Atmospheric mercury pollution remains a serious threat for human health and ecosystems, as recognized by the Minamata Convention on Mercury adopted in 2013. Coal-fired power units are major emitters of various environmental pollutants globally, including mercury. Here, we show the historical co-benefits of air pollution control measures and coal power retirement, identify coal-fired power units with disproportionately high mercury emissions, which we define as super emitters, and propose country-specific mitigation strategies. We find that super emitters were primarily concentrated in emerging countries by 2020. Our unit-level analysis highlights that substantial, near-term and cost-effective reductions can be achieved by prioritizing the retirement of super emitters and upgrading pollution control equipment, especially in countries with large coal fleets. In the long term, mercury-specific control technologies will be needed for deeper reductions. Our findings provide insights into customized and stage-based mitigation strategies for phasing down key sources of mercury emissions from coal-fired power units.

The chemical composition of fine particulate matter (PM2.5) critically shapes its impacts on climate, air quality and human health, yet its high-resolution spatiotemporal variability covering continental to global scales remains poorly constrained owing to sparse ground observations. Here we develop a 10-km-resolution global dataset of PM2.5 chemical composition for 2010-2020, including sulfate (SO42-), nitrate (NO3-), ammonium (NH4+), organic matter (OM) and black carbon (BC), using a physically constrained deep transfer learning framework. Model evaluation against surface observations yields high accuracies, with correlation coefficients ranging from 0.80 to 0.91 across compositions. We identify pronounced spatiotemporal heterogeneity in PM2.5 composition distribution and long-term evolution. During the study period, the global reduction in PM2.5 concentration was driven primarily by decreases in SO42-, with Europe and Asia contributing most prominently. The fractional contributions of BC and OM increased significantly and exhibited a sustained upward trend in North America (by 4.72% and 5.86%, respectively) and Africa (by 2.32% and 6.94%), whereas secondary inorganic aerosols declined in all the continents except Africa. Recent studies have reported substantial differences in toxicity among PM2.5 compositions. Composition-specific exposure data therefore enable more accurate assessments of PM2.5-related health risks and underscore the importance of sustained and comprehensive monitoring of PM2.5 composition.

Air pollution remains a leading global health threat, with fine particulate matter (PM2.5) contributing to millions of premature deaths annually. Chemical transport models (CTMs) are essential tools for evaluating how emission controls improve air quality and save lives, but they are computationally intensive. Reduced form models accelerate simulations but sacrifice spatial-temporal granularity, accuracy, and flexibility. Here we present CleanAir, a deep-learning-based model developed as an efficient alternative to CTMs in simulating daily PM2.5 and its chemical compositions in response to precursor emission reductions at 36 km resolution, which could predict PM2.5 concentration for a full year within 10 seconds on a single GPU, a speed five orders of magnitude faster. Built on a Residual Symmetric 3D U-Net architecture and trained on more than 2,400 emission reduction scenarios generated by a well-validated Community Multiscale Air Quality (CMAQ) model, CleanAir generalizes well across unseen meteorological years and emission patterns. It produces results comparable to CMAQ in both absolute concentrations and emission-induced changes, enabling efficient, full-coverage simulations across short-term interventions and long-term planning horizons. This advance empowers researchers and policymakers to rapidly evaluate a wide range of air quality strategies and assess the associated health impacts, thereby supporting more responsive and informed environmental decision-making.

Particulate nitrate is a major component of haze pollution, yet substantial debate exists regarding its effective control policies. Here, we develop a multiphase dissociation equilibrium equation, based on which the influence of meteorological conditions and chemical profiles can be decoupled, and the analytical expression of nitrate sensitivity to different species can be derived. With this framework, four nitrate control regimes are identified, namely the nitric-acid-control, ammonia-control, transition, and meteorology regimes. The framework also explains the trade-offs between sulfate and nitrate, and elucidates the influence of sulfate and non-volatile cations on the regime transitions. Across most Northern Hemisphere continental regions, and nearly all of China, particulate nitrate is generally more sensitive to nitric acid than to ammonia due to the universal NH3 excess. The framework provides transparent criteria for interpreting nitrate sensitivity and control regime classifications, serving as an important scientific basis for nitrate control policies and reactive nitrogen interactions.

Fine particulate matter (PM2.5) is the leading environmental risk factor for premature mortality. Although mass concentrations of PM2.5 have declined in the past decade, the unequal toxicity in different sources remains overlooked in air-quality management. Here, we integrated source-specific toxicities of PM2.5, derived from field measurements and cellular assays, with high-resolution emission inventories to conduct health-oriented source apportionment in the Yangtze River Delta (YRD). The average of toxicity-adjusted population-weighted PM2.5 exposure was high in central and northern Anhui and along the Yangtze River, approximately 6.6 times that of Zhejiang and Shanghai. Residential solid fuel combustion was the dominant contributor to PM2.5 toxicity, especially in winter (74.4%, 95% CI: [62.9%, 85.9%]). The toxicity contribution of residential combustion was negatively correlated with city-level income (r = -0.83; p < 0.0001). Targeting residential emissions effectively reduced relative toxicity-adjusted risks in low-income regions, and the costs for reducing a unit of toxicity-adjusted PM2.5 were 2.7% of those for industrial emission control. Ship emission abatement might yield optimal cost-health benefits in coastal areas based on relative ranking under assumed marginal costs. Our findings, based on in vitro cellular toxicity, highlighted controlling residential and ship emissions, providing insights into tailored air-quality policies and regional health equity.

Organic aerosol (OA) is a major component of tropospheric submicron aerosols, influencing air pollution, human health, and climate change. Primary and secondary OA (POA and SOA) exhibit distinct physicochemical properties that lead to different health and climate impacts. Current chemical transport models (CTMs), however, have difficulties not only in capturing OA concentrations, especially in polluted regions, but also in reproducing the fraction of POA and SOA. Here, we develop an OA simulation framework for full-volatility-range organic precursors, with a particular focus on improving OA formation from semivolatile and low-volatility organic compounds (S/LVOC) and intermediate-volatility organic compounds (IVOC), based on the atmospheric chemical transport model GEOS-Chem. Cooperating with a newly developed bottom-up global anthropogenic emission inventory, MEIC-global-FVOC, the improved OA scheme shows a good model performance when evaluated against a comprehensive dataset of worldwide measurements for OC, POA, and SOA, driven by increased POA formation from S/LVOC and SOA formation from IVOC. The model indicates that several populated regions in Asia, Africa, America, and Europe suffer from high OA exposure with annual mean over 5 μg m-3, highlighting the importance of controlling OA pollution. In East Asia, South Asia, the northern part of Africa, and Europe, anthropogenic SOA and POA are the largest two contributors to OA pollution, suggesting the need for reducing residential combustion that contributes over half of anthropogenic S/LVOC and IVOC emissions. For other regions, most of OA is from natural sources, which may be easily affected by extreme events (e.g., wildfires) and the warming climate. The estimated global OA burden in 2018 is 2.50 Tg, with a fraction of 75% from SOA. The SOA burden is higher than previous estimates, resulting from increased formation of S/LVOC and IVOC, highlighting that the role of SOA should be given more attention in assessing aerosol climate impact. The estimation of OA burden is sensitive to pyrogenic emission estimates and wet deposition parameterization, which need more constraints in future studies.

Abstract Global surface ozone (O3), intensified by climate change, poses increasing health and ecosystem threats. Despite stringent air policies, China’s persistent O3 pollution exemplifies a global challenge intertwined with climate actions reshaping emission pathways. Optimal mitigation remains contentious, primarily due to inconsistent conclusions regarding the sensitivity of summer regional O3 formation. Here we show the path dependency in O3 mitigation strategies for synergistic clean air and climate action goal achievement over multidecade scales. This path dependence is validated by observed concurrent plateaus (2020-2023) in deweathered O3 concentrations and sensitivity trends across Chinese megacity clusters. Leveraging this understanding of path dependency, we quantitatively reveal that the optimal future strategy involves prioritizing early volatile organic compound reductions, as their high effectiveness for regional O3 mitigation gradually diminishes towards 2050. This challenges prevailing nitrogen oxides priority paradigms. Our work reframes O3 control, providing a paradigm for resilient air quality-climate governance.

Full-volatility-range organic compounds (FVOC), including low-volatility, semi-volatile, intermediate-volatility, and non-methane volatile organic compounds (L/S/IVOC and NMVOC), are key components influencing atmospheric chemistry and climate. Accurate estimates of their emissions are crucial for modeling air quality and understanding aerosol budgets. However, the long-term trends of global FVOC emissions remain unquantified. Here, we develop a volatility-bin-resolved global anthropogenic FVOC emission inventory for 1970-2020 under the Multi-resolution Emission Inventory model for Climate and air pollution research (MEIC) framework (MEIC-global-FVOC), integrating measured gas- and particle-phase emission factors. The MEIC-global-FVOC emission inventory fills the gaps in current inventories by capturing IVOC emissions and likely accounting for part of the evaporative fraction of low-volatility and semi-volatile organic compound (L/SVOC) emissions, leading to 28%-41% higher estimates for global organic compound emissions than current inventories. Global FVOC emissions increase from 106.1 Mt to 165.0 Mt during 1970-2020, with higher growth rates of L/SVOC (9.7% per decade) and IVOC (11.5% per decade) than that of NMVOC (8.4% per decade), mainly driven by increased activities in scattered sources such as residential biofuel combustion and volatile chemical products (VCPs) use. Global L/S/IVOC emissions increase from 35.0 Mt to 58.4 Mt during 1970-2020. Regionally, L/S/IVOC emissions increase rapidly in developing regions due to rising biofuel combustion, whereas emissions in developed regions decrease or remain stable under the offsetting effects of improved PM controls, coal phase-out, sustained rural biofuel use, and increased VCPs demand. Despite notable uncertainties, the MEIC-global-FVOC emission inventory provides more complete estimates of global organic compound emissions and a new dataset for advancing chemical transport modeling and understanding air pollution sources.

This study, performed under the umbrella of the Task Force on Hemispheric Transport of Air Pollution (TF-HTAP), responds to the need of the global and regional atmospheric modelling community of having a mosaic emission inventory of air pollutants that conforms to specific requirements: global coverage, long time series, spatially distributed emissions with high time resolution, and a high sectoral resolution. The mosaic approach of integrating official regional emission inventories based on locally reported data, with a global inventory based on a globally consistent methodology, allows modellers to perform simulations of a high scientific quality while also ensuring that the results remain relevant to policymakers. HTAP_v3.2, an ad-hoc global mosaic of anthropogenic inventories, is an update to the HTAP_v3 global mosaic inventory and has been developed by integrating official inventories over specific areas (North America, Europe, Asia including China, Japan and Korea) with the independent Emissions Database for Global Atmospheric Research (EDGAR) inventory for the remaining world regions. The results are spatially and temporally distributed emissions of SO2, NOx, CO, NMVOC, NH3, PM10, PM2.5, Black Carbon (BC), and Organic Carbon (OC), with a spatial resolution of 0.1 × 0.1° and time intervals of months and years covering the period 2000–2020 (https://doi.org/10.5281/zenodo.17086684, Crippa, 2025, https://edgar.jrc.ec.europa.eu/dataset_htap_v32, last access: 27 October 2025). The emissions are further disaggregated to 16 anthropogenic emitting sectors. This paper describes the methodology applied to develop such an emission mosaic, reports on source allocation, differences among existing inventories, and best practices for the mosaic compilation. One of the key strengths of the HTAP_v3.2 emission mosaic is its temporal coverage, enabling the analysis of emission trends over the past two decades. The development of a global emission mosaic over such long time series represents a unique product for global air quality modelling and for better-informed policy making, reflecting the community effort expended by the TF-HTAP to disentangle the complexity of transboundary transport of air pollution.

Accurate and timely CO2 emission inventories are essential for tracking climate change mitigation progress. This study conducts near real-time CO2 emission estimates for China using different timely-updated activity data sources such as annual bulletins and monthly statistics. Comparing the results with emissions estimated by regular statistics, we find that emission estimates relying solely on either monthly statistics or annual bulletins have uncertainties. Prioritizing annual bulletins where available and supplementing with monthly statistics for the most recent year can balance accuracy and timeliness, yielding a median error of 1.3% across different update intervals. However, near real-time estimates of relative changes in annual CO2 emissions bear large uncertainties regardless of the approach. For the six years investigated, near real-time estimates usually failed to capture the trends in annual emissions, indicating that those near real-time approaches are unreliable for estimating changes in CO2 emissions due to notable differences between timely-updated and regular statistics.

Heatwaves and ozone (O3) pollution threaten human and ecosystem health, with their compounding effects particularly severe in cities. While ground-based observations are indicative of urban O3 pollution during heatwaves, limited vertical insights into the intensified and prolonged O3 pollution hinder a comprehensive understanding of the underlying mechanisms and mitigation strategies. Here, leveraging airship vertical measurements and meteorology-chemistry coupled modeling, we reveal that heatwave-reinforced turbulence redistributes precursors vertically, altering photochemical stratification and accelerating O3 production both at the surface and aloft over megacities in China. Stringent emission controls targeting nitrogen oxides could mitigate the heatwave-exacerbated O3 extremes by narrowing the vertical disparity of photochemical sensitivity. Although heatwaves are projected to intensify, emission reductions due to China's carbon neutrality pledge could alleviate urban O3 pollution by 41-47% during heatwaves and help tackle the dual challenges of air pollution and global warming while enhancing the climate resilience of city clusters.

Addressing climate change and air pollution exhibits strong synergy, and the Chinese government is actively promoting the integrated management of these two issues. Since 2019, the China Clean Air Policy Partnership has released annual reports on China’s progress in climate and air pollution governance. These reports track and analyze the challenges and propose solutions for China’s pursuit of carbon neutrality and clean air by developing and monitoring key indicators across five areas. This report is the fourth annual report. Building on previous research, it further refines the collaborative governance monitoring indicator system, including the addition of climate change and extreme weather, atmospheric greenhouse gases, and enhanced efficiency of pollution removal technologies. The report includes the following components: (1) an analysis of the interactions between air pollution and climate change; (2) a discussion of governance systems and practices, with an emphasis on policy implementation and local experiences; (3) coverage of structural changes and emission reduction technologies, including energy and industrial transitions, transportation, low-carbon buildings, carbon capture and storage, and power systems; (4) an overview of atmospheric dynamics and emission pathways, examining emission drivers and offering insights for future coordinated governance; and (5) an evaluation of the health impacts and benefits of joint actions. These efforts underscore China’s commitment to integrated control, resulting in slowed carbon emission growth, improved air quality, and enhanced health benefits.

China faces the challenge of simultaneously improving fine particulate matter (PM 2.5 ) and ozone (O 3 ) air quality while tackling climate change. However, most studies have focused on either the cocontrol of PM 2.5 and O 3 or the synergy between CO 2 reduction and PM 2.5 improvement, leaving comprehensive strategies for tackling all three underexplored. Here, we evaluate various combinations of clean-air and climate policies in China using an integrated framework that consists of a technology-based emission projection model, a goal-oriented measure selection and optimization method, a chemical transport model, and a detailed cost–benefit assessment approach. Our results show that while maintaining current clean-air policies is effective in controlling PM 2.5 , their impact on O 3 remains limited. Maximizing cost-effective end-of-pipe measures targeting NO x and volatile organic compounds can achieve cocontrol of PM 2.5 and O 3 , but remain vulnerable to unfavorable meteorological conditions. Near-term climate policies alone contribute limited improvements in air quality, particularly for O 3 , due to the lower synergy between CO 2 and O 3 precursor emission sources. Only a combined strategy, integrating stringent clean-air actions and ambitious climate mitigation, can achieve the “triple control” of PM 2.5 , O 3 , and CO 2 . Under this strategy, national annual mean PM 2.5 and national O 3 -8 h 90th percentile concentrations are reduced to 19.1 μg m −3 and 126.9 μg m −3 by 2030, respectively, yielding net benefits of over 3,000 billion RMB. Our findings can inform China’s next phase of environmental policies and emission control strategies in developing regions.

China’s successful implementation of two phases of stringent clean air actions from 2013 to 2020 (Phase I: 2013–2017; Phase II: 2018–2020) has substantially reduced PM _2.5 concentration in Beijing–Tianjin–Hebei and its surrounding areas (BTHSA)—one of China’s most polluted regions. However, the specific role of regional transport in this improvement remains unclear. Here, we investigated the drivers of PM _2.5 mitigation in the BTHSA and systematically quantified the contribution of regional transport during 2013–2020, by conducting multi-scenario analysis using a combination of a bottom–up emission inventory and a chemical transport model with an embedded source apportionment module. The simulated regional average PM _2.5 concentration across the BTHSA declined by 56.1% from 2013 to 2020, primarily driven by anthropogenic emission control (78.1%), while the remaining 21.9% explained by meteorological variability. Within the anthropogenic impacts, reductions in local emissions, intra-regional transport, and extra-regional transport accounted for 49.0%, 32.9%, and 18.1%, respectively. Nevertheless, the relative importance of these drivers shifted, with local contributions declining while regional transport influences intensified and surpassed local abatement in Phase II, where intra-regional transport remained dominant, but the influence of extra-regional transport rose markedly to 1.4 times its Phase I level. Spatially, emission reductions in Hebei and Shandong contributed the most to the regional PM _2.5 decline, representing over 50% of the transport-related improvement. Trends in regional contributions to Beijing, the core city of the BTHSA region, suggest the need for dynamically adjusting joint control policies, expanding coordinated mitigation efforts to extra-regional cities in Southern Shandong, as well as to key regions like the Yangtze River Delta. These findings underscore the growing importance of regional transport in air quality improvement and support more adaptive regional collaboration strategies moving forward.

Smoke from extreme wildfires in Canada adversely affected air quality in many regions in 20231,2. Here we use satellite observations, machine learning and a chemical transport model to quantify global and regional PM2.5 (particulate matter less than 2.5 μm in diameter) exposure and human health impacts related to the 2023 Canadian wildfires. We find that the fires increased annual PM2.5 exposure worldwide by 0.17 μg m-3 (95% confidence interval, 0.09-0.26 μg m-3). North America had the largest increase in annual mean exposure (1.08 μg m-3; 0.82-1.34 μg m-3), but there were also increases in Europe (0.41 μg m-3; 0.32-0.50 μg m-3) owing to long-range transport. Annual mean PM2.5 exposure in Canada increased by 3.82 μg m-3 (3.00-4.64 μg m-3). In the USA, the contribution of the Canadian fires to increased PM2.5 was 1.49 μg m-3 (1.22-1.77 μg m-3), four times as large as the contribution from the 2023 wildfires originating in the USA. We find that 354 million (277-421 million) people in North America and Europe were exposed to daily PM2.5 air pollution caused by Canadian wildfires in 2023. We estimate that 5,400 (3,400-7,400) acute deaths in North America and 64,300 (37,800-90,900) chronic deaths in North America and Europe were attributable to PM2.5 exposure to the 2023 Canadian wildfires. Our results highlight the far-reaching PM2.5 pollution and health burden that large wildfires can have in a single year.

As the precursor of ozone and secondary particles, the incomplete knowledge of NOx emission dynamics constrains our ability to fully elucidate air pollution formation. Flux measurements offer new insights into NOx emissions and titration effects. Here, we present eddy covariance flux measurements of NOx, O3 and Ox in urban Beijing during the summer of 2023. The measured NOx flux was positive with the 24-h average of 10.4 ± 10.7 nmol/m2/s, which is at the lower end of the flux reported in urban regions. The low NOx emission strength in urban Beijing may be related to the successful control measures of diesel vehicles, while confirming significant dependence of NOx flux with traffic flow. The discrepancies by a factor of 1.3-14.5 between measured NOx flux and estimates in emission inventories are observed, indicating most emission inventories may fail to accurately characterize NOx emissions in Beijing. Better agreements between flux measurements and satellite retrieval are obtained. In contrast to NOx, predominantly downward fluxes were observed for ozone (O3) and Ox (=O3 + NO2). A multiple linear regression (MLR) method is developed to examine the impacts of NOx emissions on the ozone downward flux, revealing that NOx emissions induce 54 ± 53 % of ozone downward flux average over the observation period. This study demonstrates valuable information on emission strength and chemical transformation provided by eddy covariance flux measurements of NOx, O3 and Ox.

China faces the challenge of simultaneously improving fine particulate matter (PM2.5) and ozone (O3) air quality while tackling climate change. However, most studies have focused on either the cocontrol of PM2.5 and O3 or the synergy between CO2 reduction and PM2.5 improvement, leaving comprehensive strategies for tackling all three underexplored. Here, we evaluate various combinations of clean-air and climate policies in China using an integrated framework that consists of a technology-based emission projection model, a goal-oriented measure selection and optimization method, a chemical transport model, and a detailed cost-benefit assessment approach. Our results show that while maintaining current clean-air policies is effective in controlling PM2.5, their impact on O3 remains limited. Maximizing cost-effective end-of-pipe measures targeting NOx and volatile organic compounds can achieve cocontrol of PM2.5 and O3, but remain vulnerable to unfavorable meteorological conditions. Near-term climate policies alone contribute limited improvements in air quality, particularly for O3, due to the lower synergy between CO2 and O3 precursor emission sources. Only a combined strategy, integrating stringent clean-air actions and ambitious climate mitigation, can achieve the "triple control" of PM2.5, O3, and CO2. Under this strategy, national annual mean PM2.5 and national O3-8 h 90th percentile concentrations are reduced to 19.1 μg m-3 and 126.9 μg m-3 by 2030, respectively, yielding net benefits of over 3,000 billion RMB. Our findings can inform China's next phase of environmental policies and emission control strategies in developing regions.

The city of Urumqi experiences severe air pollution, for which the variations and causes remain unclear. Here, we comprehensively investigate the seasonality, trends, and drivers of fine and coarse particulate matter (PM2.5 and PM2.5-10) and ozone (O3) in Urumqi in the Chinese national context by statistical analysis of surface and satellite observations during 2015-2023. Wintertime PM2.5 in Urumqi is twice as high as that averaged over Eastern China due to intensive emissions and unfavourable topography and meteorology, monthly PM2.5-10 in Urumqi of above 40 mu g m-3 year-round is driven by spring and autumn natural dust episodes and ubiquitous anthropogenic fugitive dust. During 2015-2023, PM2.5 and PM2.5-10 decreased by 34-38% credited to emissions control. However, the maximum daily 8h average (MDA8) O3 in Urumqi increases rapidly at a rate of 2.5 ppbv yr- 1 due to increased volatile organic compounds (VOCs), decreased nitrogen oxides (NOx) in earlier years of 2015-2016, and decreased PM2.5. Impacts from VOCs and NOxemission changes are supported by summertime O3 formation regime transits from VOCs-limited to NOxand VOCs co-limited during 2015-2023 as depicted by surface O3- nitrogen dioxide (NO2) correlations and satellite formaldehyde (HCHO)/NO2 ratios. Responses of PM2.5, PM2.5- 10, O3, and related gases during the 2020 winter and summer epidemic lockdowns conform to our findings for their 2015-2023 trends. This study concludes that joint NOxand VOCs emissions control would be particularly effective in reducing PM2.5 and O3 in Urumqi. This study also provides references for studying air quality in other places with limited observations.