Fire Smoke Elevated the Carbonaceous PM2.5 Concentration and Mortality Burden in the Contiguous U.S. and Southern Canada

Fire Smoke Elevated the Carbonaceous PM2.5 Concentration and Mortality Burden in the Contiguous U.S. and Southern Canada | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Fire Smoke Elevated the Carbonaceous PM 2.5 Concentration and Mortality Burden in the Contiguous U.S. and Southern Canada Zhihao Jin, Gonzalo A. Ferrada, Danlu Zhang, Noah Scovronick, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5478994/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Despite emerging evidence on the health impacts of fine particulate matter (PM 2.5 ) from wildland fire smoke, the specific effects of PM 2.5 composition on health outcomes remain uncertain. We developed a three-level, chemical transport model-based framework to estimate daily full-coverage concentrations of smoke-derived carbonaceous PM 2.5 , specifically Organic Carbon (OC) and Elemental Carbon (EC), at a 1 km 2 spatial resolution from 2002 to 2019 across the contiguous U.S. (CONUS) and Southern Canada (SC). Cross-validation demonstrated that the framework performed well at both the daily and monthly levels. Modeling results indicated that increases in wildland fire smoke have offset approximately one-third of the improvements in background air quality. In recent years, wildland fire smoke has become more frequent and carbonaceous PM 2.5 concentrations have intensified, especially in the Western CONUS and Southwestern Canada. Smoke exposure is also occurring earlier throughout the year, leading to more population being exposed. We estimated that long-term exposure to fire smoke carbonaceous PM 2.5 is responsible for 7,462 and 259 non-accidental deaths annually in the CONUS and SC, respectively, with associated annual monetized damage of 68.4 billion USD for the CONUS and 1.97 billion CAD for SC. The Southeastern CONUS, where prescribed fires are prevalent, contributed most to these health impacts and monetized damages. Given the challenges posed by climate change for managing prescribed and wildland fires, our findings offer critical insights to inform policy development and assess future health burdens associated with fire smoke exposure. fire smoke PM2.5 mortality carbonaceous PM2.5 wildland fire prescribed fire Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Over the past half-century, wildland fire activity has significantly increased, not only in the U.S. but also in other temperate and high-latitude ecosystems, including Canada and Europe ( 1 , 2 ). Notably, human-induced climate change was responsible for an additional 4.2 million hectares of forest fire area between 1984 and 2015, nearly doubling the area expected to be affected by fire in the absence of climate change ( 3 ). As a result of climate change, large-scale wildland fire events have become more frequent and intense, and fire seasons have lengthened in the Contiguous U.S. (CONUS) in recent decades. Previous research has indicated that wildland fire smoke have accounted for nearly 25% of the ambient fine particulate matter (PM 2.5 , particles with a diameter of less than 2.5 µm) across the U.S. in recent years, and up to 50% in certain Western U.S. regions ( 4 ). One of the primary wildfire management strategies is prescribed burning. Prescribed fires not only reduce the biomass available for subsequent wildfires, but they also support carbon sequestration, facilitate ecological resilience, and play a critical role in restoring fire-adapted ecosystems that have been degraded due to decades of fire exclusion ( 5 , 6 ). Over 65% of the prescribed burn areas are in the Southeastern U.S ( 7 ). This imbalance in the application of prescribed fires has resulted in comparable regional average PM emissions from prescribed fires in the Southeastern U.S. and wildfires in the Western U.S. ( 8 ). In the context of climate change, as the use of prescribed burning is increasing to control wildfires, smoke from these burns is becoming a growing public health concern, particularly in the Southeastern U.S. ( 9 , 10 ). The National Prescribed Fire Acts (116th and 118th Congress) emphasizes the importance of public health and safety risks associated with the expanded use of prescribed fires. However, it states that smoke from prescribed fires is generally less harmful and of shorter duration compared to wildfire smoke, stating that it exposes children to fewer adverse health effects. Such a statement, however, is based on limited research, which may lead to an underestimation of prescribed burning's health risks. With anthropogenic climate change playing an increasingly critical role in escalating wildfire activity, the negative impacts of smoke on air quality and public health are likely to worsen in the future ( 11 ). Fire smoke contains considerable amount of PM 2.5 , significantly deteriorating the air quality in downwind communities that are tens to hundreds of kilometers away ( 12 ). Smoke PM 2.5 is characterized by substantial concentrations of carbonaceous matter, including organic carbon (OC) and elemental carbon (EC), which are produced by the combustion and incomplete burning of organic materials such as wood, leaves, and other vegetation. This distinguishes fire smoke PM 2.5 from typical ambient PM 2.5 , which tends to present greater oxidative potential ( 13 – 15 ). The unique characteristics of fire smoke PM 2.5 may alter the composition of regional PM 2.5 and potentially affect its toxicity. While numerous studies have linked exposure to PM 2.5 with various adverse health impacts ( 16 – 19 ), epidemiological research linking exposure to fire smoke PM 2.5 with adverse health outcomes is still in its early stage. Long-term exposure to smoke PM 2.5 has been linked to all-cause mortality in the CONUS, particularly among vulnerable populations such as the elderly ( 20 ). It is estimated that 11,415 non-accidental deaths per year in the CONUS can be attributed to smoke PM 2.5 , with cardiovascular diseases contributing the most ( 20 ). Short-term exposure to wildfire smoke PM 2.5 has been associated with increased risks of respiratory morbidity, mental health issues, and excess mortality, ( 21 – 24 ). Evidence on the health effects of different chemical components of smoke PM 2.5 is sparse. For example, OC has been identified to be an important component influencing PM 2.5 toxicity to several reactions harming organic systems and a key contributor to all-cause mortality ( 25 – 27 ). EC, due to its small size, can penetrate deeply into the respiratory tract and serve as a transporter for various toxic substances ( 28 ). Research on the health effects of smoke PM 2.5 has been hindered owing to the scarcity of long-term exposure data, especially data with comprehensive spatial coverage and high spatial-temporal resolution. Most epidemiological studies on smoke PM have relied on local ground-based monitoring stations, satellite images, uncalibrated chemical transport model simulations (CTM) or simple classifications of smoke-affected areas to investigate the health impacts of fire smoke ( 29 – 34 ). These methods were either based on coarse resolution smoke estimates or did not quantify smoke PM, potentially introducing exposure misclassification. Emerging research has shown great promise to generate long-term and high-resolution smoke-derived PM 2.5 concentrations by calibrating CTM simulations. For instance, Cleland et al. tested the model performance of predicting 1 km 2 wildfire smoke PM 2.5 based on CTMs simulations and different combinations of concentration datasets ( 35 ). The model that fused ground-based observations, satellite aerosol optical depth (AOD)-derived concentrations and CTMs simulations provided the best estimate (R 2 = 0.71) in fire-impacted regions, highlighting the importance of integrating multiple datasets. Similarly, Zhang et al. developed CMAQ-based models to estimate daily 1 km 2 smoke PM 2.5 total mass, which achieved strong model performance with R 2 of 0.75 and 0.68 in smoke-impacted regions and non-smoke regions, respectively ( 36 ). Nevertheless, few studies have adopted CTM-based models to estimate smoke PM 2.5 speciation with high spatial and temporal resolution. This is largely because CTM simulations for PM 2.5 speciation often face higher uncertainties compared to those for total PM 2.5 mass, demanding more advanced calibration techniques ( 37 , 38 ). In this study, we developed a three-level, CTM-based modeling framework to estimate daily concentrations of fire smoke carbonaceous PM 2.5 , specifically OC and EC, at 1 km 2 spatial resolution from 2002 to 2019 with full coverage across the CONUS and Southern Canada (SC). This framework integrated information from CMAQ simulations of PM 2.5 mass and speciation, ground-based observations and multiple auxiliary spatial and spatiotemporal datasets. This innovative approach allows us to fill important research gaps described above, namely, to differentiate exposure by specific carbonaceous constituents of smoke PM 2.5 , and to estimate fire smoke-related health burdens over the long-term. By leveraging the high spatial and temporal resolution of our model predictions, we then analyzed the spatiotemporal patterns of smoke impact frequency and concentrations of carbonaceous PM 2.5 from fires smoke. Furthermore, we estimated the populations exposed to fire smoke. Lastly, we investigated the impacts of long-term exposure to fire smoke on mortality burden and associated monetized damages. Results Model performance The CV results revealed strong model performance ( Table S1-S6 ). For daily-level predictions, Smoke-off models showed higher accuracy, with random CV R² values above 0.75 for base learners and 0.77 for meta-learners. Meta-learners for EC achieved average random CV R² values of 0.80 in Smoke-off and 0.71 in Smoke-on scenarios, while OC performance dropped in Smoke-on with a reduced R² of 0.67 and increased RMSE of 1.20 µg/m³. Spatial CV showed low prediction errors for OC and EC in Smoke-off ( Fig. S1 ), with RMSEs below 0.8 µg/m³ and 0.2 µg/m³, respectively. In contrast, Smoke-on scenarios exhibited relatively higher RMSEs, especially in areas prone to fire smoke. For temporal CV, R 2 values were consistently higher for Smoke-off scenarios ( Fig. S2 ), ranging between 0.65 and 0.85. In Smoke-on scenarios, OC and EC displayed fluctuate R² values between 0.58 and 0.65, with notable declines in 2002. At monthly and annual levels, model performance improved significantly ( Table S1-S6 ). Adjustments to residuals through GAMs further enhanced accuracy ( Table S7 ), with random CV R² values surpassing 0.91 for monthly and 0.97 for annual predictions. Most overestimations and underestimations observed at the daily level were reduced by averaging at the monthly level ( Fig. S3 ). Overall, the RASL model demonstrates reliable daily predictions across various CV experiments and further improvements in adjusting the long-term prediction residuals. Feature importance analysis ( Fig. S4 ) indicated CMAQ simulations of carbonaceous PM 2.5 were primary predictors, with MAIAC AOD, urbanization factors, meteorological factors, and spatial and temporal characteristics also influential. Spatial and temporal patterns of smoke carbonaceous PM Figure S5 illustrates the average number of smoke days per year across our study domain. The central-south (i.e., Missouri, Arkansas, and Oklahoma) and southeastern (Alabama, Georgia, and Florida) regions of the CONUS exhibited the highest average number of smoke days (200 + days per year). The Western CONUS (i.e., California, Oregon, Idaho, and Montana), North-Central CONUS (i.e., North Dakota, South Dakota, and Minnesota) and adjacent areas in SC such as British Columbia also experienced a significant number of smoke days (~ 150 days), though the sources of smoke in these regions differ. In the Western CONUS and SC, the primary source of smoke is wildfires. Conversely, in the central-south and southeastern regions of the CONUS, prescribed fires are the main source of smoke ( 39 ). The northeastern regions of the study domain are less frequently impacted by smoke (~ 100 days) but can still experience significant smoke pollution from long-range transport ( 40 , 41 ). We summarized the background, total, and smoke-specific concentrations of PM 2.5 OC and EC across different time scales and climate regions in Fig. S6 . From 2002 to 2019, both the CONUS and SC experienced a declining trend in background carbonaceous PM 2.5, with annual concentrations falling below 0.95 µg/m 3 for OC and 0.20 µg/m 3 for EC by 2019. When comparing the background carbonaceous PM 2.5 between the periods 2002–2010 and 2011–2019, improvements in annual background PM 2.5 were observed, with reductions of 0.27 µg/m³ for the CONUS and 0.17 µg/m³ for SC after 2011. The CONUS climate regions of SouthEast, South, Central, and NorthEast exhibited higher background concentrations and more significant reductions over the years (Fig. 1 ). Additionally, urban areas displayed higher background concentrations of carbonaceous PM 2.5 ( Fig. S7 ), with long-term average concentrations reaching approximately 2 µg/m 3 for OC and 1 µg/m 3 for EC. Elevated background PM 2.5 OC levels were also common in many rural and forested areas of the South, Central and SouthEast CONUS climate regions, while high levels of PM 2.5 EC were mostly concentrated in urban centers across the study domain. Considering the impacts of fire smoke, we observed a distinct contribution of smoke PM 2.5 OC, whereas the contribution of smoke PM 2.5 EC was smaller and showed less fluctuation over the years. The mean annual concentration of smoke PM 2.5 OC was 0.22 µg/m³ for the CONUS and 0.14 µg/m³ for SC, while smoke EC concentrations were 0.08 µg/m³ for the CONUS and 0.05 µg/m³ for SC. Combining smoke PM 2.5 OC and EC, smoke carbonaceous PM 2.5 accounted for 19% and 16% of the total concentrations in the CONUS and SC, respectively. At the monthly level, smoke carbonaceous PM 2.5 exhibited a notable increase during the peak months of wildfire season (July–November) ( Fig. S8 ), with average monthly concentrations rising to 2.38 µg/m³ for the CONUS and 2.60 µg/m³ for SC, accounting for 58% and 57% of total concentrations, respectively. The Western and Southern CONUS climate regions (i.e., NorthWest, West, West North Central, South, and SouthEast) and the Southwestern Canada experienced more fire smoke impacts, including wildland fire and prescribed fire, resulting in increased smoke carbonaceous PM 2.5 at both regional and national scales (Fig. 1 ). Additionally, the annual mean concentrations of smoke carbonaceous PM 2.5 increased by 0.08 µg/m³ for the CONUS and 0.05 µg/m³ for SC before and after 2011. This intensifying trend in fire smoke has offset nearly one-third of the improvements in background concentration. Megafire years of 2012, 2015, 2017, and 2018 experienced significantly higher concentrations of both smoke PM 2.5 OC and EC at the monthly and annual levels. Our 1 km 2 prediction maps revealed different spatial distributions between long-term smoke OC and EC concentrations ( Fig. S7 ). High smoke OC concentrations (> 0.50 µg/m³) were primarily observed in rural areas of the Western CONUS and SC, and were sporadically distributed in the Southeastern CONUS. Elevated EC concentrations (> 0.15 µg/m³) were often collocated with high smoke OC, with the Southeastern CONUS, particularly Georgia, Florida, Mississippi, and Texas, exhibiting even higher smoke EC concentrations (> 0.20 µg/m³). Long-term average concentrations of smoke carbonaceous PM 2.5 were highest in the Western CONUS and Southwestern Canada, particularly in California, Idaho, and Montana (Fig. 2 ). The Southeastern CONUS also experienced comparable concentrations of smoke carbonaceous PM 2.5, with urban centers such as Charlotte and Atlanta, and their surrounding areas, showing higher exposure levels than urban areas in other parts of the study domain. During megafire years (2012, 2015, 2017, and 2018), elevated smoke carbonaceous PM 2.5 were concentrated in rural regions of the Western CONUS, as well as urban centers such as Los Angeles, San Francisco, Seattle and Vancouver, as captured in the annual prediction maps ( Fig. S9 and Fig. S10 ). Increasing impact of wildland fire smoke on air quality and population exposure We evaluated the populations affected by heavy fire smoke in the CONUS and SC. A “heavy fire smoke grid-day” was defined as a grid-day when the total carbonaceous PM 2.5 concentration exceeded 1 µg/m³ and the smoke carbonaceous PM 2.5 constituted more than 50% of the total concentration. We compared the cumulative daily populations affected by heavy fire smoke during the periods 2002–2010, 2011–2019, and specifically the five years from 2015–2019 (Fig. 3 ) . Our analysis revealed a clear increasing trend in population exposed over the years, corresponding to the broadening of wildfire smoke impacts. During 2002–2010, there was an average of 467 million person-days of exposure to heavy fire smoke prior to July 1st, while this number rose by 70% to 793 million person-days during 2011–2019. By year’s end, cumulative exposure increased from 986 million person-days in 2002–2010 to 1.8 billion person-days in 2011–2019, representing an 83% increase. To account for the influence of population growth on these trends, we also calculated the annual exposure days per capita. On average, individuals experienced 3.0 days of heavy smoke exposure per year in 2002–2010, which increased to 5.1 days per year in 2011–2019. The megafire years of 2017 and 2018 had particularly severe impacts, with cumulative exposure exceeding 3 billion person-days (Fig. 3 ), equivalent to 9.0 exposure days per person per year. Separate figures for cumulative person-days of exposure to heavy fire smoke in the CONUS and SC across years are provided in Fig. S11 . Excess mortality due to smoke exposure The annual non-accidental mortality rate and total deaths attributable to smoke carbonaceous PM 2.5 from 2003 to 2020 are mapped at the county level in the CONUS and the census division (CD) level in SC ( Fig. S12) . Consistent with the spatial distribution of high smoke carbonaceous PM 2.5 concentrations in the Western and Southeastern CONUS and Southwestern Canada, counties and CDs in these areas exhibited elevated annual mortality rates, exceeding 3 deaths per 100,000 people ( Fig. S12A ). The CONUS counties with higher annual death counts (> 3 deaths per year) were generally located in California, Oregon, Washington, Florida, Georgia, South Carolina, North Carolina, and Texas ( Fig. S12B ), areas characterized by high populations density, frequent smoke impact and elevated smoke carbonaceous PM 2.5 . On average, 7,462 non-accidental deaths per year in the CONUS were attributable to long-term exposure to fire smoke carbonaceous PM 2.5 ( Table S8 ), with the southeastern CONUS contributing 4,702 deaths per year (63.0%), and the western region contributing 1,568 deaths per year (21.0%). In SC, higher annual deaths were observed in urban areas of British Columbia, Ontario, and Quebec ( Fig. S12B ), with an average of 259 non-accidental deaths per year ( Table S9 ). Annual total deaths in the southeastern region remained relatively stable over the study period (Fig. 4 ), whereas deaths in the western and northeastern regions, as well as in SC, fluctuated significantly in response to variations in wildfire intensity. The average monetized damages associated with these mortality estimates were approximately 68.4 billion USD per year for the CONUS and 1.97 billion CAD per year for SC ( Table S8 and Table S9 ). The southeastern region contributed 43.0 billion USD annually, and the western region contributed 14.2 billion USD. In 2018 and 2019, both mortality and monetized damages nearly doubled compared to the average levels, leading to monetized damages exceeding 120 billion USD for the CONUS and 4.6 billion CAD for SC (Fig. 4 ). Discussion To the best of our knowledge, this is the first study to model full-coverage concentration of fire smoke-derived carbonaceous PM 2.5 with high spatial and temporal resolution across both the CONUS and SC. We developed a three-level RASL framework to estimate both daily and long-term smoke carbonaceous PM 2.5 from 2002 to 2019. Our analysis identified frequent smoke impact and elevated concentrations of smoke-derived carbonaceous PM 2.5 in the Western and Southeastern CONUS, as well as in SC. Over the past decade, wildfire seasons have started earlier and intensified, resulting in increased population exposure to wildfire smoke, particularly during recent megafire years. We estimated that long-term exposure to smoke carbonaceous PM 2.5 resulted in an average of 7,462 and 259 non-accidental deaths per year in the CONUS and SC, respectively, with the Southeastern CONUS contributing the most deaths. The long-term concentration of PM 2.5 mass from fire smoke in the CONUS and Canada has been examined in previous literature. Our findings on the spatial and temporal patterns of smoke carbonaceous PM 2.5 are consistent with these prior studies. For example, Childs et al. estimated smoke PM 2.5 over the CONUS at a 10 km resolution and observed increased smoke pollution and smoke-impacted days over the last decade, especially in the Western U.S. and in the years 2017 and 2018 ( 42 ). The FireWork model provided real-time forecasts of biomass burning PM 2.5 across North America, showing that most wildfire events were concentrated in the Western CONUS, as well as in Western, Northern, and Central Canada ( 43 ). While previous studies have offered valuable insights into wildfire smoke PM 2.5 , our use of daily 1 km 2 resolution and focus on carbonaceous PM 2.5 provides a more precise accounting of fire smoke impacts and a more detailed understanding of its composition. Notably, our study revealed that the frequent smoke days and elevated concentration in the Southeastern CONUS are consistent over time due to the frequent use of prescribed burns, which has not been extensively discussed in prior wildfire smoke modeling studies or government datasets. For example, shortcomings were identified in the current prescribed fire permit databases given that some states in the Southeastern U.S. do not require prescribed burn permits and rely on voluntary reporting, such as Texas, Arkansas, and Missouri ( 39 ). Despite these limitations, our study successfully captured the elevated smoke days and smoke carbonaceous PM 2.5 in these states. Additionally, our results indicated a significant intensification of wildfires with higher concentrations of smoke carbonaceous PM 2.5 across our study period, which has offset nearly one-third of the improvements in background air quality across the CONUS and SC, largely due to efforts such as the Clean Air Act ( 44 ). This intensifying trend in wildfire smoke stagnated or even reversed the declining trend of background concentrations in most regions. Our findings align with existing literature on smoke PM 2.5 mass. Burke et al. reported that areas affected by wildfire smoke have doubled over the past two decades in the CONUS, and that wildfire smoke has influenced the average annual PM 2.5 trend in nearly three-quarters of the states in the CONUS since 2016, accounting for 25% of the improvement in air quality ( 4 , 45 ). However, few studies have investigated the wildfire smoke impact on PM 2.5 concentration levels in Canada. Our results provide insights into wildfire smoke dynamics in Canada and the transboundary smoke effects. Nonetheless, additional research encompassing Alaska, Northern Canada, and Mexico is needed to achieve a more comprehensive understanding of the wildfire smoke activity across North America. In addition to the increased smoke concentration, our findings indicate that exposure to wildfire smoke is becoming more common, implying more frequent wildfire smoke impact, wider smoke-impacted areas, and earlier and longer wildfire seasons. Consequently, the duration of exposure has lengthened, increasing from an average of 3.0 exposure days per year exposed to heavy fire smoke during 2002–2010 to 5.1 days per year during 2011–2019 for population across our study domain. Climate change plays a pivotal role in these shifts, as rising temperatures and prolonged droughts have shifted season regimes and created more favorable conditions for wildfires by drying out vegetation and extending the wildfire season ( 46 – 51 ). These conditions also facilitate megafires becoming more common, leading to extremely high concentrations of carbonaceous PM 2.5 . Our study exhibited significantly higher concentration levels and wider population exposure during megafire years such as 2017 and 2018, leading to increased health burden. Without significant mitigation efforts, future climate models predict an alarming increase in wildfire frequency and severity, which poses further risks to ecosystems, air quality, and public health ( 52 , 53 )​. Prescribed fires are widely recognized as one of the most effective ways to prevent potential wildfires and sustain biodiversity in all regions beyond the Southeastern CONUS ( 54 – 56 ). Our study has revealed lower concentrations of smoke carbonaceous PM 2.5 in the Southeastern CONUS during megafire years compared to the Western CONUS. However, our analysis also indicates that prescribed fires do not necessarily translate into better air quality and may lead to consistent smoke pollution with elevated carbonaceous PM 2.5 concentration over time, causing adverse health effects from long-term exposure. In the Southeastern CONUS region, where prescribed fires are the primary sources of smoke, we estimated 4,702 attributable deaths per year, which is higher than the combined deaths in the Western CONUS (1,568 deaths per year) and Northeastern CONUS (1,192 deaths per year) regions. Several factors contribute to this outcome. First, unlike the Western U.S., where wildfires occur sporadically but at high intensity, the Southeastern CONUS experiences regular smoke pollution from frequent prescribed fires. Our results indicated that annual regional average concentrations of smoke carbonaceous PM 2.5 in the Southeast CONUS are comparable to those in the Western CONUS (~ 0.4 µg/m³). While prescribed fires effectively reduce wildfires risks, they also result in frequent and localized smoke pollution in the Southeastern CONUS, where prescribed burns are conducted throughout the year ( 39 ). Second, prescribed fire smoke frequently affects densely populated urban and suburban areas in the Southeast CONUS, such as Atlanta and Charlotte, even though the fires are smaller and controlled ( 39 ). In contrast, wildfires in the Western CONUS typically occur in more remote, forested areas, such as the Cascades and Rocky Mountains. While wildfire smoke can travel long distances to urban centers like Los Angeles, San Francisco, and even the Northeastern CONUS, the primary impact is often concentrated in less densely populated areas. Third, residents of Southeastern states such as Georgia and Florida are more accustomed to prescribed fires ( 57 ). Engebretson et al. reported that Southern-state residents demonstrate significantly higher tolerance of potential health impacts from prescribed fires compared to those in Western states ( 58 ). While this acceptance reduces social barriers to the use of prescribed fires, it also lowers public vigilance regarding exposure to fire smoke pollutants. In contrast, the perception of wildfire risk in the Western U.S. has been heightened by the prevalence of megafires and media coverage, leading to a higher public awareness of the dangers posed by wildfire smoke and harm-reduction behaviors. The average monetized damages associated with the attributable deaths in the Southeastern U.S. is 43.0 billion USD per year, and this cost for the CONUS has exceeded 120 billion USD in recent megafire years. In contrast, the U.S. allocated $ 1.73 billion to wildland fire management in 2024, with $ 214.5 million dedicated to fuels management ( 59 ). Considerable evidence in the scientific literature supports prescribed fire as a cheap and effective method for mitigating wildfire risk and reducing carbon emissions ( 60 – 62 ). However, most of these studies overlook the significant health impacts associated with smoke exposure from prescribed fire, which can be transported to nearby populated areas. When considering health-related costs, the cost-effectiveness of prescribed burns is called into question, as these fires can cause damage over 200 times greater than the budget. To better inform policy, it is crucial to develop a more accurate and comprehensive cost-benefit assessment for prescribed burns by incorporating health-related monetized damages from smoke exposure. Additionally, policies should prioritize minimizing human smoke exposure by improving monitoring networks of both PM 2.5 mass and carbonaceous components and preparing communities for potential health impacts. Enhancing communication strategies to warn residents and provide resources, such as air quality alerts and protective equipment, should be an essential component of these policies. Beyond policy improvements, a more sophisticated prescribed fire management system is necessary — one that considers the conditions of each fire, including risk factors such as weather and proximity to populations, and the long-term benefits of prescribed burns in reducing the severity and frequency of wildfires ( 63 ). Achieving this balance between prescribed fire and public health is essential to ensure that prescribed burns remain a valuable tool for ecosystem health and wildfire prevention. Our study has several implications. First, it provides a high-resolution fire smoke product with full coverage for the CONUS and SC, offering insights into the occurrence and distribution of fire smoke impacts, and quantitative estimates of background and smoke carbonaceous PM 2.5 concentrations. The comprehensive spatial and temporal coverage of our predictions will enable future research on the health and environmental impacts of exposure to altered PM 2.5 composition by fire smoke. Second, our findings highlight that wildland fires have intensified over the past decade, leading to an increase in deaths associated with long-term exposure to smoke carbonaceous PM 2.5 . Finally, by combining the prescribed fire permit databases with our study’s high-resolution smoke predictions, future efforts could better track prescribed fire activities in terms of locations, durations, sizes and transmissions. As climate change continues to challenge wildfire risk mitigation and biodiversity conservation, our study underscores the importance of incorporating potential health impacts in the cost-benefit analysis of prescribed fire policies and management tools. Several limitations of our study should be noted. First, there is currently no research specifically investigating the mortality attributable to smoke carbonaceous PM 2.5 . As a result, our study relies on mortality risk estimates based on total smoke PM 2.5 . As carbonaceous PM 2.5 is an important component influencing PM 2.5 toxicity and a key contributor to all-cause mortality, using the carbonaceous fraction as a proxy for total smoke PM 2.5 may lead to an underestimation of non-accidental deaths and monetized damages attributable to smoke carbonaceous PM 2.5 . Future research focusing on the mortality risk associated with specific components of smoke PM 2.5 is needed to more accurately estimate the impact of carbonaceous PM 2.5 from fire smoke. Second, the annual smoke-mortality relationships for both the CONUS and SC applied in our study may not be entirely applicable to the 2020 baseline mortality rate, which was impacted by the COVID pandemic. In conclusion, our study highlights the growing impact of fire smoke carbonaceous PM 2.5 and its adverse effects on public health across the CONUS and SC. With wildfires intensifying and becoming more frequent due to climate change, our findings underscore the urgent need for comprehensive prescribed fire management strategies that balance ecological benefits with the reduction of smoke-related health risks. This work provides a valuable foundation for future research and policymaking to address the dual challenges of wildfire prevention and public health protection in an increasingly fire-prone environment. Materials and Methods Study domain and Period Our study domain included the CONUS and SC (Fig. 5 ). The daily predictions were developed at a 1 km 2 spatial resolution. In total, our modeling grid included 9,115,328 1 km 2 grid cells for the CONUS and 2,620,640 1 km 2 grid cells for SC, from 2002 to 2019. Ground-based Carbonaceous PM Measurements The ground-based measurements of carbonaceous PM 2.5 (i.e., OC and EC) in the CONUS were obtained from a variety of sources, including the chemical speciation network (CSN) from Environmental Protection Agency (EPA), the National Park Service interagency monitoring of protected visual environments (IMPROVE) network, and the Southeastern Aerosol Research and Characterization Study (SEARCH) network ( 64 , 65 ). For SC, ground-based observations were provided by the National Air Pollution Surveillance (NAPS) program ( 66 ). In total, 401 monitors operated during our study period, with most stations operating for more than 5 years (Fig. 5 ). Extremely high observations (> 99.98th percentile) of PM 2.5 OC and EC were excluded from the model training to minimize the outliers (99.98th percentile of OC and EC: 34.75 µg/m 3 and 7.40 µg/m 3 ). PM 2.5 OC and EC stations are often co-located and take samples synchronously, with similar number of stations and observations overtime ( Table S10 ). Our final data includes approximately 448,000 daily measurements for OC or EC. Detailed summary statistics of ground-based measurements for training dataset are provided in Table S11 and Table S12 . Chemical Transport Model The state-of-the-art Community Multiscale Air Quality (CMAQ; www.epa.gov/cmaq ) modeling system estimates atmospheric concentrations of numerous chemicals and aerosols, including ozone (and its precursors), PM 2.5 , and deposition of harmful chemical species ( 67 ). CMAQ is developed and maintained by the US Environmental Protection Agency (EPA), and it has been widely used for assessing air pollution, including evaluating policies and estimating impacts on human health ( 68 – 70 ). In this study, we used CMAQ version 5.3.2 and the datasets from the EPA’s air QUAlity TimE Series (EQUATES; www.epa.gov/cmaq/equates ) project ( 71 ). In this study, we used two sets of daily simulations at 12 km 2 spatial resolution for the period of 2002–2019: ( 1 ) the baseline CMAQ simulation results from EQUATES, which includes all emission sources, and ( 2 ) a sensitivity simulation conducted using the same configurations ( Table S13 ) and inputs as in the EQUATES project, except by not including fire emissions. Hereafter, these two sets of simulations are referred to as “Smoke-on” and “Smoke-off”, respectively. Classification of Smoke Impacted Areas To better account for the specific contribution of fire smoke to PM 2.5 OC and EC concentrations, we utilized CMAQ simulations run under the different settings. The grid cells in the study domain were classified into two daily smoke scenarios: a background scenario without fire smoke impact (“Smoke-off”) and a fire smoke-impacted scenario (“Smoke-on”). The Smoke-on CMAQ model simulations were used in modeling carbonaceous PM 2.5 under Smoke-on scenarios, while the Smoke-off CMAQ model simulations were used for Smoke-off scenarios. Detailed classification methods are provided in the Supplementary Document S1 . Auxiliary Predictors To enhance model performance and predictive accuracy, we incorporated a wide range of auxiliary predictors in model development. These predictors included satellite-retrieved aerosol data products, cloud coverage, and smoke plume information, gridded meteorological factors, vegetation coverage, biogenic emissions, population, land cover, road density, topographic data, human footprint, and coordinate and time trend characteristics. These variables have been found to be important predictors in prior studies ( 36 , 42 , 72 , 73 ). All predictors at different spatial resolutions were integrated into the 1 km 2 grid cells obtained from the Multi-Angle Implementation of Atmospheric Correction (MAIAC) dataset, which served as the grid template for the MAIAC aerosol optical depth (AOD) measurements ( 74 ). Daily ground-based measurements of PM 2.5 OC and EC were assigned to their collocated grid cells. Detailed descriptions of the data sources and process steps are provided in the Supplementary Document S2 . Modeling framework After aggregating the raw measurements for OC and EC, we applied the Synthetic Minority Oversampling Technique (SMOTE) to oversample the underrepresented high-concentration measurements ( 75 ). The enriched training datasets were then classified into Smoke-off and Smoke-on scenarios. For both PM 2.5 OC and EC, and both smoke scenarios (Smoke-off and Smoke-on), we employed the proposed Residual Adjusted Super Learner (RASL) framework (Fig. 6 ). This framework operates with a three-level structure based on the super learner method, in which the cross-validated predictions from four base models (first level) were incorporated using a meta-learner algorithm (second level) to generate fused predictions of carbonaceous PM 2.5 ( 76 ). Finally, generalized additive models (GAMs; third level) were used to adjust the spatiotemporal residuals of monthly concentrations at each 1km grid cell from 2002 to 2019 across the study domain. Details of the modeling framework are provided in the Supplementary Document S3 . Model Performance Evaluation We conducted a three-stage cross-validation (CV) to evaluate the model performance at each level of the RASL framework. Three types of CV were employed: 10-fold random CV, 10-fold clustered spatial CV, and leave-one-year-out temporal CV. The clustered spatial CV better tests the model’s predictive ability when a large group of monitoring networks is missing instead of a single monitor ( 77 , 78 ). Prediction accuracy was evaluated using three metrics: the coefficient of determination (R²), root-mean-square error (RMSE), and slope. To further evaluate the model at different time scales, we averaged daily estimations into monthly and annual values, allowing us to capture both long-term trends and short-term fluctuations—important for long-term cohort studies and short-term analyses. Details of the CV experiments are provided in the Supplementary Document S4 . Calculation of mortality burden and monetized damage We employed different methods in the CONUS and SC to calculate the non-accidental mortality and monetized damages attributable to fire smoke carbonaceous PM 2.5 . For the CONUS, Ma et al provided the monthly non-accidental mortality rates for different smoke PM 2.5 concentration bins in the CONUS ( 20 ), which we then multiplied by 12 months to estimate the annual mortality rate ( Table S14 ). The annual mortality rate, along with each year’s smoke concentrations and population data, was used to estimate the following year’s deaths. To assess the monetized damages associated with these mortality estimates, we employed the Value of Statistical Life (VSL), as provided by the U.S. Department of Health and Human Services. VSLs reflect the monetary value that individuals are willing to pay to reduce the risk of death, thereby providing an economic perspective on mortality burden. We based our estimates on the 2013 VSL value and then adjusted it annually from 2003 to 2020 in accordance with HHS guidelines ( Table S15 ), to account for inflation and changes in real income for the specific dollar year ( 79 ). The year-specific VSL values were multiplied by the estimated mortality attributable to fire smoke carbonaceous PM 2.5 exposure in that year to provide an estimate of annual monetized damages. For SC, we applied the Air Quality Benefits Assessment Tool (AQBAT) developed by Health Canada, which is designed to estimate the human health impacts and economic valuation of changes in Canada’s ambient air quality ( 80 ). AQBAT provided the concentration response functions between chronic exposure to fire smoke PM 2.5 and mortality and, which enabled us to estimate the number of attributable mortality and related monetized damages across Canada ( Table S9 ). Declarations Acknowledgements Research reported in this publication was supported by the National Institute of Environmental Health Sciences of the National Institutes of Health under Award Number 1R01ES034175. Dr. K. Chen received support from the National Heart, Lung, and Blood Institute of the National Institutes of Health (R01HL169171). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. Author Contributions: Z.J. conducted the model development, predictions, and drafted the manuscript. G.F. and J.F. conducted the CMAQ simulations. D.Z. contributed to the dataset preparation. N.S. contributed to the monetized damages estimate and edited the manuscript. K.C. provided the mortality rate results. Y.L. conceived of and supervised the conduct of this study and edited the manuscript. Competing Interest Statement: The authors declare no conflict of interests. References Fernandez-Anez N et al (2021) Current wildland fire patterns and challenges in Europe: A synthesis of national perspectives. Air Soil Water Res 14:11786221211028185 Senande-Rivera M, Insua-Costa D, Miguez-Macho G (2022) Spatial and temporal expansion of global wildland fire activity in response to climate change. Nat Commun 13:1208 Abatzoglou JT, Williams AP (2016) Impact of anthropogenic climate change on wildfire across western US forests. Proceedings of the National Academy of Sciences 113, 11770–11775 Burke M et al (2021) The changing risk and burden of wildfire in the United States. Proceedings of the National Academy of Sciences 118, e2011048118 Fernandes PM (2015) Empirical support for the use of prescribed burning as a fuel treatment. Curr Forestry Rep 1:118–127 Fowler C, Konopik E (2007) The history of fire in the southern United States. Hum Ecol Rev, 165–176 Melvin M (2018) 2018 National prescribed fire use survey report. Coalition of Prescribed Fire Councils, Inc., Newton, GA) Brey SJ, Barnes EA, Pierce JR, Wiedinmyer C, Fischer EV (2018) Environmental conditions, ignition type, and air quality impacts of wildfires in the southeastern and western United States. Earths Future 6:1442–1456 Haikerwal A et al (2015) Impact of smoke from prescribed burning: Is it a public health concern? J Air Waste Manag Assoc 65:592–598 Jaffe DA et al (2020) Wildfire and prescribed burning impacts on air quality in the United States. J Air Waste Manag Assoc 70:583–615 Xu R et al (2020) Global Climate Change, and Human Health. N Engl J Med 383:2173–2181 Jaffe D, Hafner W, Chand D, Westerling A, Spracklen D (2008) Interannual variations in PM 2.5 due to wildfires in the Western United States. Environ Sci Technol 42:2812–2818 Sullivan A et al (2008) A method for smoke marker measurements and its potential application for determining the contribution of biomass burning from wildfires and prescribed fires to ambient PM 2.5 organic carbon. J Geophys Research: Atmos 113 Zhang Q et al (2007) Ubiquity and dominance of oxygenated species in organic aerosols in anthropogenically-influenced Northern Hemisphere midlatitudes. Geophys Res Lett 34 Bond TC et al (2013) Bounding the role of black carbon in the climate system: A scientific assessment. J Geophys research: Atmos 118:5380–5552 Orellano P, Reynoso J, Quaranta N, Bardach A, Ciapponi A (2020) Short-term exposure to particulate matter (PM 10 and PM 2.5 ), nitrogen dioxide (NO 2 ), and ozone (O 3 ) and all-cause and cause-specific mortality: Systematic review and meta-analysis. Environ Int 142:105876 Zhang L, Wilson JP, Zhao N, Zhang W, Wu Y (2022) The dynamics of cardiovascular and respiratory deaths attributed to long-term PM 2.5 exposures in global megacities. Sci Total Environ 842:156951 Pun VC, Kazemiparkouhi F, Manjourides J, Suh HH (2017) Long-term PM 2.5 exposure and respiratory, cancer, and cardiovascular mortality in older US adults. Am J Epidemiol 186:961–969 Li Z et al (2019) Impact of ambient PM 2.5 on adverse birth outcome and potential molecular mechanism. Ecotoxicol Environ Saf 169:248–254 Ma Y et al (2024) Long-term exposure to wildland fire smoke PM 2.5 and mortality in the contiguous United States. Proceedings of the National Academy of Sciences 121, e2403960121 Ye T et al (2022) Short-term exposure to wildfire-related PM 2.5 increases mortality risks and burdens in Brazil. Nat Commun 13:7651 Eisenman DP, Galway LP (2022) The mental health and well-being effects of wildfire smoke: a scoping review. BMC Public Health 22:2274 Karanasiou A et al (2021) Short-term health effects from outdoor exposure to biomass burning emissions: A review. Sci Total Environ 781:146739 Zhu Q et al (2024) Wildfires are associated with increased emergency department visits for anxiety disorders in the western United States. Nat Mental Health 2:379–387 Danesh Yazdi M et al (2024) Long-term exposure to PM 2.5 species and all-cause mortality among Medicare patients using mixtures analyses. Environ Res 246:118175 Hvidtfeldt UA et al (2019) Long-term residential exposure to PM 2.5 constituents and mortality in a Danish cohort. Environ Int 133:105268 Wang Y et al (2022) Long-term exposure to PM 2.5 major components and mortality in the southeastern United States. Environ Int 158:106969 Chen J et al (2017) A review of biomass burning: Emissions and impacts on air quality, health and climate in China. Sci Total Environ 579:1000–1034 Chen H, Samet JM, Bromberg PA, Tong H (2021) Cardiovascular health impacts of wildfire smoke exposure. Part Fibre Toxicol 18:2 Azevedo JM, Gonçalves FLT, de Fátima M, Andrade (2011) Long-range ozone transport and its impact on respiratory and cardiovascular health in the north of Portugal. Int J Biometeorol 55:187–202 Marlier ME et al (2013) El Niño and health risks from landscape fire emissions in southeast Asia. Nat Clim Change 3:131–136 Reid CE et al (2016) Critical Review of Health Impacts of Wildfire Smoke Exposure. Environ Health Perspect 124:1334–1343 Kiser D et al (2020) Particulate matter and emergency visits for asthma: a time-series study of their association in the presence and absence of wildfire smoke in Reno, Nevada, 2013–2018. Environ Health 19:92 Liu JC et al (2017) Wildfire-specific fine particulate matter and risk of hospital admissions in urban and rural counties. Epidemiol (Cambridge Mass) 28:77 Cleland SE et al (2020) Estimating wildfire smoke concentrations during the October 2017 California fires through BME space/time data fusion of observed, modeled, and satellite-derived PM 2.5 . Environ Sci Technol 54:13439–13447 Zhang D et al (2023) Wildland Fires Worsened Population Exposure to PM 2.5 Pollution in the Contiguous United States. Environ Sci Technol 57:19990–19998 Li Y et al (2021) Dominance of wildfires impact on air quality exceedances during the 2020 record-breaking wildfire season in the United States. Geophysical Research Letters 48, e2021GL094908 Pan X et al (2020) Six global biomass burning emission datasets: intercomparison and application in one global aerosol model. Atmos Chem Phys 20:969–994 Cummins K et al (2023) The Southeastern U.S. Prescribed Fire Permit Database: Hot Spots and Hot Moments in Prescribed Fire across the Southeastern U.S.A. Fire 6:372 Rogers HM, Ditto JC, Gentner DR (2020) Evidence for impacts on surface-level air quality in the northeastern US from long-distance transport of smoke from North American fires during the Long Island Sound Tropospheric Ozone Study (LISTOS) 2018. Atmos Chem Phys 20:671–682 Wang Z et al (2024) Intensification of mid-latitude cyclone by aerosol‐radiation interaction increases transport of Canadian wildfire smoke to northeastern US. Geophysical Research Letters 51, e2024GL108444 Childs ML et al (2022) Daily Local-Level Estimates of Ambient Wildfire Smoke PM 2.5 for the Contiguous US. Environ Sci Technol 56:13607–13621 Munoz-Alpizar R et al (2013–2016) Multi-Year PM 2.5 Wildfire Pollution Exposure over North America as Determined from Operational Air Quality Forecasts. Atmosphere 8, 179 (2017) Association AL (2020) State of the Air 2020 Burke M et al (2023) The contribution of wildfire to PM 2.5 trends in the USA. Nature 622:761–766 Zhai J, Ning Z, Dahal R, Yang S (2023) Wildfire Susceptibility of Land Use and Topographic Features in the Western United States: Implications for the Landscape Management. Forests 14:807 Williams AP et al (2019) Observed impacts of anthropogenic climate change on wildfire in California. Earths Future 7:892–910 Jain P, Castellanos-Acuna D, Coogan SC, Abatzoglou JT, Flannigan MD (2022) Observed increases in extreme fire weather driven by atmospheric humidity and temperature. Nat Clim Change 12:63–70 Halofsky JE, Peterson DL, Harvey BJ (2020) Changing wildfire, changing forests: the effects of climate change on fire regimes and vegetation in the Pacific Northwest, USA. Fire Ecol 16:1–26 Westerling AL, Hidalgo HG, Cayan DR, Swetnam TW (2006) Warming and earlier spring increase western US forest wildfire activity. science 313, 940–943 Yu G, Feng Y, Wang J, Wright DB (2023) Performance of Fire Danger Indices and Their Utility in Predicting Future Wildfire Danger Over the Conterminous United States. Earth's Future 11, eEF003823 (2023) Di Virgilio G et al (2019) Climate change increases the potential for extreme wildfires. Geophys Res Lett 46:8517–8526 Liu Y et al (2021) Projection of future wildfire emissions in western USA under climate change: contributions from changes in wildfire, fuel loading and fuel moisture. Int J Wildland Fire 31:1–13 Tian X, Zhao F, Shu L, Wang M (2013) Distribution characteristics and the influence factors of forest fires in China. For Ecol Manag 310:460–467 Valkó O, Deák B (2021) Increasing the potential of prescribed burning for the biodiversity conservation of European grasslands. Curr Opin Environ Sci Health 22:100268 Weir JR, Scasta JD (2022) Global application of prescribed fire. Csiro Publishing Kolden CA We’re Not Doing Enough Prescribed Fire in the Western United States to Mitigate Wildfire Risk. http://dx.doi.org/10.3390/fire2020030 Engebretson JM et al (2016) Characterizing public tolerance of smoke from wildland fires in communities across the United States. J Forest 114:601–609 U. S. D. o. t. Interior (2024) Budget Justifications and Performance Information Fiscal Year 2024: WILDLAND FIRE MANAGEMENT Wiedinmyer C, Hurteau MD (2010) Prescribed fire as a means of reducing forest carbon emissions in the western United States. Environ Sci Technol 44:1926–1932 Arkle RS, Pilliod DS, Welty JL (2012) Pattern and process of prescribed fires influence effectiveness at reducing wildfire severity in dry coniferous forests. For Ecol Manag 276:174–184 Tolhurst KG, McCarthy G (2016) Effect of prescribed burning on wildfire severity: a landscape-scale case study from the 2003 fires in Victoria. Australian Forestry 79:1–14 Williamson GJ, Bowman DMJS, Price OF, Henderson SB, Johnston FH (2016) A transdisciplinary approach to understanding the health effects of wildfire and prescribed fire smoke regimes. Environ Res Lett 11:125009 Solomon PA et al (2014) US national PM 2.5 chemical speciation monitoring networks—CSN and IMPROVE: description of networks. J Air Waste Manag Assoc 64:1410–1438 Hansen DA et al (2003) The southeastern aerosol research and characterization study: Part 1—Overview. J Air Waste Manag Assoc 53:1460–1471 Dabek-Zlotorzynska E et al (2011) Canadian National Air Pollution Surveillance (NAPS) PM 2.5 speciation program: Methodology and PM 2.5 chemical composition for the years 2003–2008. Atmos Environ 45:673–686 Appel KW et al (2021) The Community Multiscale Air Quality (CMAQ) model versions 5.3 and 5.3.1: system updates and evaluation. Geosci Model Dev 14:2867–2897 Wang L et al (2015) Source apportionment of PM 2.5 in top polluted cities in Hebei, China using the CMAQ model. Atmos Environ 122:723–736 Thongthammachart T et al (2021) An integrated model combining random forests and WRF/CMAQ model for high accuracy spatiotemporal PM 2.5 predictions in the Kansai region of Japan. Atmos Environ 262:118620 Zhang Q, Xue D, Liu X, Gong X, Gao H (2019) Process analysis of PM 2.5 pollution events in a coastal city of China using CMAQ. J Environ Sci 79:225–238 U. E. O. o. R. a. Development. (2020) CMAQ. 5.3.2. in Zenodo Meng X et al (2018) Estimating PM 2.5 speciation concentrations using prototype 4.4 km-resolution MISR aerosol properties over Southern California. Atmos Environ 181:70–81 Wei J et al (2023) Separating Daily 1 km PM 2.5 Inorganic Chemical Composition in China since 2000 via Deep Learning Integrating Ground, Satellite, and Model Data. Environ Sci Technol 57:18282–18295 Lyapustin A, Wang Y, Korkin S, Huang D (2018) MODIS Collection 6 MAIAC algorithm. Atmos Meas Tech 11:5741–5765 Chawla NV, Bowyer KW, Hall LO (2002) Kegelmeyer, SMOTE: synthetic minority over-sampling technique. J Artif Intell Res 16:321–357 Van der Laan MJ, Polley EC (2007) A. E. Hubbard, Super learner. Stat Appl Genet Mol Biol 6 Geng G, Meng X, He K, Liu Y (2020) Random forest models for PM 2.5 speciation concentrations using MISR fractional AODs. Environ Res Lett 15:034056 Ploton P et al (2020) Spatial validation reveals poor predictive performance of large-scale ecological mapping models. Nat Commun 11:4540 T. U. S. D. o. H. a. H. Services (2024) HHS Standard Values for Regulatory Analysis, (2024) (Aaron Kearsley) Judek S, Stieb D, Jovic B, Edwards B (2012) Air Quality Benefits Assessment Tool (AQBAT) user guide: Version 2. Health Canada, Ottawa, Ontario Additional Declarations The authors declare no competing interests. Supplementary Files SmokeCarbonPM25SupplementaryRS.docx Supplementary Material Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5478994","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":379716817,"identity":"3d48ea64-72da-465e-a92a-079c8bc4a087","order_by":0,"name":"Zhihao Jin","email":"","orcid":"","institution":"Gangarosa Department of Environmental Health, Rollins School of Public Health, Emory University","correspondingAuthor":false,"prefix":"","firstName":"Zhihao","middleName":"","lastName":"Jin","suffix":""},{"id":379719024,"identity":"ef1a2c8e-8bdf-43d4-aa51-091592bab85e","order_by":1,"name":"Gonzalo A. Ferrada","email":"","orcid":"","institution":"NOAA Earth System Research Laboratories","correspondingAuthor":false,"prefix":"","firstName":"Gonzalo","middleName":"A.","lastName":"Ferrada","suffix":""},{"id":379719025,"identity":"e482a242-e64f-45cf-b7ac-d9be7344f587","order_by":2,"name":"Danlu Zhang","email":"","orcid":"","institution":"Deparent of Biostatistics, Rollins School of Public Health, Emory University","correspondingAuthor":false,"prefix":"","firstName":"Danlu","middleName":"","lastName":"Zhang","suffix":""},{"id":379719026,"identity":"3690d3e9-9f56-43ad-813d-928d1d767da9","order_by":3,"name":"Noah Scovronick","email":"","orcid":"","institution":"Gangarosa Department of Environmental Health, Rollins School of Public Health, Emory University","correspondingAuthor":false,"prefix":"","firstName":"Noah","middleName":"","lastName":"Scovronick","suffix":""},{"id":379719027,"identity":"bf18609a-cb42-4d56-9843-09001e1e6c40","order_by":4,"name":"Joshua S. Fu","email":"","orcid":"","institution":"Deparent of Civil and Environmental Engineering, University of Tennessee","correspondingAuthor":false,"prefix":"","firstName":"Joshua","middleName":"S.","lastName":"Fu","suffix":""},{"id":379719028,"identity":"ee521cb6-d751-4edc-9d79-75620eb6c160","order_by":5,"name":"Kai Chen","email":"","orcid":"","institution":"Department of Environmental Health Sciences, Yale School of Public Health","correspondingAuthor":false,"prefix":"","firstName":"Kai","middleName":"","lastName":"Chen","suffix":""},{"id":379719029,"identity":"eb1885f8-325e-4673-942f-a8d9e12d34de","order_by":6,"name":"Yang Liu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA0ElEQVRIiWNgGAWjYNCCAgkGfjjnAFFaDCQYJBtI1AJEcJWEtBgcP3v4BYOBReLm492JjyvbGOT4biQQ0HImL80C6LDEbWfObjY828ZgLElQy4EcMwOwlhu52yQb2xgSNxDUcv4NRMvm+W/BWuoJa7mRY/wApGWDBC9YS4IBIS2SN96YAZVJGM84k7vZsOGchOHMMw/wa+E7n2P84UNFnWx/+9mNDxvKbOT5jhOwReEAA5sEXA0jmwR+5SAg38DA/AHB/UNYxygYBaNgFIw8AABA4EkKTJO4iAAAAABJRU5ErkJggg==","orcid":"","institution":"Gangarosa Department of Environmental Health, Rollins School of Public Health, Emory University","correspondingAuthor":true,"prefix":"","firstName":"Yang","middleName":"","lastName":"Liu","suffix":""}],"badges":[],"createdAt":"2024-11-19 00:14:23","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-5478994/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5478994/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":69521901,"identity":"01513f37-695f-45fe-b86a-ec81564b5047","added_by":"auto","created_at":"2024-11-21 09:00:39","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":322492,"visible":true,"origin":"","legend":"\u003cp\u003eAnnual average concentration of background and smoke PM\u003csub\u003e2.5\u003c/sub\u003e OC and EC in the CONUS and SC, categorized by climate regions and spanning the study period from 2002 to 2019.\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-5478994/v1/88a6fb3b61c907d628cea4ac.png"},{"id":69521900,"identity":"fb9aca86-947a-4d3a-a20f-b9583e6d18ce","added_by":"auto","created_at":"2024-11-21 09:00:39","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":836914,"visible":true,"origin":"","legend":"\u003cp\u003eLong-term average annual concentration of smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e from 2002 to 2019, with major urban centers (located at red points) zoomed in for detailed visualization.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-5478994/v1/b642463e45d485b391d0171c.png"},{"id":69521898,"identity":"9be174dd-4b8c-4fab-a1c9-a73344760d03","added_by":"auto","created_at":"2024-11-21 09:00:38","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":128645,"visible":true,"origin":"","legend":"\u003cp\u003eCumulative person-days exposed to heavy fire smoke in the CONUS and SC (Unit: million person-day).\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-5478994/v1/c317c5d0b6e6412321dd8880.png"},{"id":69522526,"identity":"9791cb27-1d57-4b2f-93d0-aa8b781bfa13","added_by":"auto","created_at":"2024-11-21 09:08:39","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":212068,"visible":true,"origin":"","legend":"\u003cp\u003eAnnual total deaths attributable to fire smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e in the CONUS (subplot-A) and SC (subplot-B). \u003cstrong\u003eNote\u003c/strong\u003e: Climate regions in the CONUS were grouped into three major regions based on the U.S. National Prescribed Fire Use Survey Report and our analysis of smoke-impacted areas (see \u003cstrong\u003eFig. S5\u003c/strong\u003e). These regions are defined as follows: the western region (i.e., climate regions: West, SouthWest, NorthWest, and West North Central), the southeastern region (i.e., climate regions: SouthEast, Central, and South), and the northeastern region (i.e., climate regions: NorthEast, East North Central). The annual total monetized damages for the CONUS and SC are labeled at the top of each column (units: billion USD for the CONUS and billion CAD for SC).\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-5478994/v1/3e2d12b3e089adcf25c0c32c.png"},{"id":69521899,"identity":"79ee3dfb-5e9f-40a1-8b44-d69502ffd04d","added_by":"auto","created_at":"2024-11-21 09:00:38","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":365749,"visible":true,"origin":"","legend":"\u003cp\u003eStudy domain and ground-based monitoring networks of PM\u003csub\u003e2.5\u003c/sub\u003e OC and EC.\u003c/p\u003e","description":"","filename":"image5.png","url":"https://assets-eu.researchsquare.com/files/rs-5478994/v1/dd0eda32960dbfb077c74d44.png"},{"id":69522527,"identity":"5c9df0c3-fcf5-4426-a3e5-347aee4b5af9","added_by":"auto","created_at":"2024-11-21 09:08:39","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":109201,"visible":true,"origin":"","legend":"\u003cp\u003eModeling framework of the three-level residual adjusted super learner (RASL).\u003c/p\u003e","description":"","filename":"image6.png","url":"https://assets-eu.researchsquare.com/files/rs-5478994/v1/97ee55f4cb4265dd54a218ea.png"},{"id":69523787,"identity":"7f36c0cb-9325-4d6b-b42f-caa1d2084954","added_by":"auto","created_at":"2024-11-21 09:16:40","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2655278,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5478994/v1/c7220cf2-bf63-458c-ba87-d46c83de0220.pdf"},{"id":69521905,"identity":"d3be1974-97da-43ac-a34a-a83a52f071d0","added_by":"auto","created_at":"2024-11-21 09:00:39","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":7484342,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Material\u003c/p\u003e","description":"","filename":"SmokeCarbonPM25SupplementaryRS.docx","url":"https://assets-eu.researchsquare.com/files/rs-5478994/v1/452c25e279b5c5142906d57a.docx"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003eFire Smoke Elevated the Carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e Concentration and Mortality Burden in the Contiguous U.S. and Southern Canada\u003c/p\u003e","fulltext":[{"header":"Introduction","content":"\u003cp\u003eOver the past half-century, wildland fire activity has significantly increased, not only in the U.S. but also in other temperate and high-latitude ecosystems, including Canada and Europe (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e). Notably, human-induced climate change was responsible for an additional 4.2\u0026nbsp;million hectares of forest fire area between 1984 and 2015, nearly doubling the area expected to be affected by fire in the absence of climate change (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e). As a result of climate change, large-scale wildland fire events have become more frequent and intense, and fire seasons have lengthened in the Contiguous U.S. (CONUS) in recent decades. Previous research has indicated that wildland fire smoke have accounted for nearly 25% of the ambient fine particulate matter (PM\u003csub\u003e2.5\u003c/sub\u003e, particles with a diameter of less than 2.5 \u0026micro;m) across the U.S. in recent years, and up to 50% in certain Western U.S. regions (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eOne of the primary wildfire management strategies is prescribed burning. Prescribed fires not only reduce the biomass available for subsequent wildfires, but they also support carbon sequestration, facilitate ecological resilience, and play a critical role in restoring fire-adapted ecosystems that have been degraded due to decades of fire exclusion (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e). Over 65% of the prescribed burn areas are in the Southeastern U.S (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e). This imbalance in the application of prescribed fires has resulted in comparable regional average PM emissions from prescribed fires in the Southeastern U.S. and wildfires in the Western U.S. (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e). In the context of climate change, as the use of prescribed burning is increasing to control wildfires, smoke from these burns is becoming a growing public health concern, particularly in the Southeastern U.S. (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e). The National Prescribed Fire Acts (116th and 118th Congress) emphasizes the importance of public health and safety risks associated with the expanded use of prescribed fires. However, it states that smoke from prescribed fires is generally less harmful and of shorter duration compared to wildfire smoke, stating that it exposes children to fewer adverse health effects. Such a statement, however, is based on limited research, which may lead to an underestimation of prescribed burning's health risks.\u003c/p\u003e \u003cp\u003eWith anthropogenic climate change playing an increasingly critical role in escalating wildfire activity, the negative impacts of smoke on air quality and public health are likely to worsen in the future (\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e). Fire smoke contains considerable amount of PM\u003csub\u003e2.5\u003c/sub\u003e, significantly deteriorating the air quality in downwind communities that are tens to hundreds of kilometers away (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e). Smoke PM\u003csub\u003e2.5\u003c/sub\u003e is characterized by substantial concentrations of carbonaceous matter, including organic carbon (OC) and elemental carbon (EC), which are produced by the combustion and incomplete burning of organic materials such as wood, leaves, and other vegetation. This distinguishes fire smoke PM\u003csub\u003e2.5\u003c/sub\u003e from typical ambient PM\u003csub\u003e2.5\u003c/sub\u003e, which tends to present greater oxidative potential (\u003cspan additionalcitationids=\"CR14\" citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e). The unique characteristics of fire smoke PM\u003csub\u003e2.5\u003c/sub\u003e may alter the composition of regional PM\u003csub\u003e2.5\u003c/sub\u003e and potentially affect its toxicity.\u003c/p\u003e \u003cp\u003eWhile numerous studies have linked exposure to PM\u003csub\u003e2.5\u003c/sub\u003e with various adverse health impacts (\u003cspan additionalcitationids=\"CR17 CR18\" citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e), epidemiological research linking exposure to fire smoke PM\u003csub\u003e2.5\u003c/sub\u003e with adverse health outcomes is still in its early stage. Long-term exposure to smoke PM\u003csub\u003e2.5\u003c/sub\u003e has been linked to all-cause mortality in the CONUS, particularly among vulnerable populations such as the elderly (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e). It is estimated that 11,415 non-accidental deaths per year in the CONUS can be attributed to smoke PM\u003csub\u003e2.5\u003c/sub\u003e, with cardiovascular diseases contributing the most (\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e). Short-term exposure to wildfire smoke PM\u003csub\u003e2.5\u003c/sub\u003e has been associated with increased risks of respiratory morbidity, mental health issues, and excess mortality, (\u003cspan additionalcitationids=\"CR22 CR23\" citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e). Evidence on the health effects of different chemical components of smoke PM\u003csub\u003e2.5\u003c/sub\u003e is sparse. For example, OC has been identified to be an important component influencing PM\u003csub\u003e2.5\u003c/sub\u003e toxicity to several reactions harming organic systems and a key contributor to all-cause mortality (\u003cspan additionalcitationids=\"CR26\" citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e). EC, due to its small size, can penetrate deeply into the respiratory tract and serve as a transporter for various toxic substances (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eResearch on the health effects of smoke PM\u003csub\u003e2.5\u003c/sub\u003e has been hindered owing to the scarcity of long-term exposure data, especially data with comprehensive spatial coverage and high spatial-temporal resolution. Most epidemiological studies on smoke PM have relied on local ground-based monitoring stations, satellite images, uncalibrated chemical transport model simulations (CTM) or simple classifications of smoke-affected areas to investigate the health impacts of fire smoke (\u003cspan additionalcitationids=\"CR30 CR31 CR32 CR33\" citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e). These methods were either based on coarse resolution smoke estimates or did not quantify smoke PM, potentially introducing exposure misclassification. Emerging research has shown great promise to generate long-term and high-resolution smoke-derived PM\u003csub\u003e2.5\u003c/sub\u003e concentrations by calibrating CTM simulations. For instance, Cleland et al. tested the model performance of predicting 1 km\u003csup\u003e2\u003c/sup\u003e wildfire smoke PM\u003csub\u003e2.5\u003c/sub\u003e based on CTMs simulations and different combinations of concentration datasets (\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e). The model that fused ground-based observations, satellite aerosol optical depth (AOD)-derived concentrations and CTMs simulations provided the best estimate (R\u003csup\u003e2\u003c/sup\u003e\u0026thinsp;=\u0026thinsp;0.71) in fire-impacted regions, highlighting the importance of integrating multiple datasets. Similarly, Zhang et al. developed CMAQ-based models to estimate daily 1 km\u003csup\u003e2\u003c/sup\u003e smoke PM\u003csub\u003e2.5\u003c/sub\u003e total mass, which achieved strong model performance with R\u003csup\u003e2\u003c/sup\u003e of 0.75 and 0.68 in smoke-impacted regions and non-smoke regions, respectively (\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e). Nevertheless, few studies have adopted CTM-based models to estimate smoke PM\u003csub\u003e2.5\u003c/sub\u003e speciation with high spatial and temporal resolution. This is largely because CTM simulations for PM\u003csub\u003e2.5\u003c/sub\u003e speciation often face higher uncertainties compared to those for total PM\u003csub\u003e2.5\u003c/sub\u003e mass, demanding more advanced calibration techniques (\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn this study, we developed a three-level, CTM-based modeling framework to estimate daily concentrations of fire smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e, specifically OC and EC, at 1 km\u003csup\u003e2\u003c/sup\u003e spatial resolution from 2002 to 2019 with full coverage across the CONUS and Southern Canada (SC). This framework integrated information from CMAQ simulations of PM\u003csub\u003e2.5\u003c/sub\u003e mass and speciation, ground-based observations and multiple auxiliary spatial and spatiotemporal datasets. This innovative approach allows us to fill important research gaps described above, namely, to differentiate exposure by specific carbonaceous constituents of smoke PM\u003csub\u003e2.5\u003c/sub\u003e, and to estimate fire smoke-related health burdens over the long-term. By leveraging the high spatial and temporal resolution of our model predictions, we then analyzed the spatiotemporal patterns of smoke impact frequency and concentrations of carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e from fires smoke. Furthermore, we estimated the populations exposed to fire smoke. Lastly, we investigated the impacts of long-term exposure to fire smoke on mortality burden and associated monetized damages.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eModel performance\u003c/h2\u003e \u003cp\u003eThe CV results revealed strong model performance (\u003cb\u003eTable S1-S6\u003c/b\u003e). For daily-level predictions, Smoke-off models showed higher accuracy, with random CV R\u0026sup2; values above 0.75 for base learners and 0.77 for meta-learners. Meta-learners for EC achieved average random CV R\u0026sup2; values of 0.80 in Smoke-off and 0.71 in Smoke-on scenarios, while OC performance dropped in Smoke-on with a reduced R\u0026sup2; of 0.67 and increased RMSE of 1.20 \u0026micro;g/m\u0026sup3;. Spatial CV showed low prediction errors for OC and EC in Smoke-off (\u003cb\u003eFig. S1\u003c/b\u003e), with RMSEs below 0.8 \u0026micro;g/m\u0026sup3; and 0.2 \u0026micro;g/m\u0026sup3;, respectively. In contrast, Smoke-on scenarios exhibited relatively higher RMSEs, especially in areas prone to fire smoke. For temporal CV, R\u003csup\u003e2\u003c/sup\u003e values were consistently higher for Smoke-off scenarios (\u003cb\u003eFig. S2\u003c/b\u003e), ranging between 0.65 and 0.85. In Smoke-on scenarios, OC and EC displayed fluctuate R\u0026sup2; values between 0.58 and 0.65, with notable declines in 2002. At monthly and annual levels, model performance improved significantly (\u003cb\u003eTable S1-S6\u003c/b\u003e). Adjustments to residuals through GAMs further enhanced accuracy (\u003cb\u003eTable S7\u003c/b\u003e), with random CV R\u0026sup2; values surpassing 0.91 for monthly and 0.97 for annual predictions. Most overestimations and underestimations observed at the daily level were reduced by averaging at the monthly level (\u003cb\u003eFig. S3\u003c/b\u003e). Overall, the RASL model demonstrates reliable daily predictions across various CV experiments and further improvements in adjusting the long-term prediction residuals. Feature importance analysis (\u003cb\u003eFig. S4\u003c/b\u003e) indicated CMAQ simulations of carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e were primary predictors, with MAIAC AOD, urbanization factors, meteorological factors, and spatial and temporal characteristics also influential.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eSpatial and temporal patterns of smoke carbonaceous PM\u003c/h3\u003e\n\u003cp\u003eFigure S5 illustrates the average number of smoke days per year across our study domain. The central-south (i.e., Missouri, Arkansas, and Oklahoma) and southeastern (Alabama, Georgia, and Florida) regions of the CONUS exhibited the highest average number of smoke days (200\u0026thinsp;+\u0026thinsp;days per year). The Western CONUS (i.e., California, Oregon, Idaho, and Montana), North-Central CONUS (i.e., North Dakota, South Dakota, and Minnesota) and adjacent areas in SC such as British Columbia also experienced a significant number of smoke days (~\u0026thinsp;150 days), though the sources of smoke in these regions differ. In the Western CONUS and SC, the primary source of smoke is wildfires. Conversely, in the central-south and southeastern regions of the CONUS, prescribed fires are the main source of smoke (\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e). The northeastern regions of the study domain are less frequently impacted by smoke (~\u0026thinsp;100 days) but can still experience significant smoke pollution from long-range transport (\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eWe summarized the background, total, and smoke-specific concentrations of PM\u003csub\u003e2.5\u003c/sub\u003e OC and EC across different time scales and climate regions in \u003cb\u003eFig. S6\u003c/b\u003e. From 2002 to 2019, both the CONUS and SC experienced a declining trend in background carbonaceous PM\u003csub\u003e2.5,\u003c/sub\u003e with annual concentrations falling below 0.95 \u0026micro;g/m\u003csup\u003e3\u003c/sup\u003e for OC and 0.20 \u0026micro;g/m\u003csup\u003e3\u003c/sup\u003e for EC by 2019. When comparing the background carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e between the periods 2002\u0026ndash;2010 and 2011\u0026ndash;2019, improvements in annual background PM\u003csub\u003e2.5\u003c/sub\u003e were observed, with reductions of 0.27 \u0026micro;g/m\u0026sup3; for the CONUS and 0.17 \u0026micro;g/m\u0026sup3; for SC after 2011. The CONUS climate regions of SouthEast, South, Central, and NorthEast exhibited higher background concentrations and more significant reductions over the years (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Additionally, urban areas displayed higher background concentrations of carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e (\u003cb\u003eFig. S7\u003c/b\u003e), with long-term average concentrations reaching approximately 2 \u0026micro;g/m\u003csup\u003e3\u003c/sup\u003e for OC and 1 \u0026micro;g/m\u003csup\u003e3\u003c/sup\u003e for EC. Elevated background PM\u003csub\u003e2.5\u003c/sub\u003e OC levels were also common in many rural and forested areas of the South, Central and SouthEast CONUS climate regions, while high levels of PM\u003csub\u003e2.5\u003c/sub\u003e EC were mostly concentrated in urban centers across the study domain.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eConsidering the impacts of fire smoke, we observed a distinct contribution of smoke PM\u003csub\u003e2.5\u003c/sub\u003e OC, whereas the contribution of smoke PM\u003csub\u003e2.5\u003c/sub\u003e EC was smaller and showed less fluctuation over the years. The mean annual concentration of smoke PM\u003csub\u003e2.5\u003c/sub\u003e OC was 0.22 \u0026micro;g/m\u0026sup3; for the CONUS and 0.14 \u0026micro;g/m\u0026sup3; for SC, while smoke EC concentrations were 0.08 \u0026micro;g/m\u0026sup3; for the CONUS and 0.05 \u0026micro;g/m\u0026sup3; for SC. Combining smoke PM\u003csub\u003e2.5\u003c/sub\u003e OC and EC, smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e accounted for 19% and 16% of the total concentrations in the CONUS and SC, respectively. At the monthly level, smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e exhibited a notable increase during the peak months of wildfire season (July\u0026ndash;November) (\u003cb\u003eFig. S8\u003c/b\u003e), with average monthly concentrations rising to 2.38 \u0026micro;g/m\u0026sup3; for the CONUS and 2.60 \u0026micro;g/m\u0026sup3; for SC, accounting for 58% and 57% of total concentrations, respectively.\u003c/p\u003e \u003cp\u003eThe Western and Southern CONUS climate regions (i.e., NorthWest, West, West North Central, South, and SouthEast) and the Southwestern Canada experienced more fire smoke impacts, including wildland fire and prescribed fire, resulting in increased smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e at both regional and national scales (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Additionally, the annual mean concentrations of smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e increased by 0.08 \u0026micro;g/m\u0026sup3; for the CONUS and 0.05 \u0026micro;g/m\u0026sup3; for SC before and after 2011. This intensifying trend in fire smoke has offset nearly one-third of the improvements in background concentration. Megafire years of 2012, 2015, 2017, and 2018 experienced significantly higher concentrations of both smoke PM\u003csub\u003e2.5\u003c/sub\u003e OC and EC at the monthly and annual levels.\u003c/p\u003e \u003cp\u003eOur 1 km\u003csup\u003e2\u003c/sup\u003e prediction maps revealed different spatial distributions between long-term smoke OC and EC concentrations (\u003cb\u003eFig. S7\u003c/b\u003e). High smoke OC concentrations (\u0026gt;\u0026thinsp;0.50 \u0026micro;g/m\u0026sup3;) were primarily observed in rural areas of the Western CONUS and SC, and were sporadically distributed in the Southeastern CONUS. Elevated EC concentrations (\u0026gt;\u0026thinsp;0.15 \u0026micro;g/m\u0026sup3;) were often collocated with high smoke OC, with the Southeastern CONUS, particularly Georgia, Florida, Mississippi, and Texas, exhibiting even higher smoke EC concentrations (\u0026gt;\u0026thinsp;0.20 \u0026micro;g/m\u0026sup3;). Long-term average concentrations of smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e were highest in the Western CONUS and Southwestern Canada, particularly in California, Idaho, and Montana (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). The Southeastern CONUS also experienced comparable concentrations of smoke carbonaceous PM\u003csub\u003e2.5,\u003c/sub\u003e with urban centers such as Charlotte and Atlanta, and their surrounding areas, showing higher exposure levels than urban areas in other parts of the study domain. During megafire years (2012, 2015, 2017, and 2018), elevated smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e were concentrated in rural regions of the Western CONUS, as well as urban centers such as Los Angeles, San Francisco, Seattle and Vancouver, as captured in the annual prediction maps (\u003cb\u003eFig. S9\u003c/b\u003e and \u003cb\u003eFig. S10\u003c/b\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\n\u003ch3\u003eIncreasing impact of wildland fire smoke on air quality and population exposure\u003c/h3\u003e\n\u003cp\u003eWe evaluated the populations affected by heavy fire smoke in the CONUS and SC. A \u0026ldquo;heavy fire smoke grid-day\u0026rdquo; was defined as a grid-day when the total carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e concentration exceeded 1 \u0026micro;g/m\u0026sup3; and the smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e constituted more than 50% of the total concentration. We compared the cumulative daily populations affected by heavy fire smoke during the periods 2002\u0026ndash;2010, 2011\u0026ndash;2019, and specifically the five years from 2015\u0026ndash;2019 (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e\u003cb\u003e)\u003c/b\u003e. Our analysis revealed a clear increasing trend in population exposed over the years, corresponding to the broadening of wildfire smoke impacts. During 2002\u0026ndash;2010, there was an average of 467\u0026nbsp;million person-days of exposure to heavy fire smoke prior to July 1st, while this number rose by 70% to 793\u0026nbsp;million person-days during 2011\u0026ndash;2019. By year\u0026rsquo;s end, cumulative exposure increased from 986\u0026nbsp;million person-days in 2002\u0026ndash;2010 to 1.8\u0026nbsp;billion person-days in 2011\u0026ndash;2019, representing an 83% increase. To account for the influence of population growth on these trends, we also calculated the annual exposure days per capita. On average, individuals experienced 3.0 days of heavy smoke exposure per year in 2002\u0026ndash;2010, which increased to 5.1 days per year in 2011\u0026ndash;2019. The megafire years of 2017 and 2018 had particularly severe impacts, with cumulative exposure exceeding 3\u0026nbsp;billion person-days (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e), equivalent to 9.0 exposure days per person per year. Separate figures for cumulative person-days of exposure to heavy fire smoke in the CONUS and SC across years are provided in \u003cb\u003eFig. S11\u003c/b\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e\n\u003ch3\u003eExcess mortality due to smoke exposure\u003c/h3\u003e\n\u003cp\u003eThe annual non-accidental mortality rate and total deaths attributable to smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e from 2003 to 2020 are mapped at the county level in the CONUS and the census division (CD) level in SC (\u003cb\u003eFig. S12)\u003c/b\u003e. Consistent with the spatial distribution of high smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e concentrations in the Western and Southeastern CONUS and Southwestern Canada, counties and CDs in these areas exhibited elevated annual mortality rates, exceeding 3 deaths per 100,000 people (\u003cb\u003eFig. S12A\u003c/b\u003e). The CONUS counties with higher annual death counts (\u0026gt;\u0026thinsp;3 deaths per year) were generally located in California, Oregon, Washington, Florida, Georgia, South Carolina, North Carolina, and Texas (\u003cb\u003eFig. S12B\u003c/b\u003e), areas characterized by high populations density, frequent smoke impact and elevated smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e. On average, 7,462 non-accidental deaths per year in the CONUS were attributable to long-term exposure to fire smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e (\u003cb\u003eTable S8\u003c/b\u003e), with the southeastern CONUS contributing 4,702 deaths per year (63.0%), and the western region contributing 1,568 deaths per year (21.0%). In SC, higher annual deaths were observed in urban areas of British Columbia, Ontario, and Quebec (\u003cb\u003eFig. S12B\u003c/b\u003e), with an average of 259 non-accidental deaths per year (\u003cb\u003eTable S9\u003c/b\u003e). Annual total deaths in the southeastern region remained relatively stable over the study period (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e), whereas deaths in the western and northeastern regions, as well as in SC, fluctuated significantly in response to variations in wildfire intensity.\u003c/p\u003e \u003cp\u003eThe average monetized damages associated with these mortality estimates were approximately 68.4\u0026nbsp;billion USD per year for the CONUS and 1.97\u0026nbsp;billion CAD per year for SC (\u003cb\u003eTable S8\u003c/b\u003e and \u003cb\u003eTable S9\u003c/b\u003e). The southeastern region contributed 43.0\u0026nbsp;billion USD annually, and the western region contributed 14.2\u0026nbsp;billion USD. In 2018 and 2019, both mortality and monetized damages nearly doubled compared to the average levels, leading to monetized damages exceeding 120\u0026nbsp;billion USD for the CONUS and 4.6\u0026nbsp;billion CAD for SC (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eTo the best of our knowledge, this is the first study to model full-coverage concentration of fire smoke-derived carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e with high spatial and temporal resolution across both the CONUS and SC. We developed a three-level RASL framework to estimate both daily and long-term smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e from 2002 to 2019. Our analysis identified frequent smoke impact and elevated concentrations of smoke-derived carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e in the Western and Southeastern CONUS, as well as in SC. Over the past decade, wildfire seasons have started earlier and intensified, resulting in increased population exposure to wildfire smoke, particularly during recent megafire years. We estimated that long-term exposure to smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e resulted in an average of 7,462 and 259 non-accidental deaths per year in the CONUS and SC, respectively, with the Southeastern CONUS contributing the most deaths.\u003c/p\u003e \u003cp\u003eThe long-term concentration of PM\u003csub\u003e2.5\u003c/sub\u003e mass from fire smoke in the CONUS and Canada has been examined in previous literature. Our findings on the spatial and temporal patterns of smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e are consistent with these prior studies. For example, Childs et al. estimated smoke PM\u003csub\u003e2.5\u003c/sub\u003e over the CONUS at a 10 km resolution and observed increased smoke pollution and smoke-impacted days over the last decade, especially in the Western U.S. and in the years 2017 and 2018 (\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e). The FireWork model provided real-time forecasts of biomass burning PM\u003csub\u003e2.5\u003c/sub\u003e across North America, showing that most wildfire events were concentrated in the Western CONUS, as well as in Western, Northern, and Central Canada (\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e). While previous studies have offered valuable insights into wildfire smoke PM\u003csub\u003e2.5\u003c/sub\u003e, our use of daily 1 km\u003csup\u003e2\u003c/sup\u003e resolution and focus on carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e provides a more precise accounting of fire smoke impacts and a more detailed understanding of its composition. Notably, our study revealed that the frequent smoke days and elevated concentration in the Southeastern CONUS are consistent over time due to the frequent use of prescribed burns, which has not been extensively discussed in prior wildfire smoke modeling studies or government datasets. For example, shortcomings were identified in the current prescribed fire permit databases given that some states in the Southeastern U.S. do not require prescribed burn permits and rely on voluntary reporting, such as Texas, Arkansas, and Missouri (\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e). Despite these limitations, our study successfully captured the elevated smoke days and smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e in these states.\u003c/p\u003e \u003cp\u003eAdditionally, our results indicated a significant intensification of wildfires with higher concentrations of smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e across our study period, which has offset nearly one-third of the improvements in background air quality across the CONUS and SC, largely due to efforts such as the Clean Air Act (\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e). This intensifying trend in wildfire smoke stagnated or even reversed the declining trend of background concentrations in most regions. Our findings align with existing literature on smoke PM\u003csub\u003e2.5\u003c/sub\u003e mass. Burke et al. reported that areas affected by wildfire smoke have doubled over the past two decades in the CONUS, and that wildfire smoke has influenced the average annual PM\u003csub\u003e2.5\u003c/sub\u003e trend in nearly three-quarters of the states in the CONUS since 2016, accounting for 25% of the improvement in air quality (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e). However, few studies have investigated the wildfire smoke impact on PM\u003csub\u003e2.5\u003c/sub\u003e concentration levels in Canada. Our results provide insights into wildfire smoke dynamics in Canada and the transboundary smoke effects. Nonetheless, additional research encompassing Alaska, Northern Canada, and Mexico is needed to achieve a more comprehensive understanding of the wildfire smoke activity across North America.\u003c/p\u003e \u003cp\u003eIn addition to the increased smoke concentration, our findings indicate that exposure to wildfire smoke is becoming more common, implying more frequent wildfire smoke impact, wider smoke-impacted areas, and earlier and longer wildfire seasons. Consequently, the duration of exposure has lengthened, increasing from an average of 3.0 exposure days per year exposed to heavy fire smoke during 2002\u0026ndash;2010 to 5.1 days per year during 2011\u0026ndash;2019 for population across our study domain. Climate change plays a pivotal role in these shifts, as rising temperatures and prolonged droughts have shifted season regimes and created more favorable conditions for wildfires by drying out vegetation and extending the wildfire season (\u003cspan additionalcitationids=\"CR47 CR48 CR49 CR50\" citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e). These conditions also facilitate megafires becoming more common, leading to extremely high concentrations of carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e. Our study exhibited significantly higher concentration levels and wider population exposure during megafire years such as 2017 and 2018, leading to increased health burden. Without significant mitigation efforts, future climate models predict an alarming increase in wildfire frequency and severity, which poses further risks to ecosystems, air quality, and public health (\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e, \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e)​.\u003c/p\u003e \u003cp\u003ePrescribed fires are widely recognized as one of the most effective ways to prevent potential wildfires and sustain biodiversity in all regions beyond the Southeastern CONUS (\u003cspan additionalcitationids=\"CR55\" citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e). Our study has revealed lower concentrations of smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e in the Southeastern CONUS during megafire years compared to the Western CONUS. However, our analysis also indicates that prescribed fires do not necessarily translate into better air quality and may lead to consistent smoke pollution with elevated carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e concentration over time, causing adverse health effects from long-term exposure. In the Southeastern CONUS region, where prescribed fires are the primary sources of smoke, we estimated 4,702 attributable deaths per year, which is higher than the combined deaths in the Western CONUS (1,568 deaths per year) and Northeastern CONUS (1,192 deaths per year) regions. Several factors contribute to this outcome. First, unlike the Western U.S., where wildfires occur sporadically but at high intensity, the Southeastern CONUS experiences regular smoke pollution from frequent prescribed fires. Our results indicated that annual regional average concentrations of smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e in the Southeast CONUS are comparable to those in the Western CONUS (~\u0026thinsp;0.4 \u0026micro;g/m\u0026sup3;). While prescribed fires effectively reduce wildfires risks, they also result in frequent and localized smoke pollution in the Southeastern CONUS, where prescribed burns are conducted throughout the year (\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e). Second, prescribed fire smoke frequently affects densely populated urban and suburban areas in the Southeast CONUS, such as Atlanta and Charlotte, even though the fires are smaller and controlled (\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e). In contrast, wildfires in the Western CONUS typically occur in more remote, forested areas, such as the Cascades and Rocky Mountains. While wildfire smoke can travel long distances to urban centers like Los Angeles, San Francisco, and even the Northeastern CONUS, the primary impact is often concentrated in less densely populated areas. Third, residents of Southeastern states such as Georgia and Florida are more accustomed to prescribed fires (\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e). Engebretson et al. reported that Southern-state residents demonstrate significantly higher tolerance of potential health impacts from prescribed fires compared to those in Western states (\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e). While this acceptance reduces social barriers to the use of prescribed fires, it also lowers public vigilance regarding exposure to fire smoke pollutants. In contrast, the perception of wildfire risk in the Western U.S. has been heightened by the prevalence of megafires and media coverage, leading to a higher public awareness of the dangers posed by wildfire smoke and harm-reduction behaviors.\u003c/p\u003e \u003cp\u003eThe average monetized damages associated with the attributable deaths in the Southeastern U.S. is 43.0\u0026nbsp;billion USD per year, and this cost for the CONUS has exceeded 120\u0026nbsp;billion USD in recent megafire years. In contrast, the U.S. allocated \u003cspan\u003e$\u003c/span\u003e1.73\u0026nbsp;billion to wildland fire management in 2024, with \u003cspan\u003e$\u003c/span\u003e214.5\u0026nbsp;million dedicated to fuels management (\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e). Considerable evidence in the scientific literature supports prescribed fire as a cheap and effective method for mitigating wildfire risk and reducing carbon emissions (\u003cspan additionalcitationids=\"CR61\" citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e). However, most of these studies overlook the significant health impacts associated with smoke exposure from prescribed fire, which can be transported to nearby populated areas. When considering health-related costs, the cost-effectiveness of prescribed burns is called into question, as these fires can cause damage over 200 times greater than the budget. To better inform policy, it is crucial to develop a more accurate and comprehensive cost-benefit assessment for prescribed burns by incorporating health-related monetized damages from smoke exposure. Additionally, policies should prioritize minimizing human smoke exposure by improving monitoring networks of both PM\u003csub\u003e2.5\u003c/sub\u003e mass and carbonaceous components and preparing communities for potential health impacts. Enhancing communication strategies to warn residents and provide resources, such as air quality alerts and protective equipment, should be an essential component of these policies. Beyond policy improvements, a more sophisticated prescribed fire management system is necessary \u0026mdash; one that considers the conditions of each fire, including risk factors such as weather and proximity to populations, and the long-term benefits of prescribed burns in reducing the severity and frequency of wildfires (\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e). Achieving this balance between prescribed fire and public health is essential to ensure that prescribed burns remain a valuable tool for ecosystem health and wildfire prevention.\u003c/p\u003e \u003cp\u003eOur study has several implications. First, it provides a high-resolution fire smoke product with full coverage for the CONUS and SC, offering insights into the occurrence and distribution of fire smoke impacts, and quantitative estimates of background and smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e concentrations. The comprehensive spatial and temporal coverage of our predictions will enable future research on the health and environmental impacts of exposure to altered PM\u003csub\u003e2.5\u003c/sub\u003e composition by fire smoke. Second, our findings highlight that wildland fires have intensified over the past decade, leading to an increase in deaths associated with long-term exposure to smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e. Finally, by combining the prescribed fire permit databases with our study\u0026rsquo;s high-resolution smoke predictions, future efforts could better track prescribed fire activities in terms of locations, durations, sizes and transmissions. As climate change continues to challenge wildfire risk mitigation and biodiversity conservation, our study underscores the importance of incorporating potential health impacts in the cost-benefit analysis of prescribed fire policies and management tools.\u003c/p\u003e \u003cp\u003eSeveral limitations of our study should be noted. First, there is currently no research specifically investigating the mortality attributable to smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e. As a result, our study relies on mortality risk estimates based on total smoke PM\u003csub\u003e2.5\u003c/sub\u003e. As carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e is an important component influencing PM\u003csub\u003e2.5\u003c/sub\u003e toxicity and a key contributor to all-cause mortality, using the carbonaceous fraction as a proxy for total smoke PM\u003csub\u003e2.5\u003c/sub\u003e may lead to an underestimation of non-accidental deaths and monetized damages attributable to smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e. Future research focusing on the mortality risk associated with specific components of smoke PM\u003csub\u003e2.5\u003c/sub\u003e is needed to more accurately estimate the impact of carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e from fire smoke. Second, the annual smoke-mortality relationships for both the CONUS and SC applied in our study may not be entirely applicable to the 2020 baseline mortality rate, which was impacted by the COVID pandemic.\u003c/p\u003e \u003cp\u003eIn conclusion, our study highlights the growing impact of fire smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e and its adverse effects on public health across the CONUS and SC. With wildfires intensifying and becoming more frequent due to climate change, our findings underscore the urgent need for comprehensive prescribed fire management strategies that balance ecological benefits with the reduction of smoke-related health risks. This work provides a valuable foundation for future research and policymaking to address the dual challenges of wildfire prevention and public health protection in an increasingly fire-prone environment.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\n \u003ch2\u003eStudy domain and Period\u003c/h2\u003e\n \u003cp\u003eOur study domain included the CONUS and SC (Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e). The daily predictions were developed at a 1 km\u003csup\u003e2\u003c/sup\u003e spatial resolution. In total, our modeling grid included 9,115,328 1 km\u003csup\u003e2\u003c/sup\u003e grid cells for the CONUS and 2,620,640 1 km\u003csup\u003e2\u003c/sup\u003e grid cells for SC, from 2002 to 2019.\u003c/p\u003e\n\u003c/div\u003e\n\u003ch3\u003eGround-based Carbonaceous PM Measurements\u003c/h3\u003e\n\u003cp\u003eThe ground-based measurements of carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e (i.e., OC and EC) in the CONUS were obtained from a variety of sources, including the chemical speciation network (CSN) from Environmental Protection Agency (EPA), the National Park Service interagency monitoring of protected visual environments (IMPROVE) network, and the Southeastern Aerosol Research and Characterization Study (SEARCH) network (\u003cspan class=\"CitationRef\"\u003e64\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e65\u003c/span\u003e). For SC, ground-based observations were provided by the National Air Pollution Surveillance (NAPS) program (\u003cspan class=\"CitationRef\"\u003e66\u003c/span\u003e). In total, 401 monitors operated during our study period, with most stations operating for more than 5 years (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e). Extremely high observations (\u0026gt;\u0026thinsp;99.98th percentile) of PM\u003csub\u003e2.5\u003c/sub\u003e OC and EC were excluded from the model training to minimize the outliers (99.98th percentile of OC and EC: 34.75 \u0026micro;g/m\u003csup\u003e3\u003c/sup\u003e and 7.40 \u0026micro;g/m\u003csup\u003e3\u003c/sup\u003e). PM\u003csub\u003e2.5\u003c/sub\u003e OC and EC stations are often co-located and take samples synchronously, with similar number of stations and observations overtime (\u003cstrong\u003eTable S10\u003c/strong\u003e). Our final data includes approximately 448,000 daily measurements for OC or EC. Detailed summary statistics of ground-based measurements for training dataset are provided in \u003cstrong\u003eTable S11\u003c/strong\u003e and \u003cstrong\u003eTable S12\u003c/strong\u003e.\u003c/p\u003e\n\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\n \u003ch2\u003eChemical Transport Model\u003c/h2\u003e\n \u003cp\u003eThe state-of-the-art Community Multiscale Air Quality (CMAQ; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ewww.epa.gov/cmaq\u003c/span\u003e\u003c/span\u003e) modeling system estimates atmospheric concentrations of numerous chemicals and aerosols, including ozone (and its precursors), PM\u003csub\u003e2.5\u003c/sub\u003e, and deposition of harmful chemical species (\u003cspan class=\"CitationRef\"\u003e67\u003c/span\u003e). CMAQ is developed and maintained by the US Environmental Protection Agency (EPA), and it has been widely used for assessing air pollution, including evaluating policies and estimating impacts on human health (\u003cspan class=\"CitationRef\"\u003e68\u003c/span\u003e\u0026ndash;\u003cspan class=\"CitationRef\"\u003e70\u003c/span\u003e). In this study, we used CMAQ version 5.3.2 and the datasets from the EPA\u0026rsquo;s air QUAlity TimE Series (EQUATES; \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ewww.epa.gov/cmaq/equates\u003c/span\u003e\u003c/span\u003e) project (\u003cspan class=\"CitationRef\"\u003e71\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eIn this study, we used two sets of daily simulations at 12 km\u003csup\u003e2\u003c/sup\u003e spatial resolution for the period of 2002\u0026ndash;2019: (\u003cspan class=\"CitationRef\"\u003e1\u003c/span\u003e) the baseline CMAQ simulation results from EQUATES, which includes all emission sources, and (\u003cspan class=\"CitationRef\"\u003e2\u003c/span\u003e) a sensitivity simulation conducted using the same configurations (\u003cstrong\u003eTable S13\u003c/strong\u003e) and inputs as in the EQUATES project, except by not including fire emissions. Hereafter, these two sets of simulations are referred to as \u0026ldquo;Smoke-on\u0026rdquo; and \u0026ldquo;Smoke-off\u0026rdquo;, respectively.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\n \u003ch2\u003eClassification of Smoke Impacted Areas\u003c/h2\u003e\n \u003cp\u003eTo better account for the specific contribution of fire smoke to PM\u003csub\u003e2.5\u003c/sub\u003e OC and EC concentrations, we utilized CMAQ simulations run under the different settings. The grid cells in the study domain were classified into two daily smoke scenarios: a background scenario without fire smoke impact (\u0026ldquo;Smoke-off\u0026rdquo;) and a fire smoke-impacted scenario (\u0026ldquo;Smoke-on\u0026rdquo;). The Smoke-on CMAQ model simulations were used in modeling carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e under Smoke-on scenarios, while the Smoke-off CMAQ model simulations were used for Smoke-off scenarios. Detailed classification methods are provided in the \u003cstrong\u003eSupplementary Document S1\u003c/strong\u003e.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\n \u003ch2\u003eAuxiliary Predictors\u003c/h2\u003e\n \u003cp\u003eTo enhance model performance and predictive accuracy, we incorporated a wide range of auxiliary predictors in model development. These predictors included satellite-retrieved aerosol data products, cloud coverage, and smoke plume information, gridded meteorological factors, vegetation coverage, biogenic emissions, population, land cover, road density, topographic data, human footprint, and coordinate and time trend characteristics. These variables have been found to be important predictors in prior studies (\u003cspan class=\"CitationRef\"\u003e36\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e42\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e72\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e73\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eAll predictors at different spatial resolutions were integrated into the 1 km\u003csup\u003e2\u003c/sup\u003e grid cells obtained from the Multi-Angle Implementation of Atmospheric Correction (MAIAC) dataset, which served as the grid template for the MAIAC aerosol optical depth (AOD) measurements (\u003cspan class=\"CitationRef\"\u003e74\u003c/span\u003e). Daily ground-based measurements of PM\u003csub\u003e2.5\u003c/sub\u003e OC and EC were assigned to their collocated grid cells. Detailed descriptions of the data sources and process steps are provided in the \u003cstrong\u003eSupplementary Document S2\u003c/strong\u003e.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\n \u003ch2\u003eModeling framework\u003c/h2\u003e\n \u003cp\u003eAfter aggregating the raw measurements for OC and EC, we applied the Synthetic Minority Oversampling Technique (SMOTE) to oversample the underrepresented high-concentration measurements (\u003cspan class=\"CitationRef\"\u003e75\u003c/span\u003e). The enriched training datasets were then classified into Smoke-off and Smoke-on scenarios. For both PM\u003csub\u003e2.5\u003c/sub\u003e OC and EC, and both smoke scenarios (Smoke-off and Smoke-on), we employed the proposed Residual Adjusted Super Learner (RASL) framework (Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e). This framework operates with a three-level structure based on the super learner method, in which the cross-validated predictions from four base models (first level) were incorporated using a meta-learner algorithm (second level) to generate fused predictions of carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e (\u003cspan class=\"CitationRef\"\u003e76\u003c/span\u003e). Finally, generalized additive models (GAMs; third level) were used to adjust the spatiotemporal residuals of monthly concentrations at each 1km grid cell from 2002 to 2019 across the study domain. Details of the modeling framework are provided in the \u003cstrong\u003eSupplementary Document S3\u003c/strong\u003e.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\n \u003ch2\u003eModel Performance Evaluation\u003c/h2\u003e\n \u003cp\u003eWe conducted a three-stage cross-validation (CV) to evaluate the model performance at each level of the RASL framework. Three types of CV were employed: 10-fold random CV, 10-fold clustered spatial CV, and leave-one-year-out temporal CV. The clustered spatial CV better tests the model\u0026rsquo;s predictive ability when a large group of monitoring networks is missing instead of a single monitor (\u003cspan class=\"CitationRef\"\u003e77\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e78\u003c/span\u003e). Prediction accuracy was evaluated using three metrics: the coefficient of determination (R\u0026sup2;), root-mean-square error (RMSE), and slope. To further evaluate the model at different time scales, we averaged daily estimations into monthly and annual values, allowing us to capture both long-term trends and short-term fluctuations\u0026mdash;important for long-term cohort studies and short-term analyses. Details of the CV experiments are provided in the \u003cstrong\u003eSupplementary Document S4\u003c/strong\u003e.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\n \u003ch2\u003eCalculation of mortality burden and monetized damage\u003c/h2\u003e\n \u003cp\u003eWe employed different methods in the CONUS and SC to calculate the non-accidental mortality and monetized damages attributable to fire smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e. For the CONUS, Ma et al provided the monthly non-accidental mortality rates for different smoke PM\u003csub\u003e2.5\u003c/sub\u003e concentration bins in the CONUS (\u003cspan class=\"CitationRef\"\u003e20\u003c/span\u003e), which we then multiplied by 12 months to estimate the annual mortality rate (\u003cstrong\u003eTable S14\u003c/strong\u003e). The annual mortality rate, along with each year\u0026rsquo;s smoke concentrations and population data, was used to estimate the following year\u0026rsquo;s deaths. To assess the monetized damages associated with these mortality estimates, we employed the Value of Statistical Life (VSL), as provided by the U.S. Department of Health and Human Services. VSLs reflect the monetary value that individuals are willing to pay to reduce the risk of death, thereby providing an economic perspective on mortality burden. We based our estimates on the 2013 VSL value and then adjusted it annually from 2003 to 2020 in accordance with HHS guidelines (\u003cstrong\u003eTable S15\u003c/strong\u003e), to account for inflation and changes in real income for the specific dollar year (\u003cspan class=\"CitationRef\"\u003e79\u003c/span\u003e). The year-specific VSL values were multiplied by the estimated mortality attributable to fire smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e exposure in that year to provide an estimate of annual monetized damages.\u003c/p\u003e\n \u003cp\u003eFor SC, we applied the Air Quality Benefits Assessment Tool (AQBAT) developed by Health Canada, which is designed to estimate the human health impacts and economic valuation of changes in Canada\u0026rsquo;s ambient air quality (\u003cspan class=\"CitationRef\"\u003e80\u003c/span\u003e). AQBAT provided the concentration response functions between chronic exposure to fire smoke PM\u003csub\u003e2.5\u003c/sub\u003e and mortality and, which enabled us to estimate the number of attributable mortality and related monetized damages across Canada (\u003cstrong\u003eTable S9\u003c/strong\u003e).\u003c/p\u003e\n\u003c/div\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAcknowledgements\u003c/h2\u003e \u003cp\u003eResearch reported in this publication was supported by the National Institute of Environmental Health Sciences of the National Institutes of Health under Award Number 1R01ES034175. Dr. K. Chen received support from the National Heart, Lung, and Blood Institute of the National Institutes of Health (R01HL169171). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.\u003c/p\u003e\n\u003ch2\u003eAuthor Contributions: \u003c/h2\u003e\n\u003cp\u003eZ.J. conducted the model development, predictions, and drafted the manuscript. G.F. and J.F. conducted the CMAQ simulations. D.Z. contributed to the dataset preparation. N.S. contributed to the monetized damages estimate and edited the manuscript. K.C. provided the mortality rate results. Y.L. conceived of and supervised the conduct of this study and edited the manuscript.\u003c/p\u003e\n\u003ch2\u003eCompeting Interest Statement: \u003c/h2\u003e\n\u003cp\u003eThe authors declare no conflict of interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eFernandez-Anez N et al (2021) Current wildland fire patterns and challenges in Europe: A synthesis of national perspectives. Air Soil Water Res 14:11786221211028185\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSenande-Rivera M, Insua-Costa D, Miguez-Macho G (2022) Spatial and temporal expansion of global wildland fire activity in response to climate change. Nat Commun 13:1208\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAbatzoglou JT, Williams AP (2016) Impact of anthropogenic climate change on wildfire across western US forests. \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e 113, 11770\u0026ndash;11775\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBurke M et al (2021) The changing risk and burden of wildfire in the United States. \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e 118, e2011048118\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFernandes PM (2015) Empirical support for the use of prescribed burning as a fuel treatment. Curr Forestry Rep 1:118\u0026ndash;127\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFowler C, Konopik E (2007) The history of fire in the southern United States. Hum Ecol Rev, 165\u0026ndash;176\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMelvin M (2018) 2018 National prescribed fire use survey report. Coalition of Prescribed Fire Councils, Inc., Newton, GA)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrey SJ, Barnes EA, Pierce JR, Wiedinmyer C, Fischer EV (2018) Environmental conditions, ignition type, and air quality impacts of wildfires in the southeastern and western United States. Earths Future 6:1442\u0026ndash;1456\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHaikerwal A et al (2015) Impact of smoke from prescribed burning: Is it a public health concern? J Air Waste Manag Assoc 65:592\u0026ndash;598\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJaffe DA et al (2020) Wildfire and prescribed burning impacts on air quality in the United States. J Air Waste Manag Assoc 70:583\u0026ndash;615\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu R et al (2020) Global Climate Change, and Human Health. N Engl J Med 383:2173\u0026ndash;2181\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJaffe D, Hafner W, Chand D, Westerling A, Spracklen D (2008) Interannual variations in PM\u003csub\u003e2.5\u003c/sub\u003e due to wildfires in the Western United States. Environ Sci Technol 42:2812\u0026ndash;2818\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSullivan A et al (2008) A method for smoke marker measurements and its potential application for determining the contribution of biomass burning from wildfires and prescribed fires to ambient PM\u003csub\u003e2.5\u003c/sub\u003e organic carbon. J Geophys Research: Atmos 113\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang Q et al (2007) Ubiquity and dominance of oxygenated species in organic aerosols in anthropogenically-influenced Northern Hemisphere midlatitudes. Geophys Res Lett 34\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBond TC et al (2013) Bounding the role of black carbon in the climate system: A scientific assessment. J Geophys research: Atmos 118:5380\u0026ndash;5552\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOrellano P, Reynoso J, Quaranta N, Bardach A, Ciapponi A (2020) Short-term exposure to particulate matter (PM\u003csub\u003e10\u003c/sub\u003e and PM\u003csub\u003e2.5\u003c/sub\u003e), nitrogen dioxide (NO\u003csub\u003e2\u003c/sub\u003e), and ozone (O\u003csub\u003e3\u003c/sub\u003e) and all-cause and cause-specific mortality: Systematic review and meta-analysis. Environ Int 142:105876\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang L, Wilson JP, Zhao N, Zhang W, Wu Y (2022) The dynamics of cardiovascular and respiratory deaths attributed to long-term PM\u003csub\u003e2.5\u003c/sub\u003e exposures in global megacities. Sci Total Environ 842:156951\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePun VC, Kazemiparkouhi F, Manjourides J, Suh HH (2017) Long-term PM\u003csub\u003e2.5\u003c/sub\u003e exposure and respiratory, cancer, and cardiovascular mortality in older US adults. Am J Epidemiol 186:961\u0026ndash;969\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi Z et al (2019) Impact of ambient PM\u003csub\u003e2.5\u003c/sub\u003e on adverse birth outcome and potential molecular mechanism. Ecotoxicol Environ Saf 169:248\u0026ndash;254\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMa Y et al (2024) Long-term exposure to wildland fire smoke PM\u003csub\u003e2.5\u003c/sub\u003e and mortality in the contiguous United States. \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e 121, e2403960121\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYe T et al (2022) Short-term exposure to wildfire-related PM\u003csub\u003e2.5\u003c/sub\u003e increases mortality risks and burdens in Brazil. Nat Commun 13:7651\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEisenman DP, Galway LP (2022) The mental health and well-being effects of wildfire smoke: a scoping review. BMC Public Health 22:2274\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKaranasiou A et al (2021) Short-term health effects from outdoor exposure to biomass burning emissions: A review. Sci Total Environ 781:146739\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhu Q et al (2024) Wildfires are associated with increased emergency department visits for anxiety disorders in the western United States. Nat Mental Health 2:379\u0026ndash;387\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDanesh Yazdi M et al (2024) Long-term exposure to PM\u003csub\u003e2.5\u003c/sub\u003e species and all-cause mortality among Medicare patients using mixtures analyses. Environ Res 246:118175\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHvidtfeldt UA et al (2019) Long-term residential exposure to PM\u003csub\u003e2.5\u003c/sub\u003e constituents and mortality in a Danish cohort. Environ Int 133:105268\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang Y et al (2022) Long-term exposure to PM\u003csub\u003e2.5\u003c/sub\u003e major components and mortality in the southeastern United States. Environ Int 158:106969\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen J et al (2017) A review of biomass burning: Emissions and impacts on air quality, health and climate in China. Sci Total Environ 579:1000\u0026ndash;1034\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen H, Samet JM, Bromberg PA, Tong H (2021) Cardiovascular health impacts of wildfire smoke exposure. Part Fibre Toxicol 18:2\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAzevedo JM, Gon\u0026ccedil;alves FLT, de F\u0026aacute;tima M, Andrade (2011) Long-range ozone transport and its impact on respiratory and cardiovascular health in the north of Portugal. Int J Biometeorol 55:187\u0026ndash;202\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarlier ME et al (2013) El Ni\u0026ntilde;o and health risks from landscape fire emissions in southeast Asia. Nat Clim Change 3:131\u0026ndash;136\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eReid CE et al (2016) Critical Review of Health Impacts of Wildfire Smoke Exposure. Environ Health Perspect 124:1334\u0026ndash;1343\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKiser D et al (2020) Particulate matter and emergency visits for asthma: a time-series study of their association in the presence and absence of wildfire smoke in Reno, Nevada, 2013\u0026ndash;2018. Environ Health 19:92\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu JC et al (2017) Wildfire-specific fine particulate matter and risk of hospital admissions in urban and rural counties. Epidemiol (Cambridge Mass) 28:77\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCleland SE et al (2020) Estimating wildfire smoke concentrations during the October 2017 California fires through BME space/time data fusion of observed, modeled, and satellite-derived PM\u003csub\u003e2.5\u003c/sub\u003e. Environ Sci Technol 54:13439\u0026ndash;13447\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang D et al (2023) Wildland Fires Worsened Population Exposure to PM\u003csub\u003e2.5\u003c/sub\u003e Pollution in the Contiguous United States. Environ Sci Technol 57:19990\u0026ndash;19998\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi Y et al (2021) Dominance of wildfires impact on air quality exceedances during the 2020 record-breaking wildfire season in the United States. \u003cem\u003eGeophysical Research Letters\u003c/em\u003e 48, e2021GL094908\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePan X et al (2020) Six global biomass burning emission datasets: intercomparison and application in one global aerosol model. Atmos Chem Phys 20:969\u0026ndash;994\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCummins K et al (2023) The Southeastern U.S. Prescribed Fire Permit Database: Hot Spots and Hot Moments in Prescribed Fire across the Southeastern U.S.A. Fire 6:372\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRogers HM, Ditto JC, Gentner DR (2020) Evidence for impacts on surface-level air quality in the northeastern US from long-distance transport of smoke from North American fires during the Long Island Sound Tropospheric Ozone Study (LISTOS) 2018. Atmos Chem Phys 20:671\u0026ndash;682\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang Z et al (2024) Intensification of mid-latitude cyclone by aerosol‐radiation interaction increases transport of Canadian wildfire smoke to northeastern US. \u003cem\u003eGeophysical Research Letters\u003c/em\u003e 51, e2024GL108444\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChilds ML et al (2022) Daily Local-Level Estimates of Ambient Wildfire Smoke PM\u003csub\u003e2.5\u003c/sub\u003e for the Contiguous US. Environ Sci Technol 56:13607\u0026ndash;13621\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMunoz-Alpizar R et al (2013\u0026ndash;2016) Multi-Year PM\u003csub\u003e2.5\u003c/sub\u003e Wildfire Pollution Exposure over North America as Determined from Operational Air Quality Forecasts. \u003cem\u003eAtmosphere\u003c/em\u003e 8, 179 (2017)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAssociation AL (2020) State of the Air 2020\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBurke M et al (2023) The contribution of wildfire to PM\u003csub\u003e2.5\u003c/sub\u003e trends in the USA. Nature 622:761\u0026ndash;766\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhai J, Ning Z, Dahal R, Yang S (2023) Wildfire Susceptibility of Land Use and Topographic Features in the Western United States: Implications for the Landscape Management. Forests 14:807\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilliams AP et al (2019) Observed impacts of anthropogenic climate change on wildfire in California. Earths Future 7:892\u0026ndash;910\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJain P, Castellanos-Acuna D, Coogan SC, Abatzoglou JT, Flannigan MD (2022) Observed increases in extreme fire weather driven by atmospheric humidity and temperature. Nat Clim Change 12:63\u0026ndash;70\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHalofsky JE, Peterson DL, Harvey BJ (2020) Changing wildfire, changing forests: the effects of climate change on fire regimes and vegetation in the Pacific Northwest, USA. Fire Ecol 16:1\u0026ndash;26\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWesterling AL, Hidalgo HG, Cayan DR, Swetnam TW (2006) Warming and earlier spring increase western US forest wildfire activity. \u003cem\u003escience\u003c/em\u003e 313, 940\u0026ndash;943\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu G, Feng Y, Wang J, Wright DB (2023) Performance of Fire Danger Indices and Their Utility in Predicting Future Wildfire Danger Over the Conterminous United States. \u003cem\u003eEarth's Future\u003c/em\u003e 11, eEF003823 (2023)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDi Virgilio G et al (2019) Climate change increases the potential for extreme wildfires. Geophys Res Lett 46:8517\u0026ndash;8526\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu Y et al (2021) Projection of future wildfire emissions in western USA under climate change: contributions from changes in wildfire, fuel loading and fuel moisture. Int J Wildland Fire 31:1\u0026ndash;13\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTian X, Zhao F, Shu L, Wang M (2013) Distribution characteristics and the influence factors of forest fires in China. For Ecol Manag 310:460\u0026ndash;467\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eValk\u0026oacute; O, De\u0026aacute;k B (2021) Increasing the potential of prescribed burning for the biodiversity conservation of European grasslands. Curr Opin Environ Sci Health 22:100268\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeir JR, Scasta JD (2022) Global application of prescribed fire. Csiro Publishing\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKolden CA We\u0026rsquo;re Not Doing Enough Prescribed Fire in the Western United States to Mitigate Wildfire Risk. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://dx.doi.org/10.3390/fire2020030\u003c/span\u003e\u003cspan address=\"10.3390/fire2020030\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEngebretson JM et al (2016) Characterizing public tolerance of smoke from wildland fires in communities across the United States. J Forest 114:601\u0026ndash;609\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eU. S. D. o. t. Interior (2024) Budget Justifications and Performance Information Fiscal Year 2024: WILDLAND FIRE MANAGEMENT\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWiedinmyer C, Hurteau MD (2010) Prescribed fire as a means of reducing forest carbon emissions in the western United States. Environ Sci Technol 44:1926\u0026ndash;1932\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eArkle RS, Pilliod DS, Welty JL (2012) Pattern and process of prescribed fires influence effectiveness at reducing wildfire severity in dry coniferous forests. For Ecol Manag 276:174\u0026ndash;184\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTolhurst KG, McCarthy G (2016) Effect of prescribed burning on wildfire severity: a landscape-scale case study from the 2003 fires in Victoria. Australian Forestry 79:1\u0026ndash;14\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWilliamson GJ, Bowman DMJS, Price OF, Henderson SB, Johnston FH (2016) A transdisciplinary approach to understanding the health effects of wildfire and prescribed fire smoke regimes. Environ Res Lett 11:125009\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSolomon PA et al (2014) US national PM\u003csub\u003e2.5\u003c/sub\u003e chemical speciation monitoring networks\u0026mdash;CSN and IMPROVE: description of networks. J Air Waste Manag Assoc 64:1410\u0026ndash;1438\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHansen DA et al (2003) The southeastern aerosol research and characterization study: Part 1\u0026mdash;Overview. J Air Waste Manag Assoc 53:1460\u0026ndash;1471\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDabek-Zlotorzynska E et al (2011) Canadian National Air Pollution Surveillance (NAPS) PM\u003csub\u003e2.5\u003c/sub\u003e speciation program: Methodology and PM\u003csub\u003e2.5\u003c/sub\u003e chemical composition for the years 2003\u0026ndash;2008. Atmos Environ 45:673\u0026ndash;686\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAppel KW et al (2021) The Community Multiscale Air Quality (CMAQ) model versions 5.3 and 5.3.1: system updates and evaluation. Geosci Model Dev 14:2867\u0026ndash;2897\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang L et al (2015) Source apportionment of PM\u003csub\u003e2.5\u003c/sub\u003e in top polluted cities in Hebei, China using the CMAQ model. Atmos Environ 122:723\u0026ndash;736\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThongthammachart T et al (2021) An integrated model combining random forests and WRF/CMAQ model for high accuracy spatiotemporal PM\u003csub\u003e2.5\u003c/sub\u003e predictions in the Kansai region of Japan. Atmos Environ 262:118620\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang Q, Xue D, Liu X, Gong X, Gao H (2019) Process analysis of PM\u003csub\u003e2.5\u003c/sub\u003e pollution events in a coastal city of China using CMAQ. J Environ Sci 79:225\u0026ndash;238\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eU. E. O. o. R. a. Development. (2020) CMAQ. 5.3.2. in \u003cem\u003eZenodo\u003c/em\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMeng X et al (2018) Estimating PM\u003csub\u003e2.5\u003c/sub\u003e speciation concentrations using prototype 4.4 km-resolution MISR aerosol properties over Southern California. Atmos Environ 181:70\u0026ndash;81\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWei J et al (2023) Separating Daily 1 km PM\u003csub\u003e2.5\u003c/sub\u003e Inorganic Chemical Composition in China since 2000 via Deep Learning Integrating Ground, Satellite, and Model Data. Environ Sci Technol 57:18282\u0026ndash;18295\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLyapustin A, Wang Y, Korkin S, Huang D (2018) MODIS Collection 6 MAIAC algorithm. Atmos Meas Tech 11:5741\u0026ndash;5765\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChawla NV, Bowyer KW, Hall LO (2002) Kegelmeyer, SMOTE: synthetic minority over-sampling technique. J Artif Intell Res 16:321\u0026ndash;357\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan der Laan MJ, Polley EC (2007) A. E. Hubbard, Super learner. Stat Appl Genet Mol Biol 6\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGeng G, Meng X, He K, Liu Y (2020) Random forest models for PM\u003csub\u003e2.5\u003c/sub\u003e speciation concentrations using MISR fractional AODs. Environ Res Lett 15:034056\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePloton P et al (2020) Spatial validation reveals poor predictive performance of large-scale ecological mapping models. Nat Commun 11:4540\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eT. U. S. D. o. H. a. H. Services (2024) HHS Standard Values for Regulatory Analysis, (2024) (Aaron Kearsley)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJudek S, Stieb D, Jovic B, Edwards B (2012) Air Quality Benefits Assessment Tool (AQBAT) user guide: Version 2. Health Canada, Ottawa, Ontario\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"Emory University","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"fire smoke PM2.5, mortality, carbonaceous PM2.5, wildland fire, prescribed fire","lastPublishedDoi":"10.21203/rs.3.rs-5478994/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5478994/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eDespite emerging evidence on the health impacts of fine particulate matter (PM\u003csub\u003e2.5\u003c/sub\u003e) from wildland fire smoke, the specific effects of PM\u003csub\u003e2.5\u003c/sub\u003e composition on health outcomes remain uncertain. We developed a three-level, chemical transport model-based framework to estimate daily full-coverage concentrations of smoke-derived carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e, specifically Organic Carbon (OC) and Elemental Carbon (EC), at a 1 km\u003csup\u003e2\u003c/sup\u003e spatial resolution from 2002 to 2019 across the contiguous U.S. (CONUS) and Southern Canada (SC). Cross-validation demonstrated that the framework performed well at both the daily and monthly levels. Modeling results indicated that increases in wildland fire smoke have offset approximately one-third of the improvements in background air quality. In recent years, wildland fire smoke has become more frequent and carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e concentrations have intensified, especially in the Western CONUS and Southwestern Canada. Smoke exposure is also occurring earlier throughout the year, leading to more population being exposed. We estimated that long-term exposure to fire smoke carbonaceous PM\u003csub\u003e2.5\u003c/sub\u003e is responsible for 7,462 and 259 non-accidental deaths annually in the CONUS and SC, respectively, with associated annual monetized damage of 68.4\u0026nbsp;billion USD for the CONUS and 1.97\u0026nbsp;billion CAD for SC. The Southeastern CONUS, where prescribed fires are prevalent, contributed most to these health impacts and monetized damages. Given the challenges posed by climate change for managing prescribed and wildland fires, our findings offer critical insights to inform policy development and assess future health burdens associated with fire smoke exposure.\u003c/p\u003e","manuscriptTitle":"Fire Smoke Elevated the Carbonaceous PM2.5 Concentration and Mortality Burden in the Contiguous U.S. and Southern Canada","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-11-21 09:00:34","doi":"10.21203/rs.3.rs-5478994/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"4c7ee1c5-16a7-4a44-acbd-c481ae70dd4c","owner":[],"postedDate":"November 21st, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2024-11-21T09:00:34+00:00","versionOfRecord":[],"versionCreatedAt":"2024-11-21 09:00:34","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-5478994","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5478994","identity":"rs-5478994","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}