Introduction

Causing 5.6–6.6 million premature deaths in 20161, air pollution is an important externality that interacts with climate change. Researchers2,3,4,5 have called for a coordinated effort to combat climate change and to improve air quality, but little work has been done to analyse the global synergies between the Paris Agreement on climate change6 and air pollution. This paper combines extensive data sets and models on emissions, climate, the energy system, the dispersion and impacts of ambient air pollutants, and the economy to quantify the impact of actual climate change mitigation policies proposed in the run-up to the 21st Conference Of the Parties in Paris on three inter-related Sustainable Development Goals7: Good health (SDG3), Clean energy (SDG7), and Climate action (SDG13). As emphasised by the Intergovernmental Panel on Climate Change (IPCC)8, a comprehensive analysis of co-benefits and adverse side effects is essential to estimate the actual costs of mitigation policies. In a political context, the co-benefits on air pollution are particularly relevant because they are mainly local and short term, while the averted climate change impacts occur globally over a decadal temporal scale.

While earlier regional estimates present a broad range of values for the co-benefits of climate policy on air quality (between 2 and 196$per tonne of carbon dioxide, with mean value of 49$/tCO2)9, recent work10,11,12 highlights that the improved human health outcomes due to cleaner air can largely offset the costs to reduce greenhouse gases (GHGs), particularly in heavily polluted regions. Here we assess the global and regional mortality, morbidity, and agricultural air quality co-benefits in the context of the Paris Agreement while accounting for future uncertainty in air pollution control measures. The main contribution lies in the quantification of ancillary benefits of actual (pledged) climate and energy policy elements in the Nationally Determined Contributions (NDCs), thereby capturing the heterogeneity in country-specific ambition levels and in sector coverage, in contrast with scenarios studying global and economy-wide carbon pricing, and complementing recent studies for the US13,14,15 and China16,17,18,19. Considering the regional differentiation in pledges is crucial because the transboundary effects of air pollution can be substantial20. The results show that the NDCs as pledged in the run-up to the Paris Agreement could improve health outcomes substantially, avoiding 71–99 thousand premature deaths in the year 2030. Present study furthermore compares the value of the air quality co-benefits with the macroeconomic cost of climate change mitigation policies, the latter also depending on the relative ambition levels across countries through industry competitiveness and international trade. We find that the value of co-benefits differs widely across regions and outweighs the costs of reducing GHGs on a global level in the majority of scenarios.

Results

Climate change mitigation pathways and the energy system

The climate scenarios encompass three trajectories of GHG emissions (global aggregate shown in Fig. 1; details by region in Supplementary Table 1). The Reference (REF) assumes no climate change mitigation policies beyond those already in place and compares best to Current policy scenarios in the literature21. The NDC scenario implements the GHG emission reductions and related policies in the Paris pledges, including the emission reductions that are conditional on other aspects of the Paris Agreement such as financing. Jointly, the current pledges represented by this scenario imply a likely increase in global average temperature of 2.5–3.2 °C, in line with the range provided by other studies21. The 2 °C scenario considers policies that result in a trajectory of GHG emissions that is consistent with at least 75% probability of limiting the average rise of global temperature to 2 °C by 2100 compared to pre-industrial levels.

Air quality policies are likely to develop in parallel of climate policies and may affect the potential scope of co-benefits of climate policy22. We address the implications of air pollution policy uncertainty for the co-benefits by exploring three storylines. The Fixed Legislation (FLE) scenario considers no additional implementation of air pollution abatement technologies from 2010 onwards. In combination with a REF climate policy, this assumption implies that economic and population growth lead to increasing global emissions over time for all air pollutants, with the exception of carbon monoxide (CO), for which a rising trend is offset by ongoing progress in energy technology and corresponding efficiency improvements. Because high levels of air pollution provide a broad base for reductions, the estimates of the co-benefits of climate policy derived under the assumption of Fixed air quality Legislation will be considered here as an upper bound. A gradual adoption and diffusion of air pollution control measures is included in the Stringent Legislation (SLE) scenario, which better reflects ambitious recent policy objectives in fast-growing countries, such as China. In the Best Available Technologies (BAT) scenario, countries fully adopt the maximum technically feasible air pollutant emission reduction technologies by 2030. This hypothetical benchmark identifies how structural changes induced by climate policy can improve air quality beyond what can be expected by end-of-pipe air pollution abatement technologies alone. By implementing stringent air pollution abatement, the BAT scenario leaves less room for co-benefits of climate action and will be used here to quantify a lower bound for the co-benefits.

Air pollutants affect global and local temperature changes and regional precipitation patterns23,24. The effect of air pollution control on climate change is a priori unclear, because the radiative forcing of some pollutants, such as black carbon (BC), is positive, while other pollutants (NOx, SO2) have a cooling effect on the climate. In addition to the median estimate of global average temperature change and the 50% probability bounds for the SLE scenario, Fig. 1 (dotted lines) includes the median estimate in case of less (FLE) and more (BAT) stringent air pollution control measures. The figure shows that the temperature change relative to pre-industrial levels in the REF climate policy under more ambitious air pollution controls (BAT) exceeds the central case estimate (SLE) by 0.08 °C in 2100, while the higher end of the air pollution projection (FLE) implies temperature changes that are 0.32 °C below the SLE case in 2100. Hence, the end-of-pipe reduction of air pollutants with a cooling effect outweighs the decrease in air pollutants that contribute to global warming in our scenarios, leading to a net upward effect on global mean temperatures, in line with other work25,26. This result does not consider the effect of air pollution controls on GHG emissions and is less pronounced in more ambitious climate mitigation scenarios. Compared to the REF, the decrease in GHG emissions in the 2 °C scenario implies larger reductions in global mean temperature in the case of cleaner air (1.84 °C under BAT vs. 1.62 °C under FLE, REF—2 °C in 2100), since co-reduction of cooling aerosols plays a smaller role when stringent air pollution controls (BAT) are in place.

Ambitious climate policies are to reshape the energy landscape in the coming decades27,28,29. The decarbonisation of the energy supply mix combined with reduced energy consumption through efficiency gains will be key factors in the transformation of energy systems (Fig. 1; regional numbers are given in Supplementary Tables 25). Energy efficiency drives total global energy consumption down by approximately 10% in the NDC scenario and by more than a quarter (27%) in the 2 °C scenario in the year 2050 compared to the REF. Moreover, the rising share of renewables represents a structural change in the energy sector, especially in electricity generation.

The impact of climate policy measures on air quality

Policies that aim to mitigate climate change tend to reduce emissions of GHGs and local air pollutants22,30, particularly when both share the same underlying drivers. The extent to which CO2 reductions are correlated with changes in air pollutants differs by region and type of air pollutant (Fig. 2). For pollutants that mainly result from the combustion of fossil fuels, such as SO2 and NOx, emission reductions tend to be strongly correlated with decreasing emissions of CO2. This co-movement is less obvious for regions with large GHG abatement potential from options other than the burning of fossil fuels, such as land use in Brazil (see Supplementary Table 6), and for regions with important industrial sources of pollutants, such as SO2 from production of metals in Russia. Figure 2 shows that air quality co-benefits generally outweigh adverse side effects on the aggregate level for most pollutants and regions, although there are some trade-offs embedded in technological choices such as carbon capture and storage31,32, biomass energy33, and biofuels34. For organic carbon and CO in particular, Fig. 2 shows that not all models in the IPCC’s Fifth Assessment Report35 agree on the sign of the change in emissions on the global level, indicating uncertainty in the estimates. Figure 2 furthermore indicates the sensitivity of co-reductions with respect to the implemented air pollution control technologies (Supplementary Tables 728 provide numbers by region and scenario). Low air pollutant emission intensities (BAT) in the benchmark in key climate change mitigation sectors reduce the scope for co-benefits, as can be seen from the sulphur dioxide emission reductions in India, for instance, where the implementation of BAT implies less polluting coal-fired electricity generation facilities.

The impacts of air pollution are not confined to national borders as air pollutants are dispersed geographically20,36. The changes in the concentration of particulate matter with diameter smaller than 2.5 μm (PM2.5) and tropospheric ozone mixing ratio mapped in Fig. 3 are the results of emissions, transportation, and atmospheric chemistry reactions of pollutants (see detailed numbers in Supplementary Tables 2934). Hence, results for PM include both direct emissions from primary sources, such as BC and organic matter, and secondary PM that derives from emissions of NH3, NOx, SO2, and volatile organic compounds (VOCs). Ozone is formed by the reaction of precursor gases NOx, VOCs, and CO in the presence of sunlight. By 2030, the time horizon of most NDCs, the Paris pledges lead to globally relatively small but locally significant reductions in the concentration of PM2.5, while a decrease in ozone mixing ratio spreads more widely across the globe. Over the long run (2050), the impact of a more ambitious climate policy setting (2 °C) reveals that the potential contribution of GHG abatement policies to improved air quality is substantial, particularly in China, India, and the Middle East, where benchmark concentrations are comparably high.

Air quality co-benefits for human health and agriculture

The benefits of improved air quality include avoided premature mortality due to related cardiovascular and respiratory diseases and lung cancer37. Consistent with other research38, outdoor air pollution-related premature mortality is projected to roughly double by 2050 compared to 2010 when considering the current climate (REF) and air pollution (FLE) policies combined with population and economic growth, whereas ambitious policies (2 °C—BAT) bring the number of premature deaths below the level of 2010 despite population growth by 2050 (premature mortality per scenario, region, and pollutant are detailed in Supplementary Tables 3540). Climate policies as currently pledged under the NDCs lead to between 71 and 99 thousand avoided premature deaths globally in the year 2030 compared to current climate policies (REF), while bringing greenhouse emissions in line with a 2 °C temperature goal prevents between 178 and 346 thousand premature deaths globally in the year 2030. In the year 2050, 2 °C-compatible climate action reduces the premature deaths from air pollution by 0.7–1.5 million compared to the REF, of which more than two thirds are prevented in India and China (Fig. 4). The gap between the NDC and 2 °C scenarios is explained by countries like India, for which the current NDC up to 2030 fails to capture the potential air quality co-benefits of climate action. In addition to avoided premature mortality, reductions in air pollution bring benefits in terms of reduced sickness days, thereby boosting labour markets (see Supplementary Table 41).

Disclaimer

The views expressed are purely those of the authors and may not in any circumstances be regarded as stating an official position of the European Commission.