Quantifying Temperature-Driven Surface Ozone Formation: Insights From Field Observations

Abstract

Tropospheric ozone (O₃) pollution has emerged as an increasingly prominent environmental challenge, especially in the context of global warming. This study quantitatively assessed key factors influencing O₃ formation using field observations in cities spanning from northern to southern China and eastern China (Beijing, Haikou, and Suzhou) combined with an interpretable XGBoost-SHAP model. Model performance was validated across space and time. Temperature explained 33.6% of the total model-attributed importance for O₃ in Beijing, with effects varying by pollution scenario: highest under co-pollution (83.3 ± 39.9 μg m⁻³, 56%), followed by O₃ pollution (51.9 ± 30.6 μg m⁻³, 49%), and PM_2.5 pollution (21.3 ± 11.3 μg m⁻³, 25%). A 1°C rise increased O₃ by 6.1 ± 3.0 μg m⁻³ in Beijing, with lower sensitivities observed in Suzhou and Haikou, consistent with their distinct precursor availability and chemical regimes. During heatwaves (≥35°C), O₃ concentrations were 2.3 times higher than at normal temperatures (≤30°C). These results provide scenario-specific, quantitative insights into temperature-driven O₃ formation across contrasting pollution regimes and climatic zones, with implications for climate-informed air quality management strategies.