improve historical emission constraints using multiple data sources and model validation.
Mieville et al. [33]	Global	Historical reconstruction using satellite products (GBA2000 burnt areas, ATSR fire hotspots) and historical data.	Global emissions stable until 1970s (~7400 Tg CO2/yr), then increased to ~9950 Tg CO2/yr. Boreal/temperate fires decreased due to suppression.	Provides a century-scale perspective on global fire emissions, highlighting the impact of human management practices.	Historical data reliability and scaling of emissions over time.	Global (Forest, Savanna)	Improve historical biomass burning reconstructions using multi-proxy data integration.

), comparative assessments highlight both advancements and persistent challenges in global fire emission estimations. Methodological approaches have evolved from coarse-resolution historical reconstructions [32,33] toward more sophisticated, high-resolution, region-specific inventories that integrate burned area (BA), fire radiative power (FRP), emission factors, and in situ validation. While this methodological diversity enhances accuracy and spatial representativeness, it simultaneously introduces significant uncertainties, with emission estimates diverging by up to an order of magnitude [34].

Wildfires emit substantial amounts of gases and particulate matter, with seasonal peaks varying across regions. In Southeast Asia, emissions typically reach their maximum during the dry season between January and March [35]. In China, two distinct peaks are observed in June and October [36], while Thailand experiences its highest emissions in March [37]. Peat-dominated fires show additional peaks between August and October. The most intense emissions are reported in Yunnan, Sichuan, and Inner Mongolia, generally occurring during spring and autumn [38].

Findings from several studies indicate that biomass burning emissions are often reported in aggregated forms, encompassing both forest fires and agricultural residue burning. However, evidence suggests that emissions from agricultural residues frequently dominate the total estimates [36,39]. Therefore, a clear distinction between forest fire emissions and agricultural residue burning is essential to improve the accuracy and comparability of emission assessments.

A recurrent source of uncertainty lies in detecting small-scale or understory fires, particularly in agricultural landscapes and tropical forests, where standard MODIS burned area products often underperform [34,35]. Peatland regions such as Indonesia and Southeast Asia present additional challenges due to variability in peat combustion depth and region-specific emission factors, underscoring the need for specialized algorithms and locally derived parameters [35,40]. Similarly, in temperate and boreal regions, uncertainties stem largely from fuel load estimates and combustion efficiency, with studies emphasizing the importance of high-resolution biomass data and biome-specific models [41,42].

Across the reviewed studies, there is a clear shift toward integrating multi-source satellite observations (e.g., MODIS, Visible Infrared Imaging Radiometer Suite (VIIRS), Advanced Wide Field Sensor (AWiFS)), with advanced numerical models used to reduce discrepancies in emission estimates. Methodological improvements include fusing multiple burned area products to create consensus datasets, refining fire radiative power to emission conversion coefficients, and incorporating dynamic fire-emission modules within climate–air quality models [43,44,45]. Nevertheless, even with these advancements, inventories often diverge by nearly an order of magnitude, reflecting the combined effects of algorithm choice, data resolution, and biome-specific uncertainties. These findings underscore the necessity of harmonizing detection techniques, developing region-specific emission factor databases, and applying multi-proxy validation strategies in order to build more reliable fire emission inventories for climate and air quality assessments.

3.2. Magnitude and Composition of Emissions

Wildfire emissions are a significant source of GHGs, particularly Carbon Dioxide (CO₂), Methane (CH₄), Nitrogen Oxides (NOx), Carbon Monoxide (CO), PM_2.5, and PM₁₀. A comprehensive literature review shows that forest and biomass burning release substantial amounts of these GHGs and particulate matter, with their composition and size varying across ecosystems and fire types. The specific types and quantities of GHG emissions are summarized in

Table 1. Mean pollutant emissions from wildfire inventories per year.

Region	Location	References	CO2 (Tg)	CH4 (Tg)	NOx (Tg)	CO (Tg)	PM2.5 (Tg)	PM10 (Tg)
Temperate	China	Qiu et al. [36]	7.69	0.021	0.015	0.331	0.048	0.068
		Song et al. [38]	1.70	0.005	0.002	0.104	0.011	-
		Wu et al. [39]	62.71	0.234	-	3.530	0.338	-
		Shi et al. [41]	2027.90	3.617	3.997	87.137	9.877	-
		Yin et al. [45]	91.41	0.236	0.231	5.02	0.506	0.567
		Wu et al. [46]	962.04	2.978	1.422	54.163	4.387	5.034
		Zhang et al. [47]	6.00	0.019	-	0.370	-	-
		Ye et al. [48]	0.05	0	0.063	0.004	-	-
		Li et al. [49]	17.20	0.064	0.040	0.907	0.129	0.184
		Wang et al. [50]	1.25	0.005	0.001	0.081	0.001	0.005
		Yang & Jiang. [51]	22.45	0.043	0.034	1.127	0.140	0.200
	Nepal	Bhujel et al. [52]	0.001	0.001	-	0.001	-	-
	Slovakia	Korísteková et al. [53]	2.44	0.01	-	0.193	0.015	0.017
	Himalaya	Bar et al. [54]	13.60	0.042	0.013	0.840	0.104	0.115
	United States	Bartowitz et al. [55]	23.11	-	-	-	-	-
		Larkin et al. [34]	65.5	0.181	0.154	-	0.403	-
Tropical	Indian	Kayet et al. [56]	7.98	0.031	0.008	0.642	0.077	0.094
		Reddy et al. [57]	98.11	0.33	0.010	5.690	-	-
		Saranya et al. [58]	2.08	0.008	0.003	0.127	-	-
	Indonesia	Permadi & Oanh [40]	57.25	0.401	0.098	7.411	-	-
		Shi and Yamaguchi [35]	54.88	0.236	0.056	3.612	-	-
	Thailand	Junpen et al. [37]	171.12	0.736	-	11.263	0.970	0.906
	Brazil	Teixeira et al. [59]	7.82	0.022	0.022	0.418	0.042	-
	South Asia	Aditi et al. [60]	91.47	0.272	-	-	0.620	-
	Tropical country	Chang & Song [61]	131.67	0.892	0.115	11.500	3.492	3.775
	Tropical Continent	Shi et al. [62]	31.06	0.095	0.070	1.605	0.224	0.322
Mediterranean	Portugal	Carvalho et al. [44]	1.01	0.004	0.003	0.062	0.068	0.074
		Martins et al. [63]	4.14	0.016	0.013	0.274	0.021	0.032
	Turkey	Baldassarre et al. [43]	-	-	0.003	0.102	0.029	-
	Argentina	Bertolin et al. [64]	0.07	0.005	-	0.026	-	-
	Italy	Scarpa et al. [65]	2.02	0.011	-	0.221	0.018	0.022
	Greece	Bilgic et al. [66]	-	-	0.014	0.396	0.033	-
Boreal	Russia	Bondur et al. [42]	184.89	0.479	-	9.014	1.077	-
	Alaska	Huang et al. [67]	59.33	0.237	1.179	5.041	-	-
-	World	Lamarque et al. [32]	-	-	54.648	332.131	-	-
-	World	Mieville et. al. [33]	8820.56	-	20.956	501	-	-

[32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67].

Global fire regimes exhibit distinct characteristics across biomes, reflecting variations in climate, vegetation, and fire drivers. In tropical regions, fires are predominantly human-induced, often linked to land clearing for agricultural activities [68] extreme drought caused by El Nino [69]. In contrast, temperate forests experience a mixture of surface and crown fires, where ignition is driven by seasonal droughts or lightning events [70,71] and post-fire recovery is often facilitated by species with adaptive traits such as resprouting or serotinous cones [72]. Boreal ecosystems, dominated by coniferous vegetation and thick organic layers, are characterized by stand-replacing crown fires of high intensity, typically triggered by lightning under extreme summer dryness; these events strongly influence carbon and nitrogen dynamics and long-term ecosystem productivity [73]. In addition, Mediterranean-type ecosystems, on the other hand, experience fast-moving, high-intensity crown fires fueled by seasonal aridity and strong winds, but vegetation is often fire-adapted through resprouting and seed germination strategies that promote rapid recovery [74]. Collectively, these contrasting fire regimes highlight how climatic context and vegetation feedbacks shape the ecological outcomes of wildfires across the world’s major biomes.

The variability of pollutants in Table 1 is governed by a complex interplay of factors encompassing fuel characteristics, environmental conditions, and combustion processes. In addition to fuel-related factors, emission variability across studies is further shaped by differences in the burned-area extent, the specific fire locations considered (e.g., national vs. subregional estimates), and the temporal range of the datasets, particularly where single-year observations are compared with multi-year averages. Fuel is a factor that influences emissions based on the type, moisture content, density, chemical composition of biomass, and combustion mode (flaming vs. smoldering) [75,76,77]. Higher fuel moisture content suppresses efficient combustion by demanding other thermal energy for vaporization, thereby elevating emissions of incomplete combustion products such as CO, CH₄, and particulate matter, as demonstrated in experimental studies contrasting flaming and smoldering combustion of various wildland fuels [78,79]. Meteorological variables, including temperature, relative humidity, wind speed, and broader climatic trends, play critical roles in shaping fire behavior, altering fuel moisture dynamics, and modulating emissions through their influence on combustion intensity and duration [80,81]. Topographic context further changes emission outcomes through effects on microclimate and fire propagation; for instance, south-facing slopes tend to be drier and more prone to intense burning, thus intensifying emission output. Combustion processes may occur completely or incompletely, leading to distinct emission profiles. Complete combustion predominantly produces CO₂, whereas incomplete combustion results in higher proportions of CO and CH₄ relative to CO₂ [82].

The regression analysis is six major emission species including CO₂, CH₄, CO, NO_x, PM_2.5, and PM₁₀ against burned area, as shown in Figure 3. The emission and burned area data were compiled from 17 peer-reviewed studies and subsequently subjected to regression analysis. Burned area consistently emerges as a key driver of total emissions, though the strength of correlation varies significantly among different gases and particulates, reflecting the role of additional ecological and atmospheric parameters.

3.2.1. Carbon Dioxide (CO₂)

CO₂ exhibits the strongest positive correlation with burned area (r = 0.95, Figure 3a), confirming that larger fire-affected regions release higher amounts of carbon due to greater biomass consumption and enhanced combustion efficiency. This indicates that the extent of burned area is a key determinant of CO₂ emissions. Previous studies have reported similar findings, with a total burned area of 12,898 km² producing 67.5 Tg of CO₂ [83] and an annual average of 2.37 Mha of burned area resulting in 79.26 ± 28.65 Tg C emissions [84]. Significantly, CO₂ emissions from wildfire events can exceed those from fossil fuel combustion. For example, the estimated carbon emissions in Canada, 2023 ranged between 570 and 727 TgC (mean: 647 TgC), equivalent to approximately 2.09–2.67 Gt CO₂, with an average of 2.37 Gt CO₂ released into the atmosphere. These values are more than four times of Canada’s annual fossil fuel emissions and substantially higher than the previous years’ average of 121 TgC (443 Mt CO₂) [85]. In the Amazon, CO₂ concentrations during fire events were also observed to be approximately two ppm higher than in surrounding areas [12] indicating an increase in CO₂ due to wildfires.

3.2.2. Carbon Monoxide (CO)

CO strongly correlates with burned area (r = 0.84, Figure 3b), with larger fires consuming more biomass. Larger fire-affected regions generate proportionally higher CO emissions, with ~70% of the variability explained by fire extent. CO accumulation is critical due to its long atmospheric lifetime and indirect effects on tropospheric ozone formation and hydroxyl radicals (OH) depletion.

In addition to its correlation with burned area, CO also demonstrates a remarkably strong linear relationship with CO₂ emissions, with r = 0.95 (Figure 4). These findings suggest that, if CO data is not available in some regions, CO emissions can be estimated using a regression equation, as this approach is comparatively more practical and easier to apply for preliminary assessment. The slope of the regression (0.0581) reflects the consistent ratio of incomplete to complete combustion across fire events, underscoring that CO₂ remains the dominant gas, yet CO is non-negligible due to its atmospheric and health implications. The observed quantitative relationship, particularly the CO/CO₂ ratio, and it needs to validate through future research.

Empirical evidence further illustrates the magnitude of these effects. In Australia, CO concentrations increased by 272% following major wildfire episodes [86], while in Santiago, Chile, CO levels rose by 50% during severe fire events [87]. Similarly, forest fires have produced extreme spikes in gaseous pollutants, with CO rising from baseline levels of 250–350 µg/m³ to nearly 100,000 µg/m³ [12]. These elevated concentrations pose immediate risks to human health and exacerbate secondary atmospheric processes, including the formation of photochemical smog and delicate particulate matter.

Beyond acute episodes, long-term analyses confirm the atmospheric significance of CO emissions. Across the Northern Hemisphere, increasing wildfire activity has been linked not only to significant rises in mean and extreme CO and aerosol optical depth (AOD) values, but also to consistent perturbations in the CO₂ and CO emission balance, which plays a role in climate feedback loops. Extreme total column CO and AOD were increased to between 9.3% and 33% from 2008 and 2023, particularly in boreal regions and the western United States [88]. This reinforces the interpretation that wildfire-driven CO emissions are not only regional pollutants but also essential tracers of combustion efficiency with cascading consequences for atmospheric chemistry and climate.

3.2.3. Methane (CH₄)

The relationship between CH₄ emissions and burned area is positive (r = 0.69, Figure 3c), yet the substantial data dispersion indicates that fire extent alone cannot fully account for CH₄ variability. This suggests that additional factors, including combustion mode, fuel characteristics, and ecosystem type, strongly influence CH₄ emissions. Zhao et al. [89] previously reported that variations in combustion efficiency, fuel moisture, and fire intensity modulate methane production independently of burned area. In addition, smoldering combustion in peat-rich ecosystems generates high CH₄ emissions as compared to flaming combustion in drier, mineral soils [90]. While the statistical relationship underscores the relevance of burned area as a primary determinant, the observed variability highlights the necessity of incorporating ecological and combustion-specific parameters when quantifying wildfire-related CH₄ emissions [91,92]. Empirical evidence supports this complexity, whereby during the 2020 California forest fire, forest fire activity accounted for approximately 13% of the state’s total CH₄ emissions, underscoring the significant role of biomass burning in regional GHGs budgets [93].

In contrast, CH₄ emissions showed a strong linear correlation with CO₂ emissions (r = 0.81, Figure 5), suggesting that CO₂ can serve as a robust predictor for estimating CH₄ release. The regression slope (0.003) indicates that for each 1 Tg of CO₂ emitted, approximately 0.004 Tg of CH₄ is released, consistent with their common origin in biomass combustion. According to Andreae & Merlet [94] and Akagi et al. [92], CH₄ emissions can be estimated from CO₂ emissions using the emission ratio, since CO₂ represents the dominant from carbon released during biomass burning and thus serves as a reliable reference for trace gas estimation. Unlike burned area, CO₂ emissions directly capture combustion processes, making them a more reliable basis for CH₄ estimation. Given CH₄’s global warming potential 28 times higher than CO₂ [95], even relatively small CH₄ fluxes derived from CO₂-based predictions contribute disproportionately to climate forcing, underscoring the need for integrated fire emission assessments.

3.2.4. Nitrogen Oxides (NO_x)

NO_x correlates strongly with burned area (r = 0.84, Figure 3d), indicating that larger fire-affected regions tend to produce proportionally higher NO_x emissions. Multiple factors, including forest type, fuel composition, land cover characteristics, and the extent of the burned area, shape the magnitude of NO_x release. Schreier et al. [96] reported that dry deciduous forests with higher nitrogen content release more NO_x per unit of fire radiative power (FRP) as compared to more humid coniferous forests, while shrub-dominated ecosystems produce higher emissions than pine forests [76]. In addition, Schreier et al. [97] demonstrated a strong relationship between NO₂ emissions and the size of burned areas, reinforcing the role of fire extent in regulating atmospheric chemistry. Beyond emission factors, wildfire events substantially increase atmospheric concentrations of reactive nitrogen species. For example, Wan et al. [98] observed that enhanced levels of CO and NO₂ following the Australian wildfires highlight the significant contribution of large-scale biomass burning to regional atmospheric composition. The findings of Singh et al. [86] indicated a 45% increase in NO_x concentrations during forest fire events in Australia. Similarly, Adame et al. [12] reported that NO₂ levels rose sharply from 25–40 to 478 µg/m³ during fire episodes. Furthermore, the variability in NO_x emissions is also strongly dependent on land cover type, as emphasized by Fredrickson et al. [99].

3.2.5. Particulate Matter (PM_2.5 and PM₁₀)

The regression analysis reveals that PM_2.5 exhibits a weak correlation with burned area (r = 0.45, Figure 3e), whereas PM₁₀ demonstrates a comparatively stronger association (r = 0.79, Figure 3f). These findings suggest that as the extent of the burned area increases, the emission of coarse particulate matter (PM₁₀) tends to rise proportionally, reflecting the predominance of larger particles generated during extensive combustion events. Regional studies support these patterns, whereby Tarín-Carrasco et al. [100] reported a significant positive correlation between burned area and PM₁₀, while Nurrohman et al. [101] projected increases in PM_2.5 emissions proportional to burned area under climate warming scenarios in Siberia. Similarly, Zhan et al. [102] observed the temporal and spatial trends of PM_2.5 linked to burned biomass in Southwest China, and Nguyen et al. [103] confirmed that coarse PM emissions scale consistently with burned area across the African landscapes. The weaker correlation of PM_2.5 compared to PM₁₀ may be attributed to faster atmospheric dispersion of fine particles and variability in combustion efficiency. These findings indicate that wildfires substantially elevate particulate matter concentrations, emphasizing the importance of monitoring PM emissions for air quality management and public health protection.

Beyond the spatial extent of a fire, combustion efficiency, typically quantified as Modified Combustion Efficiency (MCE), is a critical determinant of wildfire emission magnitudes. MCE approaches ~0.99 during pure flaming combustion, defining the characteristic of efficient, and complete combustion, but decreases to 0.65–0.85 (average ~0.80) during smoldering combustion [104]. Figure 6 demonstrates that MCE provides a more robust prediction of the CO/CO₂ (R² = 0.85) emission ratio compared to CH₄/CO₂ (R² = 0.57), as reflected by the coefficient of determination.

The CO/CO₂ ratio is inversely related to MCE, with higher ratios indicating less complete, smoldering combustion and lower MCE. In contrast, lower ratios correspond to more complete, flaming combustion and higher MCE. This trend toward lower MCE, indicative of incomplete combustion, is further supported by substantially elevated CO/CO₂ (~0.214) and CH₄/CO₂ (~0.014) ratios during smoldering phases, in contrast to lower ratios observed during efficient flaming combustion (CO/CO₂ ≈0.060; CH₄/CO₂ ≈0.004) [105].

3.3. Atmospheric Chemistry and Ozone Layer Effects

Wildfire emissions substantially modify atmospheric chemistry through the release of reactive trace gases and aerosols that interact with photochemical and radiative processes. These processes operate at two distinct levels: in the troposphere, where precursor gases and solar radiation drive ozone (O₃) formation and air-quality degradation, and in the stratosphere, where smoke plumes injected to high altitudes perturb ozone concentration and atmospheric radiative balance.

3.3.1. Tropospheric Ozone Formation

Wildfire emissions influence atmospheric radiation by releasing short-lived climate forcers and aerosols, including CO₂, black carbon (BC), and other particulate matter, which act as potent radiative forcing agents [106,107]. Tropospheric ozone concentrations increase following wildfire events, as evidenced by Doñana Natural Park, Spain, where O₃ rose by 453 ppbv (1461%). In contrast, Taiwan showed minimal change (<10%), highlighting spatial, meteorological, and seasonal variability (

Table 2. Observed changes in surface ozone concentrations before and after wildfire events.

References	Location	Year	O3 Before Fire (ppbv)	O3 After Fire (ppbv)	Difference (ppbv)	Percentage of Increased (%)
Adame et al. [12]	Doñana Natural Park, Spanyol	2017	31	484	453	1461.3
	Tropical area	2017	25	61	36	144.0
Huang et al. [113]	Indo–China	2015	39.3	48.6	9.3	23.7
	Yunnan-Guizhou	2015	45.5	52.3	6.8	14.9
	Guangdong-Guangxi	2015	30.7	32.5	1.8	5.9
	Hainan	2015	35.1	38	2.9	8.3
	Taiwan	2015	34.1	36.9	2.8	8.2
	Indo–China	2015	38.5	42.2	3.7	9.6
	Yunnan-Guizhou	2015	49	50.4	1.4	2.9
	Guangdong-Guangxi	2015	37	38.7	1.7	4.6
	Hainan	2015	39.8	41.8	2	5.0
	Taiwan	2015	36	36.7	0.7	1.9
Lapere et al. [87]	Chile	2017	55	65	10	18.2
Lei et al. [114]	Global	2005-2012	23.9	25.1	1.2	5.0
Yue et al. [115]	Projection North America	2050	40	42	2	5
	Projection Canada	2050	20	25	5	25
	Projection Alaska	2050	20	35	15	75

). Aerosols reduce net radiation at the top of the atmosphere via direct, indirect, and surface albedo effects but induce localized warming in the lower troposphere [108]. Net atmospheric radiative forcing during active fire events ranges from 10 to 14 W/m², corresponding to a warming rate of 0.002–0.005 K per day [109]. This dual effect of warming via greenhouse gases and BC and cooling via aerosol–radiation interactions demonstrates the complexity of fire-induced climate impacts [108,110]. Extreme events, such as the 2019–2020 Australian “Black Summer” fires, injected smoke containing oxygenated organic compounds into the stratosphere, perturbing gas composition and reducing total column ozone, particularly in Southern Hemisphere mid-latitudes [111,112]. Wildfire-induced alterations in atmospheric constituents, particularly ozone precursors and particulate matter, can indirectly contribute to ozone layer thinning and broader climatic effects, including elevated surface temperatures.

These interactions establish a crucial vertical linkage between tropospheric and stratospheric processes. As pyroconvective smoke columns reach altitudes exceeding the tropopause, their radiative and chemical effects extend beyond the weather-forming layer, initiating longer-term perturbations in stratospheric ozone and radiative equilibrium. These processes are further explored in the following section on stratospheric ozone perturbations.

3.3.2. Stratospheric Ozone Perturbation

Wildfire-driven pyroconvective activity can inject massive amounts of smoke and aerosols directly into the lower stratosphere, producing perturbations that persist for months and rival the climatic effects of moderate volcanic eruptions [116]. These events, commonly referred to as pyrocumulonimbus (pyroCb) injections, have been observed to elevate stratospheric aerosol optical depth (AOD), modify the radiative balance, and induce measurable ozone depletion across mid to high latitudes [117,118]. Table 3 summarizes representative cases from major wildfire events documented in the past decade that displayed perturbation in ozone.

The 2019–2020 Australian “Black Summer” fire episode stands as the most intense known stratospheric smoke injection in the satellite era, with plume altitudes reaching 30–35 km and estimated aerosol mass loads of 0.3–1.1 Tg [119,120]. The resulting smoke layer persisted for more than three months and exhibited zonal transport across the Southern Hemisphere, enhancing AOD by nearly an order of magnitude and cooling the lower stratosphere while simultaneously warming the upper troposphere through absorption of solar radiation. Subsequent analysis by Damany-Pearce et al. [121] indicated that smoke-induced radiative heating sustained the aerosol layer for 5–6 months at altitudes between 25–35 km, altering stratospheric circulation and contributing to temporary ozone reductions exceeding 2–5% at midlatitudes.

Comparable perturbations were reported following the 2017 British Columbia pyroCb event, where injection heights reached 22–23 km and stratospheric AOD increased by 0.05–0.1, persisting for 8–10 months [122,123]. These long-lived smoke layers displayed microphysical and optical properties similar to volcanic sulfate aerosols, though their composition was dominated by black carbon and organic compounds. Such smoke particles absorb ultraviolet and visible radiation, thereby modifying photolysis rates and accelerating heterogeneous reactions that can deplete ozone. Ohneiser et al. [124] reported an even longer-lasting perturbation following the 2020 Chilean pyroCb, where elevated AOD (up to 0.85) remained detectable for over 12 months, demonstrating the resilience of smoke aerosols under stable stratospheric conditions.

More recent observations in Canada (2019–2020) revealed that pyro-convective plumes penetrating to 10.9–16.5 km injected approximately 0.03–0.06 Tg of aerosol mass, with impacts on stratospheric composition persisting for about six months [125]. Although less massive than the Australian events, such injections are climatically significant at regional scales, where enhanced UV absorption and radiative heating can modulate polar vortex strength and delay ozone recovery in the Northern Hemisphere.

Overall, these findings demonstrate that large wildfires can generate significant perturbations in the stratosphere. Although the resulting impacts such as temperature anomalies and localized ozone depletion are transient, their recurrence has the potential to induce cumulative and long-lasting atmospheric consequences. These results underscore that the coupling between tropospheric fire emissions and stratospheric processes represents a critical yet underrecognized component of the Earth’s climate system, emphasizing the need for its explicit representation in future climate and chemistry–climate modeling frameworks.

Table 3. Stratospheric perturbations in ozone and aerosols associated with major wildfire events.

References	Location	Year	Injection Height (km)	Stratospheric Aerosol Mass (Tg)	Duration of Perturbation (month)
Peterson et al. [119]	Australia	2019–2020	30.0–35.0	0.3–1.1	>3
Rieger et al. [126]	Australia	2020	20.0–25.0	-	6–9
Khaykin et al. [120]	Australia	2019–2020	Up to 35.0	0.4–0.6	>3
Damany-Pearce et al. [121]	Southern midlatitudes	2020	25.0–35.0	0.8	5–6
Khaykin et al. [125]	Canada	2019–2020	10.9–16.5	0.03–0.06	6
Torres et al. [122]	British Columbia PyroCb, Canada	2017	12.0–14.0	0.18–0.35	8–10
Das et al. [123]	British Columbia PyroCb, Canada	2017	12.0–23.0	0.3	5
Ohneiser et al. [124]	Chile	2020	19.0–20.0	up to 0.85	>12

Table 3. Stratospheric perturbations in ozone and aerosols associated with major wildfire events.

References	Location	Year	Injection Height (km)	Stratospheric Aerosol Mass (Tg)	Duration of Perturbation (month)
Peterson et al. [119]	Australia	2019–2020	30.0–35.0	0.3–1.1	>3
Rieger et al. [126]	Australia	2020	20.0–25.0	-	6–9
Khaykin et al. [120]	Australia	2019–2020	Up to 35.0	0.4–0.6	>3
Damany-Pearce et al. [121]	Southern midlatitudes	2020	25.0–35.0	0.8	5–6
Khaykin et al. [125]	Canada	2019–2020	10.9–16.5	0.03–0.06	6
Torres et al. [122]	British Columbia PyroCb, Canada	2017	12.0–14.0	0.18–0.35	8–10
Das et al. [123]	British Columbia PyroCb, Canada	2017	12.0–23.0	0.3	5
Ohneiser et al. [124]	Chile	2020	19.0–20.0	up to 0.85	>12

3.4. Impact of Wildfire on Climate Change

Forest fires are significant sources of greenhouse gases, contributing significantly to climate change [127]. Beyond carbon emissions, wildfires substantially influence local and regional meteorological conditions by altering key atmospheric parameters, such as air temperature and humidity. For example, wildfire activity in Central and Eastern Siberia has been shown to affect both proximate and remote regions through smoke transport and heat fluxes [128]. Similarly, during the forest fires in Karnataka from 15–26 February 2019, near-surface temperatures increased by approximately 1.0–1.5 °C, demonstrating the direct thermal effects of biomass burning on regional climate [109].

The climatic influence of wildfires extends beyond the immediate release of emissions. Atmospheric processing of fire-derived gases and aerosols forms complex organic compounds that alter radiative fluxes, deteriorate air quality, and amplify warming potential [129]. These interactions highlight the multifaceted pathways by which wildfires influence atmospheric chemistry, radiative forcing, and feedback mechanisms linked to climate change.

Wildfires do not directly induce radiative forcing or temperature rise; these effects occur through the emissions they release. Several studies have demonstrated that fire-related emissions contribute significantly to radiative forcing and subsequent increases in temperature. For instance, the estimated total global warming potential (GWP) of warming agents from national biomass open burning in 2007 reached approximately 190 Tg CO₂-equivalents over a 20-year horizon and 95 Tg CO₂-equivalents over a 100-year horizon, while the corresponding net GWP was 110 Tg CO₂-equivalents (20-year horizon) and 73 Tg CO₂-equivalents (100-year horizon) [40]. The findings of Huang et al. [67] further highlight that greenhouse gases such as CO₂, CH₄, and N₂O not only contribute substantially to radiative forcing but also exhibit pronounced spatial variability, with a 20-year mean of 7.41 W/m² and a standard deviation of 2.87 W/m².

Table 4. Summary of environmental impacts from wildfire-related emissions.

References	Location	Results
Reddy et al. [57]	India	Seasonal temperature rose 0.61 °C marking 2012 as the 2nd warmest year since 1901 and land precipitation decreases by 0.180 ± 0.966 mm/month
Bhawar et al. [109]	India	Fire events caused 10–14 W/m2 forcing, warming rate 0.002–0.005 K/day.
Lv et al. [130]	China	Post-fire land surface temperature increased by 0.11 °C after one year.
Liu et al. [131]	Siberia	Boreal fires caused net warming (0.07–0.325 K) with summer heating and winter cooling
Huang et al. [67]	Alaska	Mean radiative forcing 7.41 ± 2.87 W/m2 with strong spatial variation
Helbig et al. [132]	North America	Surface temperature in Canadian boreal forest 2024 increased 0.27 °C (summer) and decreased 0.02 °C (winter) after fire
Tian et al. [108]	Global	Fire emissions reduced net radiation (0.565 W/m2) and surface air temperature (0.061 °C); cooling > 0.25 °C in Amazon, US, and boreal Asia
Jiang et al. [128]	Global	Aerosols induced radiative effect (20.78 W m−2), reduced rainfall, and cooled Arctic regions

summarizes recent studies quantifying the climatic impacts of wildfire emissions, including changes in radiative forcing, surface temperature, and precipitation patterns. For example, Reddy et al. [57] observed a 0.61 °C seasonal temperature rise in India during 2012 as the second warmest year since 1901, while Lv et al. [130] reported a post-fire land surface temperature increase of 0.11 °C after one year. In contrast, Tian et al. [108] identified a regional cooling effect (0.565 W/m² reduction in net radiation) in certain areas such as the Amazon and boreal Asia, highlighting the complex and region-dependent nature of fire–climate interactions.

Similarly, Bondur et al. [42] observed that the annual mean fire radiative power increase across Russia is likely linked to climate change, as evidenced by rising temperatures and the growing prevalence of arid conditions. Carvalho et al. [44] further demonstrated that climate change alone can substantially affect atmospheric pollutant levels in Portugal, particularly in July and August, as projected summer temperature increases may strongly accelerate the kinetic rates of atmospheric chemical cycles. Moreover, Lv et al. [130] reported that land surface temperatures in the affected areas increased by 0.11 °C one year after a forest fire occurrence. Overall, the impact of wildfire extend beyond the immediate emissions released during the events and may persist over the long term. Such impacts, including rising temperatures, can exacerbate the likelihood and intensity of subsequent fire occurrences, creating a reinforcing feedback loop. These observations underscore the critical importance of implementing effective fire management strategies to mitigate future fire risks.

3.5. Public Health Impacts

The health consequences of wildfire smoke are increasingly documented, with exposure linked to respiratory infections, cardiovascular complications, and premature mortality. Table 5 presents evidence from selected studies that quantified health outcomes associated with wildfire exposure. Studies in developed countries such as the United States and Australia show consistent associations between wildfire exposure and adverse health outcomes. Pye et al. [133] reported that the oxidative potential of wildfire-derived PM, particularly from Reactive Organic Compounds (ROCs), constitutes the primary driver of toxicity. Wen et al. [134] further explained that health impacts are disproportionately distributed, with low-income populations experiencing heightened vulnerability due to limited adaptive capacity and poor healthcare access.

In contrast, studies in Thailand, such as Pongpiachan et al. [135], found relatively lower associations between biomass burning and polycyclic aromatic hydrocarbons (PAHs)-related health burdens. This variability may be due to differences in combustion efficiency, local biomass composition, and existing background air pollution. However, in Indonesia, an estimated 648 premature deaths occur each year, equivalent to 26 deaths per 100,000 population, mainly driven by elevated concentrations of PM_2.5, which are linked to chronic respiratory disorders, cardiovascular diseases, and lung cancer [135,136].

Table 5. Summary of human health effects associated with wildfire emissions.

References	Location	Results
Wen et al. [134]	New South Wales, Australia	Higher health risk in low socioeconomic status and high fire-density areas.
Pongpiachan et al. [135]	Northern Thailand	Minor role of PAHs from biomass burning in local health effects
Tarín-Carrasco et al. [100]	Portugal	35% increase in respiratory mortality during fire years.
Barbosa et al. [137]	Portugal	High wildfire emissions in 2017 coincided with increased mortality and economic loss.
Schroeder et al. [138]	Brazil	Strong Spearman correlation (r = 0.66) between fire events and respiratory death.
Pye et al. [135]	Western USA	Emissions: 1250 g ROC/kg CO; particulate phase drives health burden.
Maji et al. [136]	Southeastern USA (10 states)	PM2.5 increased by 10% (prescribed) and 22% (extensive burns); mortality linked.

Table 5. Summary of human health effects associated with wildfire emissions.

References	Location	Results
Wen et al. [134]	New South Wales, Australia	Higher health risk in low socioeconomic status and high fire-density areas.
Pongpiachan et al. [135]	Northern Thailand	Minor role of PAHs from biomass burning in local health effects
Tarín-Carrasco et al. [100]	Portugal	35% increase in respiratory mortality during fire years.
Barbosa et al. [137]	Portugal	High wildfire emissions in 2017 coincided with increased mortality and economic loss.
Schroeder et al. [138]	Brazil	Strong Spearman correlation (r = 0.66) between fire events and respiratory death.
Pye et al. [135]	Western USA	Emissions: 1250 g ROC/kg CO; particulate phase drives health burden.
Maji et al. [136]	Southeastern USA (10 states)	PM2.5 increased by 10% (prescribed) and 22% (extensive burns); mortality linked.

Evidence from Brazil [138] and Portugal [137] suggests that repeated exposure during peak fire years is linked to mortality, long-term health system strain, and economic losses. These findings reinforce the need to incorporate wildfire smoke as a formal determinant of public health within national disease surveillance systems. Wildfires represent a significant public health concern, as they can contribute to human mortality both directly and indirectly. This is primarily due to inhaling toxic gases and delicate particulate matter released during combustion. Such exposures are often unrecognized in real time but may lead to severe respiratory and cardiovascular outcomes. Consequently, effective fire prevention strategies are essential to minimize the release of hazardous emissions and protect human health.

4. Overall Discussion

Wildfires are among the most significant environmental challenges globally, with complex implications for atmospheric composition, climate change, and public health. Based on the literature review, a consistent finding across multiple studies reported that wildfires predominantly occur under complete combustion conditions. Complete combustion is characterized by sufficient oxygen availability and low fuel moisture, producing efficient biomass burning. This combustion pattern is evident in emission profiles, with CO₂ being produced in substantially greater quantities than CO and CH₄. The predominance of CO₂ over other gases is not merely a reflection of fire intensity but also indicative of anthropogenic influences. Several studies suggest that many fire events are intentional, often associated with land management practices such as pre-drying biomass to reduce moisture content, thereby increasing combustion efficiency. Such practices led to a disproportionate reduction in CO emissions while elevating CO₂ outputs, reflecting the optimized flaming combustion conditions.

The magnitude of emissions from wildfires is closely linked to the burned area. The regression analysis indicates that emissions of CO₂, CO, NO_x, and PM₁₀ can be reliably estimated using burn area as a predictor through regression modeling. Empirical evidence demonstrates a strong positive correlation between the spatial extent of fires and the volume of CO₂ emissions. This indicates that, at least for CO₂, fire size is a reliable predictor of emission magnitude. Across various forest ecosystems—including shrublands, savannas, tropical forests, and temperate forests larger burned areas consistently yield higher CO₂ emissions. In contrast, CH₄ and PM_2.5 emissions demonstrate low correlation with burn area and therefore cannot be accurately estimated using this approach. It is critical to note that other pollutants, particularly CH₄ and CO, are highly sensitive to the type of vegetation, the combustion phase, and fuel characteristics. For instance, smoldering fires in peatlands may generate significant CH₄ and CO emissions even over relatively small areas [139], highlighting the need to consider fire type and fuel composition in emission assessments.

Forest type and combustion completeness are additional factors that significantly modulate emission profiles. Dense tropical forests with high biomass density generally produce larger absolute emissions than shrublands or savannas [140]. However, despite their smaller per-event emissions, fires in savannas and grasslands occur more frequently and may contribute substantially to regional emissions. The completeness of combustion also influences the relative proportions of emitted gases. Flaming, complete combustion maximizes CO₂ output, while smoldering combustion increases CO, CH₄, and aerosols emissions. Therefore, accurate characterization of combustion type is essential for reliable emission modeling and understanding the broader climatic and environmental consequences.

Methodological factors, particularly the resolution of satellite imagery and the emission factors applied, further determine the reliability of emission estimates. Satellite data are commonly used to map burned areas and assess fire activity; however, coarse-resolution data can underestimate burned area, especially in heterogeneous landscapes with fragmented vegetation. High-resolution satellite products provide more accurate spatial delineation but are often limited by temporal frequency and cost considerations. Similarly, emission factors, which translate burned biomass into estimated emissions of specific pollutants, vary widely across forest types, fuel moisture contents, and combustion conditions. Many studies rely on generalized emission factors derived from limited laboratory or regional experiments, which may not accurately represent diverse ecological and climatic contexts. Inaccuracies in these factors can introduce substantial biases into regional and global emission inventories, impacting climate modeling and policy recommendations.

Wildfire emissions have profound environmental consequences. Indirectly, the release of greenhouse gases—including CO₂, CH₄, and N₂O accelerates global warming. Reactive compounds such as NO_x and VOCs contribute to the formation of tropospheric ozone, which not only acts as a greenhouse gas but also has detrimental effects on vegetation and human health. The deposition of black carbon and aerosols on snow and ice surfaces reduces albedo, further amplifying regional warming through positive feedback mechanisms. In addition, wildfire emissions may contribute to stratospheric ozone depletion, though their role is secondary compared to anthropogenic halogenated compounds. Nonetheless, the contribution of fire-related gases to ozone chemistry cannot be ignored, particularly in regions with frequent large-scale fires.

The direct impacts of wildfire emissions on human health are equally critical. Exposure to delicate particulate matter (PM_2.5) is strongly associated with respiratory diseases such as asthma, chronic obstructive pulmonary disease (COPD), and lung cancer [141,142]. Cardiovascular health is also adversely affected due to systemic inflammation induced by pollutants. Gaseous pollutants, including CO, can impair oxygen transport and exacerbate hypoxic conditions, posing acute risks during exposure [143]. Wildfire smoke, consisting of fine particulate matter (PM_2.5), gases (CO, NO_x, O₃), and polycyclic aromatic hydrocarbons (PAHs), penetrates deep into the respiratory tract, inducing oxidative stress and inflammatory responses that can impair pulmonary function and exacerbate respiratory diseases. Ultrafine particles can traverse the alveolar–capillary barrier, enter the bloodstream, and provoke systemic inflammation, endothelial dysfunction, and cardiovascular stress, particularly in vulnerable populations such as children, the elderly, and individuals with pre-existing cardiopulmonary conditions. Epidemiological analyses estimate that landscape fire smoke causes approximately 677,745 premature deaths annually, with nearly 39% occurring in children under five [144]. Short-term exposure accounts for an estimated 54–240 premature mortalities per year, whereas long-term exposure may contribute to 570–2500 premature deaths annually [145]. The toxic constituents of wildfire smoke—including PM_2.5, volatile organic compounds (VOCs), O₃, endotoxins, microbial agents, and other hazardous gases—disrupt the respiratory tract’s immune and barrier functions, enhancing susceptibility to oxidative stress, airway inflammation, allergies, asthma, and chronic obstructive pulmonary disease [146]. Smoke plumes from fires can travel long distances, affecting populations far from the fire origin, making wildfire emissions a transboundary air pollution concern. Vulnerable populations—including children, the elderly, and those with pre-existing medical conditions—are disproportionately affected. This dual environmental and public health impact underscores the importance of comprehensive fire monitoring and management strategies.

The reviewed studies also reveals that additional contextual factors influence emission magnitude, including climatic conditions and seasonal variations. Climatic regions, rainfall patterns, temperature, and humidity affect fuel availability, moisture content, and fire behavior, thereby modulating combustion efficiency and emission profiles. For instance, dry seasons with prolonged droughts increase the likelihood of large, high-intensity fires with higher CO₂ emissions. At the same time, wet conditions may limit fire spread and favor smoldering combustion with higher CH₄ and CO emissions. Despite this, many studies fail to report detailed climatic classifications or to integrate regional climatic data into their emission analyses.

This study has several limitations. First, the final dataset comprised only 34 studies, which may not fully capture the diversity of ecosystems and fire regimes globally. Second, methodological inconsistencies across studies, particularly in emission factor selection, satellite detection capacity, and combustion completeness assumptions have introduced uncertainties in aggregated estimates. For instance, small-scale and understory fires are often underdetected by Moderate Resolution Imaging Spectroradiometer (MODIS) products, leading to systematic underestimation of burned area and emissions. Third, the scope of the meta-analysis was limited to articles indexed in Scopus and Web of Science, potentially excluding valuable regional datasets and gray literature. Fourth, cross-ecosystem comparisons remain constrained by uneven geographical representation, with tropical peatland fires in Southeast Asia disproportionately studied compared to Mediterranean or African savanna fires. Fifth, health impact data were largely derived from short-term exposure studies, with limited longitudinal analyses linking wildfire smoke exposure to chronic outcomes such as cancer and neurological disorders. Finally, while this review integrates ecological, atmospheric, and health perspectives, complex feedback mechanisms (e.g., land use change, fire suppression carbon costs, adaptive capacity of communities) were beyond the analytical scope. These limitations should be addressed in future research by expanding datasets, employing high-resolution remote sensing, harmonizing emission factor databases, and incorporating long-term epidemiological studies.

Future research should prioritize direct, field-based emission measurements that capture real-time fluxes of key gases such as CO₂, CO, and CH₄. These measurements would improve the understanding of combustion completeness under diverse environmental and management conditions and refine emission factor estimates. Additionally, integrating high-resolution remote sensing with detailed field verification of forest type, fuel load, and land cover would enhance the accuracy and robustness of emission assessments. This approach would allow for more precise mapping of burned areas and more accurate modeling of pollutant outputs across heterogeneous landscapes. Interdisciplinary research linking atmospheric science, ecology, and public health is essential to fully capture the direct and indirect consequences of wildfire emissions. By combining emission quantification with climate modeling and epidemiological assessments, researchers can generate comprehensive insights into the broader environmental and societal impacts of wildfires. Finally, efforts should be dedicated to standardize reporting protocols for burned area, fire intensity, hotspot counts, and climatic context to improve cross-study comparability. Adoption of harmonized methodologies would facilitate meta-analyses and enhance the reliability of global fire emission inventories.

5. Conclusions

Overall, this systematic review provides an integrated understanding of how wildfire emissions interact with atmospheric processes, influence climate dynamics, and affect human health across ecosystems. Synthesizing evidence from 34 peer-reviewed studies, it comprehensively evaluates GHG emissions from wildfires, emphasizing their magnitude, composition, climatic influence, and methodological evolution, and the following results were found.

• The findings demonstrate that carbon dioxide (CO₂) dominates total fire-related emissions and shows strong predictive correlations with carbon monoxide (CO) and methane (CH₄), highlighting their interdependence under varying combustion regimes. The Modified Combustion Efficiency (MCE) is identified as a key determinant of emission composition, with higher MCE values yielding complete combustion and higher CO₂ outputs, whereas lower MCE values increase incomplete combustion products such as CO, CH₄, and particulate matter (PM_2.5 and PM₁₀).
• These emissions exert measurable influences on both regional and global climate systems. Fire-related GHGs and aerosols intensify radiative forcing, increase near-surface air temperatures, and promote the formation of tropospheric and stratospheric ozone, collectively amplifying feedback mechanisms that accelerate global warming and elevate future fire risks. At the atmospheric–biospheric interface, the deposition of black carbon and aerosols reduces albedo, contributing further to localized warming and altering carbon cycle dynamics. Beyond climate implications, fire emissions substantially deteriorate air quality and human health. Exposure to fine particulates and ozone is consistently associated with increased respiratory and cardiovascular morbidity and mortality, highlighting the intersection between ecological disturbance and public health crises.
• Despite notable advances, methodological inconsistencies remain a key limitation across studies. Disparities in emission factor selection, satellite detection capability, and ecosystem representation contribute to significant uncertainty in global emission inventories. Cross-ecosystem analyses, particularly involving peat, tropical, and boreal fires, remain underrepresented. To address these gaps, future research should prioritize harmonization of emission measurement techniques, the integration of high-resolution remote sensing with in situ validation, and the expansion of standardized emission factor databases across biomes.
• Interdisciplinary frameworks linking atmospheric science, ecology, and epidemiology are essential to refine emission modeling, strengthen climate projections, and guide evidence-based mitigation and adaptation policies. Moreover, this review advances scientific understanding of the coupled nature of combustion efficiency, emission variability, and climate feedback, while providing a foundation for more reliable global emission inventories and informed policy interventions in fire-prone ecosystems.