The Empirical Kinetics Modeling Approach (EKMA) is a valuable tool for guiding ozone (O3) precursor emission controls, but O3 isopleths generated through regional emission perturbations in regional air quality models (RAQMs) may hide interpretive biases for local applications. Using the WRF-CMAQ model, this study constructs city-specific EKMA diagrams by perturbing O3 precursor emissions across the Greater Bay Area (GBA) of China and integrates source apportionment to quantify these biases. Results reveal that such isopleths overestimate city-specific control efficacy by approximately 58.7% on average across the GBA cities, based on monthly O3 metrics, with biases persisting as high as 47.2% even in relatively stagnant meteorological conditions. This indicates that local isopleths derived from regional emission perturbations inadequately represent the effectiveness of local emission controls. Reinterpreting the city-specific isopleths from a regional emission reduction perspective and applying multiple air quality indicators, we observe O3 concentration rebounds in numerous GBA cities due to inappropriate NOx reductions. Nonetheless, when shifting to a risk-based perspective, short-term NOx reductions achieve overall health benefits by offsetting greater NO2 risks, which outweigh the elevated O3-related risks in VOC-limited regimes. These insights provide evidence supporting governmental efforts to prioritize short-term, cost-effective NOx-focused controls, which deliver immediate health benefits, particularly in high-NOx urban cores. Additionally, this study offers positive incentives for adopting long-term attainment strategies with balanced NOx:AVOCs (anthropogenic VOCs) ratios of 1:1 or even 2:1 to achieve air pollution abatement and decarbonization in the GBA simultaneously. Our work highlights the imperative of aligning EKMA-RAQMs diagram interpretations with experimental designs and provides a novel perspective for holistically evaluating regional emission control policies.
Liang et al. (Mon,) studied this question.