Explore the vast range of titles available.

The evolution of tropospheric ozone from 1850 to 2100 has been studied using data from Phase 6 of the Coupled Model Intercomparison Project (CMIP6). We evaluate long-term changes using coupled atmosphere–ocean chemistry–climate models, focusing on the CMIP Historical and ScenarioMIP ssp370 experiments, for which detailed tropospheric-ozone diagnostics were archived. The model ensemble has been evaluated against a suite of surface, sonde and satellite observations of the past several decades and found to reproduce well the salient spatial, seasonal and decadal variability and trends. The multi-model mean tropospheric-ozone burden increases from 247 ± 36 Tg in 1850 to a mean value of 356 ± 31 Tg for the period 2005–2014, an increase of 44 %. Modelled present-day values agree well with previous determinations (ACCENT: 336 ± 27 Tg; Atmospheric Chemistry and Climate Model Intercomparison Project, ACCMIP: 337 ± 23 Tg; Tropospheric Ozone Assessment Report, TOAR: 340 ± 34 Tg). In the ssp370 experiments, the ozone burden increases to 416 ± 35 Tg by 2100. The ozone budget has been examined over the same period using lumped ozone production (PO3) and loss (LO3) diagnostics. Both ozone production and chemical loss terms increase steadily over the period 1850 to 2100, with net chemical production (PO3-LO3) reaching a maximum around the year 2000. The residual term, which contains contributions from stratosphere–troposphere transport reaches a minimum around the same time before recovering in the 21st century, while dry deposition increases steadily over the period 1850–2100. Differences between the model residual terms are explained in terms of variation in tropopause height and stratospheric ozone burden.

Weaknesses in process representation in chemistry–climate models lead to biases in simulating surface ozone and to uncertainty in projections of future ozone change. We here develop a deep learning model to demonstrate the feasibility of ozone bias correction in a global chemistry–climate model. We apply this approach to identify the key factors causing ozone biases and to correct projections of future surface ozone. Temperature and the related geographic variables latitude and month show the strongest relationship with ozone biases. This indicates that ozone biases are sensitive to temperature and suggests weaknesses in representation of temperature-sensitive physical or chemical processes. Photolysis rates are also an important factor, highlighting the sensitivity of biases to simulated cloud cover and insolation. Atmospheric chemical species such as the hydroxyl radical, nitric acid and peroxyacyl nitrate show strong positive relationships with ozone biases on a regional scale. These relationships reveal the conditions under which ozone biases occur, although they reflect association rather than direct causation. We correct model projections of future ozone under different climate and emission scenarios following the shared socio-economic pathways. We find that changes in seasonal ozone mixing ratios from the present day to the future are generally smaller than those simulated without bias correction, especially in high-emission regions. This suggests that the ozone sensitivity to changing emissions and climate may be overestimated with chemistry–climate models. Given the uncertainty in simulating future ozone, we show that deep learning approaches can provide improved assessment of the impacts of climate and emission changes on future air quality, along with valuable information to guide future model development.

Chemistry-climate models have developed significantly over the decades, yet they still exhibit substantial systematic biases in simulating atmospheric composition due to gaps in our understanding of underlying processes. Building on deep learning's success in different domains, we explore its application to correct surface ozone biases in the state-of-the-art chemistry-climate model UKESM1. Six statistical models have been developed, and the model Transformer outperforms others due to its advanced architecture. A simple weighted ensemble approach is further proved to enhance performance by 14 % over the best single model Transformer, reducing RMSE to 0.69 ppb. Applied to future scenarios (SSP3-7.0 and SSP3-7.0-lowNTCF), the UKESM1 shows a larger overestimation of ozone changes by up to 25 ppb compared to present-day conditions. Despite biases, UKESM1 captures the non-linear ozone sensitivity to precursors, with temperature-sensitive processes identified as a dominant contributor to biases. We highlight that simulations of future surface ozone are likely to become less accurate under a warmer climate. Therefore, the bias correction approaches introduced here have substantial potential to improve the accuracy of ozone impact assessments. These methods are also applicable to other chemistry-climate models, which is critical for informing air quality and climate policy decisions.

The Earth system models (ESMs) that participated in the sixth Coupled Model Intercomparison Project (CMIP6) tend to simulate excessive cooling in surface air temperature (TAS) between 1960 and 1990. The anomalous cooling is pronounced over the Northern Hemisphere (NH) midlatitudes, coinciding with the rapid growth of anthropogenic sulfur dioxide (SO2) emissions, the primary precursor of atmospheric sulfate aerosols. Structural uncertainties between ESMs have a larger impact on the anomalous cooling than internal variability. Historical simulations with and without anthropogenic aerosol emissions indicate that the anomalous cooling in the ESMs is attributed to the higher aerosol burden in these models. The aerosol forcing sensitivity, estimated as the outgoing shortwave radiation (OSR) response to aerosol concentration changes, cannot well explain the diversity of pothole cooling (PHC) biases in the ESMs. The relative contributions to aerosol forcing sensitivity from aerosol–radiation interactions (ARIs) and aerosol–cloud interactions (ACIs) can be estimated from CMIP6 simulations. We show that even when the aerosol forcing sensitivity is similar between ESMs, the relative contributions of ARI and ACI may be substantially different. The ACI accounts for between 64 % and 87 % of the aerosol forcing sensitivity in the models and is the main source of the aerosol forcing sensitivity differences between the ESMs. The ACI can be further decomposed into a cloud-amount term (which depends linearly on cloud fraction) and a cloud-albedo term (which is independent of cloud fraction, to the first order), with the cloud-amount term accounting for most of the inter-model differences.

In this study we evaluate simulated surface SO2 and sulfate (SO42-) concentrations from the United Kingdom Earth System Model (UKESM1) against observations from ground-based measurement networks in the USA and Europe for the period 1987–2014. We find that UKESM1 captures the historical trend for decreasing concentrations of atmospheric SO2 and SO42- in both Europe and the USA over the period 1987–2014. However, in the polluted regions of the eastern USA and Europe, UKESM1 over-predicts surface SO2 concentrations by a factor of 3 while under-predicting surface SO42- concentrations by 25 %–35 %. In the cleaner western USA, the model over-predicts both surface SO2 and SO42- concentrations by factors of 12 and 1.5 respectively. We find that UKESM1’s bias in surface SO2 and SO42- concentrations is variable according to region and season. We also evaluate UKESM1 against total column SO2 from the Ozone Monitoring Instrument (OMI) using an updated data product. This comparison provides information about the model's global performance, finding that UKESM1 over-predicts total column SO2 over much of the globe, including the large source regions of India, China, the USA, and Europe as well as over outflow regions. Finally, we assess the impact of a more realistic treatment of the model's SO2 dry deposition parameterization. This change increases SO2 dry deposition to the land and ocean surfaces, thus reducing the atmospheric loading of SO2 and SO42-. In comparison with the ground-based and satellite observations, we find that the modified parameterization reduces the model's over-prediction of surface SO2 concentrations and total column SO2. Relative to the ground-based observations, the simulated surface SO42- concentrations are also reduced, while the simulated SO2 dry deposition fluxes increase.

Observed trends in tropospheric ozone, an important air pollutant and short‐lived climate forcer (SLCF), are estimated using available surface and ozonesonde profile data for 1993–2019, using a coherent methodology, and compared to modeled trends (1995–2015) from the Arctic Monitoring Assessment Program SLCF 2021 assessment. Increases in observed surface ozone at Arctic coastal sites, notably during winter, and concurrent decreasing trends in surface carbon monoxide, are generally captured by multi‐model median trends. Wintertime increases are also estimated in the free troposphere at most Arctic sites, with decreases during spring months. Winter trends tend to be overestimated by the multi‐model medians. Springtime surface ozone increases in northern coastal Alaska are not simulated while negative springtime trends in northern Scandinavia are not always reproduced. Possible reasons for observed changes and model performance are discussed including decreasing precursor emissions, changing ozone dry deposition, and variability in large‐scale meteorology. Plain Language Summary The Arctic is warming much faster than the rest of the globe due to increases in carbon dioxide, and other trace constituents like ozone, also an air pollutant. However, improved understanding is needed about long‐term changes or trends in Arctic tropospheric ozone. A coherent methodology is used to identify trends in surface and regular profile measurements over the last 20–30 years, and results from six chemistry‐climate models. Increases in observed ozone are found at the surface and in the free troposphere during winter in the high Arctic. Paradoxically, decreases in nitrogen oxide emissions at mid‐latitudes appear to be leading to increases in ozone during winter, but associated increases in Arctic tropospheric ozone tend to be overestimated in the models. Increases are also found at the surface in northern Alaska during spring but not reproduced by the models. The causes are unknown but could be related to changes in local sources or sinks of Arctic ozone or in large‐scale weather patterns. Declining mid‐latitude emissions, or increased dry deposition to northern forests, may explain negative surface ozone trends over northern Scandinavia in spring that are not always captured by the models. Further work is needed to understand changes in Arctic tropospheric ozone. Key Points Coherent ozone trend analysis methodology applied to multi‐decade, pan‐Arctic surface and ozonesonde datasets and multi‐model medians Increasing winter Arctic tropospheric ozone overestimated by models in the free troposphere, and spring surface changes not captured Spring (summer) decreases (increases) in observed ozone throughout the troposphere, not always simulated by models

Quantifying forcings from anthropogenic perturbations to the Earth system (ES) is important for understanding changes in climate since the pre-industrial (PI) period. Here, we quantify and analyse a wide range of present-day (PD) anthropogenic effective radiative forcings (ERFs) with the UK's Earth System Model (ESM), UKESM1, following the protocols defined by the Radiative Forcing Model Intercomparison Project (RFMIP) and the Aerosol and Chemistry Model Intercomparison Project (AerChemMIP). In particular, quantifying ERFs that include rapid adjustments within a full ESM enables the role of various chemistry–aerosol–cloud interactions to be investigated. Global mean ERFs for the PD (year 2014) relative to the PI (year 1850) period for carbon dioxide (CO2), nitrous oxide (N2O), ozone-depleting substances (ODSs), and methane (CH4) are 1.89 ± 0.04, 0.25 ± 0.04, −0.18 ± 0.04, and 0.97 ±  0.04 W m−2, respectively. The total greenhouse gas (GHG) ERF is 2.92 ± 0.04 W m−2. UKESM1 has an aerosol ERF of −1.09 ± 0.04 W m−2. A relatively strong negative forcing from aerosol–cloud interactions (ACI) and a small negative instantaneous forcing from aerosol–radiation interactions (ARI) from sulfate and organic carbon (OC) are partially offset by a substantial forcing from black carbon (BC) absorption. Internal mixing and chemical interactions imply that neither the forcing from ARI nor ACI is linear, making the aerosol ERF less than the sum of the individual speciated aerosol ERFs. Ozone (O3) precursor gases consisting of volatile organic compounds (VOCs), carbon monoxide (CO), and nitrogen oxides (NOx), but excluding CH4, exert a positive radiative forcing due to increases in O3. However, they also lead to oxidant changes, which in turn cause an indirect aerosol ERF. The net effect is that the ERF from PD–PI changes in NOx emissions is negligible at 0.03 ± 0.04 W m−2, while the ERF from changes in VOC and CO emissions is 0.33 ± 0.04 W m−2. Together, aerosol and O3 precursors (called near-term climate forcers (NTCFs) in the context of AerChemMIP) exert an ERF of −1.03 ± 0.04 W m−2, mainly due to changes in the cloud radiative effect (CRE). There is also a negative ERF from land use change (−0.17 ± 0.04 W m−2). When adjusted from year 1850 to 1700, it is more negative than the range of previous estimates, and is most likely due to too strong an albedo response. In combination, the net anthropogenic ERF (1.76 ± 0.04 W m−2) is consistent with other estimates. By including interactions between GHGs, stratospheric and tropospheric O3, aerosols, and clouds, this work demonstrates the importance of ES interactions when quantifying ERFs. It also suggests that rapid adjustments need to include chemical as well as physical adjustments to fully account for complex ES interactions.

Despite local emission reductions, severe haze events remain a serious issue in Beijing. Previous studies have suggested that both greenhouse gas increases and aerosol decreases are likely to increase the frequency of weather patterns conducive to haze events. However, the combined effect of atmospheric circulation changes and aerosol and precursor emission changes on Beijing haze remains unclear. We use the Shared Socioeconomic Pathways (SSPs) to explore the effects of aerosol and greenhouse gas emission changes on both haze weather and Beijing haze itself. We confirm that the occurrence of haze weather patterns is likely to increase in future under all SSPs and show that even though aerosol reductions play a small role, greenhouse gas increases are the main driver, especially during the second half of the 21st century. However, the severity of the haze events decreases on decadal timescales by as much as 70 % by 2100. The main influence on the haze itself is the reductions in local aerosol emissions, which outweigh the effects of changes in atmospheric circulation patterns. This demonstrates that aerosol reductions are beneficial, despite their influence on the circulation.

Air pollution exposure is a leading public health problem in China. Despite recent air quality improvements, fine particulate matter (PM 2.5 ) exposure remains large, the associated disease burden is substantial, and population ageing is projected to increase the susceptibility to disease. Here, we used emulators of a regional chemical transport model to quantify the impacts of future emission scenarios on air pollution exposure in China. We estimated how key emission sectors contribute to these future health impacts from air pollution exposure. We found that PM 2.5 exposure declines in all scenarios across China over 2020–2050, with reductions of 15% under current air quality legislation, 36% when exploiting the full potential of air pollutant emission reduction technologies, and 39% when that technical mitigation potential is combined with emission controls for climate mitigation. However, population ageing means that the PM 2.5 disease burden under current legislation (CLE) increases by 17% in 2050 relative to 2020. In comparison to CLE in 2050, the application of the best air pollution technologies provides substantial health benefits, reducing the PM 2.5 disease burden by 16%, avoiding 536 600 (95% uncertainty interval, 95UI: 497 800–573 300) premature deaths per year. These public health benefits are mainly due to reductions in industrial (43%) and residential (30%) emissions. Climate mitigation efforts combined with the best air pollution technologies leads to an additional 2% reduction in the PM 2.5 disease burden, avoiding 57 000 (95UI: 52 800–61 100) premature deaths per year. Up to 90% of the 2020–2050 reductions in PM 2.5 exposure are already achieved by 2030, assuming efficient implementation and enforcement of currently committed air quality policies in key sectors. Achieving reductions in PM 2.5 exposure and the associated disease burden after 2030 will require further tightening of emission limits for regulated sectors, addressing other sources including agriculture and waste management, and international coordinated action to mitigate air pollution across Asia.

The effect of observational constraint on the ranges of uncertain physical and chemical process parameters was explored in a global aerosol–climate model. The study uses 1 million variants of the Hadley Centre General Environment Model version 3 (HadGEM3) that sample 26 sources of uncertainty, together with over 9000 monthly aggregated grid-box measurements of aerosol optical depth, PM2.5, particle number concentrations, sulfate and organic mass concentrations. Despite many compensating effects in the model, the procedure constrains the probability distributions of parameters related to secondary organic aerosol, anthropogenic SO2 emissions, residential emissions, sea spray emissions, dry deposition rates of SO2 and aerosols, new particle formation, cloud droplet pH and the diameter of primary combustion particles. Observational constraint rules out nearly 98 % of the model variants. On constraint, the ±1σ (standard deviation) range of global annual mean direct radiative forcing (RFari) is reduced by 33 % to −0.14 to −0.26 W m−2, and the 95 % credible interval (CI) is reduced by 34 % to −0.1 to −0.32 W m−2. For the global annual mean aerosol–cloud radiative forcing, RFaci, the ±1σ range is reduced by 7 % to −1.66 to −2.48 W m−2, and the 95 % CI by 6 % to −1.28 to −2.88 W m−2. The tightness of the constraint is limited by parameter cancellation effects (model equifinality) as well as the large and poorly defined “representativeness error” associated with comparing point measurements with a global model. The constraint could also be narrowed if model structural errors that prevent simultaneous agreement with different measurement types in multiple locations and seasons could be improved. For example, constraints using either sulfate or PM2.5 measurements individually result in RFari±1σ ranges that only just overlap, which shows that emergent constraints based on one measurement type may be overconfident.