Satellite-constrained AI uncovers nonlinear drivers of ozone trends in China
Ground-level ozone pollution in China has continued to worsen despite substantial reductions in anthropogenic emissions, raising concerns about sustainability of air quality improvements under a warming climate. Here, we combine satellite observations of ozone precursors with an interpretable artificial intelligence framework to diagnose nonlinear ozone responses and disentangle the relative influences of precursor changes and meteorological variability across China during 2014–2023. Warm-season ozone increased despite widespread declines in NO 2 columns, revealing a growing mismatch between emission reductions and ozone responses. Meteorological changes amplified ozone increases, with surface warming emerging as the dominant driver. Idealized warming experiments suggest that a 3 °C temperature increase could raise national mean ozone by 6 μg m -3 and cause more than 50,000 additional ozone-attributable deaths. However, coordinated precursor management can offset a substantial fraction of this climate-induced burden. Our findings highlight the need to integrate climate adaptation with emission-control strategies for sustainable ozone mitigation.
Authors
- Jia Xing (ORCID: https://orcid.org/0000-0001-7574-2381)
- Senlin Hu (ORCID: https://orcid.org/0000-0002-9065-3667)
- Ge Song (ORCID: https://orcid.org/0000-0001-5099-1306)
- Maolin Zhang (ORCID: https://orcid.org/0009-0007-7045-8614)
- Siwei Li
Institutions
- Wuhan University (CN)
- State Key Laboratory of Information Engineering in Surveying Mapping and Remote Sensing (CN)
- University of Tennessee at Knoxville (US)
Publication Details
- Journal
- npj Clean Air
- Published
- 2026-11-19
- DOI
- https://doi.org/10.1038/s44407-026-00114-0
- Primary Topic
- Atmospheric chemistry and aerosols
- Type
- article
- Field-Weighted Citation Impact
- 0.00