High concentrations of ozone (O3) and fine particulate matter (PM2.5) negatively impact human health, particularly in megacities. These pollutants often co-occur in summer due to stagnant meteorological conditions, high solar radiation, and temperature, which promote the formation of nitrogen oxides (NOx) and volatile organic compounds (VOCs). Previous studies have shown a positive relationship between MDA8 O3 (maximum daily 8-h average) and DA24 PM2.5 (daily 24-h average) concentrations, with a maximum turning point (MTP) observed at around 50–60 µg m−3 DA24 PM2.5 in Chinese megacity-clusters. Above this MTP, O3 levels stabilize despite increasing PM2.5, attributed to PM2.5 scavenging of radicals that inhibit O3 production. NYC and Beijing, two extensively studied megacities, have implemented strict emission control policies, leading to significant PM2.5 decreases. However, O3 reductions have been slower in NYC and even increased in Beijing, necessitating further investigation into the O3-PM2.5 relationship and its response to emission controls.

The introduction section already provides a summary of the relevant literature, highlighting the previously established positive correlation between O3 and PM2.5, the existence of a maximum turning point (MTP) in their relationship, and the contrasting trends observed in NYC and Beijing following emission control measures. The authors cite studies demonstrating the impact of PM2.5 on O3 formation through the scavenging of HO2 and NO3 radicals, and the effects of emission control policies on both pollutants in these two megacities. The discrepancy between PM2.5 and O3 trends following emission controls motivates the current research.

The study analyzed 19 years of surface measurements in NYC (2001-2019) and 6 years in Beijing (2014-2019), along with aerosol chemical composition data. The relationship between MDA8 O3 and DA24 PM2.5 was analyzed for different sub-periods, defined by PM2.5 concentration levels. A non-linear function, combining a positive linear component reflecting O3/PM2.5 co-occurrence and a negative power function representing O3 formation suppression by PM2.5, was used to fit the data. Aerosol mass fractions were divided into SAP (sulfate, ammonium, primary organic aerosol) and non-SAP components to investigate the influence of aerosol composition on the O3-PM2.5 relationship. The Community Multiscale Air Quality (CMAQ) model was used to simulate O3 and PM2.5 formation under different emission reduction scenarios, including regional equal percentage reductions. The methodology also involved analyzing aerosol chemical composition using the Aerodyne Aerosol Mass Spectrometer (AMS), and estimating synchronous emission reductions based on the relationship between top 5% DA24 PM2.5 concentration and emission reduction ratios from model simulations.

In NYC, the O3-PM2.5 linear slope increased over time, indicating a weaker O3 control effect compared to PM2.5. This increase correlated with a reduced mass fraction of SAP in PM2.5. The power function coefficient also increased, signifying an enhanced O3 suppression effect by PM2.5 at the same PM2.5 level. In Beijing, despite significant PM2.5 improvements since 2013, O3 levels increased. Similar to NYC, the O3-PM2.5 linear slope increased, and the power function coefficient also rose. The changes in the O3-PM2.5 relationship in both cities were linked to changes in aerosol composition, primarily a reduced fraction of SAP components in PM2.5. CMAQ model simulations suggested that regional equal percentage emission reductions in Beijing and other Chinese megacities could mitigate further O3 increases and avoid an increase in the O3-PM2.5 linear slope. To reach specific PM2.5 and O3 reduction goals in Beijing, BTH (Beijing-Tianjin-Hebei) regional emission reductions of 42%, 53%, and 70% were estimated, respectively. Similar analyses for the Yangtze River Delta (YRD) and Pearl River Delta (PRD) showed varying responses to emission controls, with YRD exhibiting an increased O3-PM2.5 linear slope and PRD showing only a slight increase.

The findings highlight the complex non-linear relationship between O3 and PM2.5 and how this relationship changes in response to emission control policies. The increased O3-PM2.5 linear slope observed in both NYC and Beijing following emission controls, primarily targeting SO2 and PM2.5, indicates that O3 reduction lags behind PM2.5 reduction. This is attributed to the altered chemical composition of PM2.5, specifically a decrease in the SAP fraction. The results emphasize the need for a more comprehensive approach to air pollution control that considers the interdependencies between different pollutants. Regional equal percentage emission reductions, as suggested by model simulations, offer a potential strategy to simultaneously reduce O3 and PM2.5. However, the study acknowledges the limitations of the simplified non-linear function used and suggests further research incorporating more accurate representations of the O3-PM2.5 relationship.

The study demonstrates the complex interplay between O3 and PM2.5 in response to emission control policies, highlighting the need for a holistic approach to pollution control. The increased O3-PM2.5 linear slope under current policies underscores the importance of considering the non-linear interactions between pollutants. Regional equal percentage emission reductions are proposed as a potential strategy, though further research is needed to refine this approach and explore the optimal balance of NOx and VOC emission reductions. The differences in response between NYC, Beijing, YRD, and PRD highlight the need for region-specific strategies.

The study relies on a simplified non-linear function to describe the O3-PM2.5 relationship. The accuracy of the power function coefficient, particularly in sub-periods with limited data points, may be affected. The assumption of similar aerosol composition between PM1 and PM2.5 introduces uncertainty, and the neglect of primary black carbon might also affect the results. The CMAQ model simulations have inherent uncertainties related to emission inventories, model parameters, and meteorological inputs. The conclusions regarding the effectiveness of regional equal percentage emission reductions should be considered within the context of these limitations.