Unveiling Future Individual and Compound Heat and Air Pollution Extremes in China: Insights for Mitigation

Abstract Extreme heat and air pollution increasingly threaten health and ecosystems in China, yet future risks are poorly resolved by the coarse monthly data of Coupled Model Intercomparison Project phase 6 (CMIP6). Here, we develop a machine learning–based framework to downscale monthly CMIP6 outputs to daily, enabling detailed assessment of future individual and compound heat-air pollution extremes. Cross validation shows strong performance ( R 2 = 0.88 for O 3 and 0.80 for PM 2.5 downscaling models), with meteorological uncertainties and monthly PM 2.5 biases as main error sources. Under emission-control scenarios [shared socioeconomic pathway (SSP) 1-2.6, SSP2-4.5, SSP5-8.5], PM 2.5 events decline rapidly, while O 3 and heat extremes persist or worsen, resulting in O 3 becoming the dominant pollutant by midcentury. Weak air quality controls (SSP3-7.0) cause severe increases in both pollutants, resulting in over 100 polluted days per year across eastern China in the future. Compound extremes intensify across all scenarios due to sustained warming. While near-term climate forcer mitigation reduces PM 2.5 and O 3 levels, aerosol-driven regional warming exacerbates heat-ozone extremes, highlighting the need for simultaneous greenhouse gas reductions to avoid worsening compound climate-air quality risks. The downscaled daily PM 2.5 and O 3 datasets provide a key resource for advancing air quality, climate, and health research and inform integrated mitigation strategies in China. Significance Statement Heat waves and air pollution pose major health risks in China, but Coupled Model Intercomparison Project phase 6 (CMIP6) models lack the daily detail needed for future projections. This study overcomes this limitation by using machine learning to downscale monthly data to daily pollution estimates. Results reveal that even with strong emission controls, harmful compound heat-ozone extremes persist, while weak controls lead to dramatic increases in pollution events. Effective risk management requires integrated strategies addressing both air pollution and climate warming, guiding air quality and climate policies.