Interpretation of decadal-scale ozone production efficiency in the Seoul Metropolitan Area: Implication for ozone abatement

Understanding the drivers of urban ozone (O-3) formation is challenging and requires uncovering the non-linear relationship between O-3 and its precursors. We used a novel method to differentiate between background O-3 levels and local photochemical O-3 generation using the observation-based O-3 production efficiency (OPE) from decade-long observations including nitrogen dioxide (NO2), nitrogen monoxide (NO), reactive nitrogen compounds (NOy), O-3, and 56 volatile organic compounds (VOCs) covering eight years (2009-2016). In this study, we employed O-x (=O-3+NO2), and defined OPEx as the number of O-x molecules produced from the oxidation of a single NOx molecule, and estimates the local contribution (in %) to total O-3 from the slope of the O-x-NO2(=NOy-NOx) linear regression. We thus identified the local urban O-3 formation regime using the relationship of several factors, including nitrogen oxide (NOx), VOC/NOx, NO2, NO2, and long-range transported O-x, and suggested how O-3 can be reduced in the Seoul Metropolitan Area (SMA). The results show that the regional background dominates the annual level of O-x with a contribution of about 66 +/- 11%, while photochemical production acts as a stronger factor influencing the annual variability of O-x than regional background. Regional background varies by transport evidenced by its good correlation with the west wind fraction, whereas photochemical production depends on the VOC/NOx ratio, which is high in summer and low in winter. The relationship between photochemical production and VOC/NOx indicates that the transition of the O-x formation regime occurs at a VOC/NOx ratio of approximately 4 in the SMA. Urban environments with NOx > 20 ppb typically lie in low VOC/NOx zone, where reduced NO2 emissions lead to decreased NO2 but increased OPEx, with no discernible change in photochemical O-x production. Thus, NOx emission reductions in urban environments would have a negligible effect on Ox reduction. In contrast, suburban environments such as usually NOx < 10 ppb are located in the higher VOC/NOx zone, where decreased NO2 leads to decreased OPEx with no discernible change in NO2, consequently suppressing photochemical O-x production. Together, these results suggest that any urban O-3 policy in the SMA should consider significant cuts in NOx emissions to allow the O-3 regime to shift. The OPEx method can also be applied to other urban areas for the development of effective O-3 reduction policies.