Identifying surface-level O3 photochemical regimes using meteorology-adjusted satellite observations and statistical and machine learning models in South Korea
Abstract. Ground-level ozone (O3) is a persistent urban and regional air quality challenge because its production responds nonlinearly to nitrogen oxides (NOx) and volatile organic compounds (VOCs) and is strongly influenced by meteorological conditions. The satellite-based ratio of formaldehyde (HCHO) to nitrogen dioxide (NO2) (i.e., HCHO/NO2) is a reliable indicator of O3 photochemical regimes that distinguishes whether O3 production is NOx-limited or NOx-saturated; however, the corresponding regime thresholds remain poorly constrained. A new framework was developed to identify refined regime thresholds by integrating ground and satellite observations, while accounting for meteorological variability using statistical and machine learning models. When applied to O3 air pollution in South Korea from 2019 to 2023, the framework revealed robust O3 photochemical regimes based on province-specific regime thresholds with 34 % lower uncertainty than conventional approaches. In addition, the regime thresholds reasonably captured regional characteristics of O3 responses to NO2 perturbations, with the scenario-specific correlations of 0.57–0.75. A meta-analysis indicated that a unified nationwide regime threshold was not feasible and that applying the regional regime thresholds of 1.8–3.9 provided a more reasonable basis for O3 photochemical regime inferences. The framework provides transferable, meteorology-adjusted guidance for designing region-specific NOx and VOC control strategies, supporting more effective O3 air quality management.