Exposure to ambient air pollution, including ozone and fine particulate matter (PM2.5), is the world's leading environmental health risk factor. Estimating how this burden may change in the future depends on projecting population growth and age structure as well as understanding how future meteorological changes may impact the production and removal of pollutants from the atmosphere. The net impact of these factors on a global scale has not been well-characterized. Here, we leverage recent meteorology, exposure, and mortality output from general circulation, atmospheric chemistry, and health impact models to isolate how changes in meteorology and populations will impact future global air-pollution-related mortality and the associated monetized impacts by the degree of global temperature change. In contrast to previous studies, we estimate that changes in meteorologically driven air pollution, in the absence of pollutant precursor emission changes, will result in 180 000 fewer deaths annually by 2100 relative to current levels, an annual monetized benefit of $7.3 trillion. Reductions are driven by decreases in PM2.5-attributable mortality in populated regions but are substantially offset by global increases in ozone-related mortality. We also highlight striking regional differences in the sign of net pollutant impacts by 2100, with net pollution decreases in the Northern Hemisphere driven by reductions in nitrate aerosol, while increases in both ozone and organic aerosol at higher temperatures lead to net increases in pollutant impacts in the Southern Hemisphere. Lastly, we assess sensitivities of these results to meteorological projections, health impact functions, and 10 000 future warming scenarios.

Exposure to ambient air pollution, including ozone and fine particulate matter (PM2.5), is the world's leading environmental health risk factor. Estimating how this burden may change in the future depends on projecting population growth and age structure as well as understanding how future meteorological changes may impact the production and removal of pollutants from the atmosphere. The net impact of these factors on a global scale has not been well-characterized. Here, we leverage recent meteorology, exposure, and mortality output from general circulation, atmospheric chemistry, and health impact models to isolate how changes in meteorology and populations will impact future global air-pollution-related mortality and the associated monetized impacts by the degree of global temperature change. In contrast to previous studies, we estimate that changes in meteorologically driven air pollution, in the absence of pollutant precursor emission changes, will result in 180 000 fewer deaths annually by 2100 relative to current levels, an annual monetized benefit of $7.3 trillion. Reductions are driven by decreases in PM2.5-attributable mortality in populated regions but are substantially offset by global increases in ozone-related mortality. We also highlight striking regional differences in the sign of net pollutant impacts by 2100, with net pollution decreases in the Northern Hemisphere driven by reductions in nitrate aerosol, while increases in both ozone and organic aerosol at higher temperatures lead to net increases in pollutant impacts in the Southern Hemisphere. Lastly, we assess sensitivities of these results to meteorological projections, health impact functions, and 10 000 future warming scenarios.

Background Climate actions targeting combustion sources can generate large ancillary health benefits via associated air-quality improvements. Therefore, understanding the health costs associated with ambient fine particulate matter (PM2·5) from combustion sources can guide policy design for both air pollution and climate mitigation efforts. Methods In this modelling study, we estimated the health costs attributable to ambient PM2·5 from six major combustion sources across 204 countries using updated concentration–response models and an age-adjusted valuation method. We defined major combustion sources as the sum of total coal, liquid fuel and natural gas, solid biofuel, agricultural waste burning, other fires, and 50% of the anthropogenic fugitive, combustion, and industrial dust source. Findings Global long-term exposure to ambient PM2·5 from combustion sources imposed US$1·1 (95% uncertainty interval 0·8–1·5) trillion in health costs in 2019, accounting for 56% of the total health costs from all PM2·5 sources. Comparing source contributions to PM2·5 concentrations and health costs, we observed a higher share of health costs from combustion sources compared to their contribution to population-weighted PM2·5 concentration across 134 countries, accounting for more than 87% of the global population. This disparity was primarily attributed to the non-linear relationship between PM2·5 concentration and its associated health costs. Globally, phasing out fossil fuels can generate 23% higher relative health benefits compared to their share of PM2·5 reductions. Specifically, the share of health costs for total coal was 36% higher than the source's contributions to corresponding PM2·5 concentrations and the share of health costs for liquid fuel and natural gas was 12% higher. Other than fossil fuels, South Asia was expected to show 16% greater relative health benefits than the percentage reduction in PM2·5 from the abatement of solid biofuel emissions. Interpretation In most countries, targeting combustion sources might offer greater health benefits than non-combustion sources. This finding provides additional rationale for climate actions aimed at phasing out combustion sources, especially those related to fossil fuels and solid biofuel. Mitigation efforts designed according to source-specific health costs can more effectively avoid health costs than strategies that depend solely on the source contributions to overall PM2·5 concentration. Funding The Health Effects Institute, the National Natural Science Foundation of China, and NASA.

Developing future climate projections begins with choosing future emissions scenarios. While scenarios are often based on storylines, here instead we produce a probabilistic multi-million-member ensemble of radiative forcing trajectories to assess the relevance of future forcing thresholds. We coupled a probabilistic database of future greenhouse gas emission scenarios with a probabilistically calibrated reduced complexity climate model. In 2100, we project median forcings of 5.1 watt per square meters (5th to 95th percentiles of 3.3 to 7.1), with roughly 0.5% probability of exceeding 8.5 watt per square meters, and a 1% probability of being lower than 2.6 watt per square meters. Although the probability of 8.5 watt per square meters scenarios is low, our results support their continued utility for calibrating damage functions, characterizing climate in the 22(nd) century (the probability of exceeding 8.5 watt per square meters increases to about 7% by 2150), and assessing low-probability/high-impact futures.

Over the last decades, air pollution emissions have decreased substantially; however, inequities in air pollution persist. We evaluate county-level racial/ethnic and socioeconomic disparities in emissions changes from six air pollution source sectors (industry [SO 2 ], energy [SO 2 , NO x ], agriculture [NH 3 ], commercial [NO x ], residential [particulate organic carbon], and on-road transportation [NO x ]) in the contiguous United States during the 40 years following the Clean Air Act (CAA) enactment (1970-2010). We calculate relative emission changes and examine the differential changes given county demographics using hierarchical nested models. The results show racial/ethnic disparities, particularly in the industry and energy generation source sectors. We also find that median family income is a driver of variation in relative emissions changes in all sectors—counties with median family income >$75 K vs. less generally experience larger relative declines in industry, energy, transportation, residential, and commercial-related emissions. Emissions from most air pollution source sectors have, on a national level, decreased following the United States CAA. In this work, we show that the relative reductions in emissions varied across racial/ethnic and socioeconomic groups.

Background Climate actions targeting combustion sources can generate large ancillary health benefits via associated air-quality improvements. Therefore, understanding the health costs associated with ambient fine particulate matter (PM25) from combustion sources can guide policy design for both air pollution and climate mitigation efforts. Methods In this modelling study, we estimated the health costs attributable to ambient PM25 from six major combustion sources across 204 countries using updated concentration-response models and an age-adjusted valuation method. We defined major combustion sources as the sum of total coal, liquid fuel and natural gas, solid biofuel, agricultural waste burning, other fires, and 50% of the anthropogenic fugitive, combustion, and industrial dust source. Findings Global long-term exposure to ambient PM25 from combustion sources imposed US$11 (95% uncertainty interval 08-15) trillion in health costs in 2019, accounting for 56% of the total health costs from all PM25 sources. Comparing source contributions to PM25 concentrations and health costs, we observed a higher share of health costs from combustion sources compared to their contribution to population-weighted PM25 concentration across 134 countries, accounting for more than 87% of the global population. This disparity was primarily attributed to the non-linear relationship between PM25 concentration and its associated health costs. Globally, phasing out fossil fuels can generate 23% higher relative health benefits compared to their share of PM25 reductions. Specifically, the share of health costs for total coal was 36% higher than the source's contributions to corresponding PM25 concentrations and the share of health costs for liquid fuel and natural gas was 12% higher. Other than fossil fuels, South Asia was expected to show 16% greater relative health benefits than the percentage reduction in PM25 from the abatement of solid biofuel emissions. Interpretation In most countries, targeting combustion sources might offer greater health benefits than non-combustion sources. This finding provides additional rationale for climate actions aimed at phasing out combustion sources, especially those related to fossil fuels and solid biofuel. Mitigation efforts designed according to source-specific health costs can more effectively avoid health costs than strategies that depend solely on the source contributions to overall PM25 concentration.

Release of gridded greenhouse gas emissions from 2000-2019 based on the 2021_04_21 CEDS release with direct inclusion of point sources as time series. This dataset contains just the speciated VOC emissions covered by CEDS, both at 0.5° and 0.1° spatial resolution. Data for the CEDS primary emission species can be found in the companion dataset, DOI: 10.5281/zenodo.7526345.

We report aircraft observations of extreme levels of HCl and the dihalogens Cl2, Br2, and BrCl in an industrial plume near the Great Salt Lake, Utah. Complete depletion of O3 was observed concurrently with halogen enhancements as a direct result of photochemically produced halogen radicals. Observed fluxes for Cl2, HCl, and NOx agreed with facility-reported emissions inventories. Bromine emissions are not required to be reported in the inventory, but are estimated as 173 Mg year-1 Br2 and 949 Mg year-1 BrCl, representing a major uncounted oxidant source. A zero-dimensional photochemical box model reproduced the observed O3 depletions and demonstrated that bromine radical cycling was principally responsible for the rapid O3 depletion. Inclusion of observed halogen emissions in both the box model and a 3D chemical model showed significant increases in oxidants and particulate matter (PM2.5) in the populated regions of the Great Salt Lake Basin, where winter PM2.5 is among the most severe air quality issues in the U.S. The model shows regional PM2.5 increases of 10%-25% attributable to this single industrial halogen source, demonstrating the impact of underreported industrial bromine emissions on oxidation sources and air quality within a major urban area of the western U.S.

Evidence of the physical and economic impacts of climate change is a critical input to policy development and decision-making. In addition to the magnitude of potential impacts, detailed estimates of where, when, and to whom those damages may occur; the types of impacts that will be most damaging; uncertainties in these damages; and the ability of adaptation to reduce potential risks are all interconnected and important considerations. This study utilizes the reduced-complexity model, the Framework for Evaluating Damages and Impacts (FrEDI), to rapidly project economic and physical impacts of climate change across 10 000 future scenarios for multiple impact sectors, regions, and populations within the contiguous United States (US). Results from FrEDI show that net national damages increase overtime, with mean climate-driven damages estimated to reach USD 2.9 trillion (95 % confidence interval (CI): USD 510 billion to USD 12 trillion) annually by 2090. Detailed FrEDI results show that for the analyzed sectors the majority of annual long-term (e.g., 2090) damages are associated with climate change impacts to human health, including mortality attributable to climate-driven changes in temperature and air pollution (O3 and PM2.5) exposure. Regional results also show that annual long-term climate-driven damages vary geographically. The Southeast (all regions are as defined in Fig. 5) is projected to experience the largest annual damages per capita (mean: USD 9300 per person annually; 95 % CI: USD 1800-USD 37 000 per person annually), whereas the smallest damages per capita are expected in the Southwest (mean: USD 6300 per person annually; 95 % CI: USD 840-USD 27 000 per person annually). Climate change impacts may also broaden existing societal inequalities, with, for example, Black or African Americans being disproportionately affected by additional premature mortality from changes in air quality. Lastly, FrEDI projections are extended through 2300 to estimate the net present climate-driven damages within US borders from marginal changes in greenhouse gas emissions. Combined, this analysis provides the most detailed illustration to date of the distribution of climate change impacts within US borders.

High end emissions and forcing scenarios used to drive climate models provide key insights into possible high-risk climate change impacts in the 21st century and beyond. Given rapid and ongoing societal changes (e.g., population growth, energy demand, technology, etc.), debates are increasing on the continued relevance of higher forcing scenarios, such as those designed to reach 8.5 W/m2 by the end of the century (e.g., RCP8.5 and SSP5-8.5). Here, we determine an updated probability of exceeding 8.5 W/m2 by 2100 by generating a multi-million-member ensemble from a newly developed database of probabilistic greenhouse gas (GHG) emissions scenarios, augmented here with improved characterization of aerosols and minor GHGs, coupled with the newest version of a reduced complexity climate model, with parameter distributions calibrated to match the IPCC 6th Assessment Report. The probability of exceeding 8.5 W/m2 this century is estimated to be less than 1%: however, we discuss important rationales for the continued use of 8.5 W/m2 scenarios, including for generating high temperature scenarios for use in calibrating damage functions, characterizing climate in the 22nd century (probability of exceeding 8.5 W/m2 increases to ~7% by 2150), and assessing low-probability/high-impact futures.

Fine particulate matter (PM2.5) exposure is a leading mortality risk factor in India and the surrounding region of South Asia. This study evaluates the contribution of emission sectors and fuels to PM2.5 mass for 29 states in India and 6 surrounding countries (Pakistan, Bangladesh, Nepal, Bhutan, Sri Lanka, and Myanmar) by combining source-specific emission estimates, stretched grid simulations from a chemical transport model, high resolution hybrid PM2.5, and disease-specific mortality estimates. We find that 1.02 (95% Confidence Interval (CI): 0.78-1.26) million deaths in South Asia attributable to ambient PM2.5 in 2019 were primarily from three leading sectors: residential combustion (28%), industry (15%), and power generation (12%). Solid biofuel is the leading combustible fuel contributing to the PM2.5-attributable mortality (31%), followed by coal (17%), and oil and gas (14%). State-level analyses reveal higher residential combustion contributions (35%-39%) in states (Delhi, Uttar-Pradesh, Haryana) with high ambient PM2.5 (>95 μg/m3). The combined mortality burden associated with residential combustion (ambient) and household air pollution (HAP) in India is 0.72 million (95% CI:0.54-0.89) (68% attributable to HAP, 32% attributable to residential combustion). Our results illustrate the potential to reduce PM2.5 mass and improve population health by reducing emissions from traditional energy sources across multiple sectors in South Asia.

Pandemics greatly affect transportation, economic and household activities and their associated air pollutant emissions. In less affluent regions, household energy use is often the dominant pollution source and is sensitive to the affluence change caused by a persisting pandemic. Air quality studies on COVID-19 have shown declines in pollution levels over industrialized regions as an immediate response to pandemic-caused lockdown and weakened economy. Yet few have considered the response of residential emissions to altered household affluence and energy choice supplemented by social distancing. Here we quantify the potential effects of long-term pandemics on ambient fine particulate matter pollution (PM2.5) and resulting premature mortality worldwide, by comprehensively considering the changes in transportation, economic production and household energy use. We find that a persisting COVID-like pandemic would reduce the global gross domestic product by 10.9 % and premature mortality related to black carbon, primary organic aerosols and secondary inorganic aerosols by 9.5 %. The global mortality decline would reach 13.0 % had the response of residential emissions been excluded. Among the 13 aggregated regions worldwide, the least affluent regions exhibit the greatest fractional economic losses with no comparable magnitudes of mortality reduction. This is because their weakened affluence would cause switch to more polluting household energy types on top of longer stay-at-home time, largely offsetting the effect of reduced transportation and economic production. International financial, technological and vaccine aids could reduce such environmental inequality.

Nationally reported greenhouse gas inventories are a core component of the Paris Agreement transparency framework. Comparisons with emission estimates derived from atmospheric observations help identify improvements to reduce uncertainties and increase confidence in reported values. To facilitate comparisons over the contiguous United States, we present a 0.1°×0.1° gridded inventory of annual 2012-2018 anthropogenic methane emissions, allocated to 26 individual source categories, with scale-dependent error estimates. Our inventory is consistent with the U.S. Environmental Protection Agency (EPA) Inventory of U.S. Greenhouse Gas Emissions and Sinks (GHGI), submitted to the United Nations in 2020. Total emissions and patterns (spatial/temporal) reflect the activity and emission factor data underlying the GHGI, including many updates relative to a previous gridded GHGI that has been extensively compared with observations. These underlying data are not generally available in global gridded inventories, and comparison to EDGAR v6 shows large spatial differences, particularly for the oil and gas sectors. We also find strong regional variability across all sources in annual 2012-2018 spatial trends, highlighting the importance of understanding regional and facility-level activities. Our inventory represents the first time series of gridded GHGI methane emissions and enables robust comparisons of emissions and their trends with atmospheric observations.

Atmospheric methane directly affects surface temperatures and indirectly affects ozone, impacting human welfare, the economy, and environment. The social cost of methane (SC-CH4) metric estimates the costs associated with an additional marginal metric ton of emissions. Current SC-CH4 estimates do not consider the indirect impacts associated with ozone production from changes in methane. We use global model simulations and a new BenMAP webtool to estimate respiratory-related deaths associated with increases in ozone from a pulse of methane emissions in 2020. By using an approach consistent with the current SC-CH4 framework, we monetize and discount annual damages back to present day values. We estimate that the methane-ozone mechanism is attributable to 760 (95% CI: 330-1200) respiratory-related deaths per million metric tons of methane globally, for a global net present damage of $1800/mT (95% CI: $760-$2800/Mt CH4; 2% Ramsey discount rate); this would double the current SC-CH4 if included. These physical impacts are consistent with recent studies, but comparing direct costs is challenging. Economic damages are sensitive to uncertainties in the exposure and health risks associated with tropospheric ozone, assumptions about future projections of NOx emissions, socioeconomic conditions, and mortality rates, monetization parameters, and other factors. Our estimates are highly sensitive to uncertainties in ozone health risks. We also develop a reduced form model to test sensitivities to other parameters. The reduced form tool runs with a user-supplied emissions pulse, as well as socioeconomic and precursor projections, enabling future integration of the methane-ozone mechanism into the SC-CH4 modeling framework.

Ambient fine particulate matter (PM 2.5 ) is the world’s leading environmental health risk factor. Quantification is needed of regional contributions to changes in global PM 2.5 exposure. Here we interpret satellite-derived PM 2.5 estimates over 1998-2019 and find a reversal of previous growth in global PM 2.5 air pollution, which is quantitatively attributed to contributions from 13 regions. Global population-weighted (PW) PM 2.5 exposure, related to both pollution levels and population size, increased from 1998 (28.3 μg/m 3 ) to a peak in 2011 (38.9 μg/m 3 ) and decreased steadily afterwards (34.7 μg/m 3 in 2019). Post-2011 change was related to exposure reduction in China and slowed exposure growth in other regions (especially South Asia, the Middle East and Africa). The post-2011 exposure reduction contributes to stagnation of growth in global PM 2.5 -attributable mortality and increasing health benefits per µg/m 3 marginal reduction in exposure, implying increasing urgency and benefits of PM 2.5 mitigation with aging population and cleaner air.

Assessments of high-forcing climate scenarios provide unique insight into possible high-risk climate change impacts in the 21st century and beyond. Given rapid and ongoing societal changes (e.g., population growth, energy demand, technology, etc.), debates are increasing on the continued relevance of these 'high-forcing' scenarios, such as those designed to reach 8.5 W/m2 by the end of the century (e.g., RCP8.5 and SSP5-8.5). Here, we determine an updated probability of exceeding 8.5 W/m2 by 2100 by generating a multi-million-member ensemble from a newly developed database of probabilistic greenhouse gas (GHG) emissions scenarios, augmented here with improved characterization of aerosols and minor GHGs, coupled with the newest version of a reduced complexity climate model, with parameter distributions calibrated to match the IPCC 6th Assessment Report. The probability of exceeding 8.5 W/m2 this century is estimated to be less than 1%, however, we discuss important rationales for the continued use of 8.5 W/m2 scenarios, including for generating high temperature scenarios for use in calibrating damage functions, characterizing climate in the 22nd century (probability of exceeding 8.5 W/m2 increases to ~7% by 2150), and assessing low-probability/high-impact futures.

Abstract. High levels of fine particulate matter (PM2.5) pollution in East Asia often exceed local air quality standards. Observations from the Korea United States-Air Quality (KORUS-AQ) field campaign in May and June 2016 showed that development of extreme pollution (haze) occurred through a combination of long-range transport and favorable meteorological conditions that enhanced local production of PM2.5. Atmospheric models often have difficulty simulating PM2.5 chemical composition during haze, which is of concern for the development of successful control measures. We use observations from KORUS-AQ to examine the ability of the GEOS-Chem chemical transport model to simulate PM2.5 composition throughout the campaign and identify the mechanisms driving the pollution event. In the surface level, the model underestimates campaign average sulfate aerosol by −64 % but overestimates nitrate aerosol by 36 %. The largest underestimate in sulfate occurs during the pollution event in conditions of high relative humidity, where models typically struggle to generate the high concentrations due to missing heterogeneous chemistry in aerosol liquid water in the polluted boundary layer. Hourly surface observations show that the model nitrate bias is driven by an overestimation of the nighttime peak. In the model, nitrate formation is limited by the supply of nitric acid, which is biased by +100 % against aircraft observations. We hypothesize that this is due to a missing sink, which we implement here as a factor of five increase in dry deposition. We show that the resulting increased deposition velocity is consistent with observations of total nitrate as a function of photochemical age. The model does not account for factors such as the urban heat island effect or the heterogeneity of the built-up urban landscape resulting in insufficient model turbulence and surface area over the study area that likely results in insufficient dry deposition. Other species such as NH3 could be similarly affected but were not measured during the campaign. Nighttime production of nitrate is driven by NO2 hydrolysis in the model, while observations show that unexpectedly elevated nighttime ozone (not present in the model) should result in N2O5 hydrolysis as the primary pathway. The model is unable to represent nighttime ozone due to an overly rapid collapse of the afternoon mixed layer and excessive titration by NO. We attribute this to missing nighttime heating driving deeper nocturnal mixing that would be expected to occur in a city like Seoul. This urban heating is not considered in air quality models run at large enough scales to treat both local chemistry and long-range transport. Key model failures in simulating nitrate, mainly overestimated daytime nitric acid, incorrect representation of nighttime chemistry, and an overly shallow and insufficiently turbulent nighttime mixed layer, exacerbate the model’s inability to simulate the buildup of PM2.5 during haze pollution. To address the underestimate in sulfate most evident during the haze event, heterogeneous aerosol uptake of SO2 is added to the model which previously only considered aqueous production of sulfate from SO2 in cloud water. Implementing a simple parameterization of this chemistry improves the model abundance of sulfate but degrades the SO2 simulation implying that emissions are underestimated. We find that improving model simulations of sulfate has direct relevance to determining local vs. transboundary contributions to PM2.5. During the haze pollution event, the inclusion of heterogeneous aerosol uptake of SO2 decreases the fraction of PM2.5 attributable to long-range transport from 66 % to 54 %. Locally-produced sulfate increased from 1 % to 46 % of locally-produced PM2.5, implying that local emissions controls would have a larger effect than previously thought. However, this additional uptake of SO2 is coupled to the model nitrate prediction which affects the aerosol liquid water abundance and chemistry driving sulfate-nitrate-ammonium partitioning. An additional simulation of the haze pollution with heterogeneous uptake of SO2 to aerosol and simple improvements to the model nitrate simulation results in 30 % less sulfate due to 40 % less nitrate and aerosol water, and results in an underestimate of sulfate during the haze event. Future studies need to better consider the impact of model physical processes such as dry deposition and boundary layer mixing on the simulation of nitrate and the effect of improved nitrate simulations on the overall simulation of secondary inorganic aerosol (sulfate+nitrate+ammonium) in East Asia. Foreign emissions are rapidly changing, increasing the need to understand the impact of local emissions on PM2.5 in South Korea to ensure continued air quality improvements.

. High levels of fine particulate matter (PM 2.5 ) pollution in East Asia often exceed local air quality standards. Observations from the Korea United States-Air Quality (KORUS-AQ) field campaign in May and June 2016 showed that development of extreme 30 pollution (haze) occurred through a combination of long-range transport and favorable meteorological conditions that enhanced local production of PM 2.5 . Atmospheric models often have difficulty simulating PM 2.5 chemical composition during haze, which is of concern for the development of successful control measures. We use observations from KORUS-AQ to examine the ability of the GEOS-Chem chemical transport model to simulate PM 2.5 composition throughout the campaign and identify the mechanisms driving the pollution event. In the surface level, the model underestimates campaign average sulfate aerosol 35 by -64% but overestimates nitrate aerosol by 36%. The largest underestimate in sulfate occurs during the pollution event in conditions of high relative humidity, where models typically struggle to generate the high concentrations due to missing heterogeneous chemistry in aerosol liquid water in the polluted boundary layer. Hourly surface observations show that the model nitrate bias is driven by an overestimation of the nighttime peak. In the model, nitrate formation is limited by the supply of nitric acid, which is biased by +100% against aircraft observations. We hypothesize that this is due to a missing sink, which we implement here as a factor of five increase in dry deposition. We show that the resulting increased deposition velocity is consistent with observations of total nitrate as a function of photochemical age. The model does not account for factors such as the urban heat island effect or the heterogeneity of the built-up urban landscape resulting in insufficient model turbulence and surface area over the study area that likely results in insufficient dry deposition. Other species such as NH 3 could be similarly affected but were not measured during the campaign. Nighttime production of nitrate is driven by NO 2 hydrolysis in 45 the model, while observations show that unexpectedly elevated nighttime ozone (not present in the model) should result in N 2 O 5 hydrolysis as the primary pathway. The model is unable to represent nighttime ozone due to an overly rapid collapse of the afternoon mixed layer and excessive titration by NO. We attribute this to missing nighttime heating driving deeper nocturnal mixing that would be expected to occur in a city like Seoul. This urban heating is not considered in air quality models run at large enough scales to treat both local chemistry and long-range transport. Key model failures in simulating nitrate, 50 mainly overestimated daytime nitric acid, incorrect representation of nighttime chemistry, and an overly shallow and insufficiently turbulent nighttime mixed layer, exacerbate the model’s inability to simulate the buildup of PM 2.5 during haze pollution. To address the underestimate in sulfate most evident during the haze event, heterogeneous aerosol uptake of SO 2 is added to the model which previously only considered aqueous production of sulfate from SO 2 in cloud water. Implementing a simple parameterization of this chemistry improves the model abundance of sulfate but degrades the SO 2 simulation implying that emissions are underestimated. We find that improving model simulations of sulfate has direct relevance to determining local vs. transboundary contributions to PM 2.5 . During the haze pollution event, the inclusion of heterogeneous aerosol uptake of the emissions inventory for key emitted species in the nested East Asia domain for May 2016. Model dry deposition for gas-phase species is based on the resistance-in-series scheme from Wesely (1989) as implemented by Wang et al. (1998), where species deposition is limited by aerodynamic resistance, quasi-laminar layer resistance, and canopy or surface resistance. Species with low surface resistance, such as HNO 3 , are limited in their deposition velocity by aerodynamic resistance only. Aerosol deposition is from Zhang et al. (2001). The original model wet deposition scheme is described by Liu et al. (2001) for water-soluble aerosols and Amos et al. (2012) for gases. Wet deposition includes scavenging from moist convective updrafts and rainout and washout from precipitation. We include the revised wet deposition scheme of Luo et al. (2019) that uses an empirical washout rate for nitric acid two orders of magnitude higher than the previous value and replaces the standard constant value for in-cloud condensation water content with the value calculated by the meteorological fields (GEOS-FP). GEOS-Chem uses a non-local boundary layer mixing scheme (Holtslag and Boville, 1993; Lin and McElroy, 2010) where mixing is calculated explicitly from meteorological variables provided by GEOS-FP (i.e. 160 sensible and latent heat flux, temperature, friction velocity). The mixing height is restricted from dropping below a minimum mechanical mixing depth, defined as a function of local friction velocity (Lin and McElroy, 2010). (SALA) (SALC) (Alexander and in four size the first bin and 38% of the second bin are included in PM 2.5 . The recommended definition of dry PM 2.5 is given by Eq 1.

Background: Air pollution emissions have substantially decreased in the US since the 1970 Clean Air Act. We examined racial and economic disparities in emissions reductions from specific pollution sources. Methods: We evaluated racial/ethnic and economic disparities in emissions reductions from six pollution sectors [industry (SO2), energy (SO2), agriculture (NH3), commercial (NOx), residential (organic carbon), and on-road transport (NOx)] across the contiguous US from 1970–2010. We used the CEDGBD-MAP emissions inventory to estimate the average relative emissions change from one decennial to the following and its link to demographics (percent White, Black, Asian, American Indian Hispanic, family income, property value, poverty, and unemployment) at the county level. We constructed hierarchical models with the relative emissions change as the dependent variable and demographics as independent variables. We modeled each pollutant and demographic variable separately, adjusting for potential confounding, and tested for nonlinearities. Results: On average, national emissions decreased from 1970 to 2010 except in the agriculture and the residential sectors. The largest reductions across the study period were in industry and energy SO2 emissions (-66% and -89%, respectively) and on-road transport NOx (-59%). Although average emissions decreased nationally, trends varied across counties. We found racial disparities in industry, energy, and agriculture emissions. An increase in % American Indian, Hispanic, or Asian was associated with increases in industry SO2, agriculture NH3, and energy NOx. We also found economic disparities: an increase in median family income was associated with decreases in emissions from all sectors except agriculture, and increases in unemployment or poverty were associated with increases in industry and energy SO2 emissions. Conclusion: Although air pollution emissions, on average, decreased since 1970, the trends in emissions varied across counties. On average, low-income and communities of color have experienced smaller improvements or a relative increase in emissions despite the nationwide downward trends.