The synergistic control of PM2.5 and ozone (O3) are the focus of air quality improvement during the 14th Five-Year Plan in China. The production of O3 shows a highly nonlinear relationship with its precursors volatile organic compounds (VOCs) and nitrogen oxides (NOx). In this study, we conducted online observations of O3, VOCs, and NOx at an urban site in downtown Nanjing from April to September of 2020 and 2021. The average concentrations of O3 and its precursors between these two years were compared, and then the O3-VOCs-NOx sensitivity and the VOCs sources were analyzed using the observation-based box model (OBM) and positive matrix factorization (PMF), respectively. The results showed that the mean daily maximum O3 concentrations, VOCs, and NOx concentrations decreased by 7% (P=0.031), 17.6% (P<0.001), and 14.0% (P=0.004) from April to September of 2021 compared with those from the same period in 2020, respectively. The average relative incremental reactivity (RIR) values of NOx and anthropogenic VOCs during the O3 non-attainment days in 2020 and 2021 were 0.17 and 0.14 and 0.21 and 0.14, respectively. The positive RIR values of NOx and VOCs indicated that O3 production was controlled by both VOCs and NOx. The O3 production potential contours (EKMA curves) based on the 50×50 scenario simulations also supported this conclusion. The PMF results showed that industrial and traffic-related emissions were the main sources of VOCs. The five PMF-resolved factors were identified as industrial emissions, including industrial liquefied petroleum gas (LPG) use, the benzene-related industry, petrochemistry, toluene-related industry, and solvent and paint use, which contributed 55%-57% of the average mass concentration of total VOCs. The summed relative contributions of vehicular exhaust and gasoline evaporation were 43%-45%. Petrochemistry and solvent and paint use showed the two highest RIR values, suggesting that VOCs from these two sources should be reduced with priority to control O3. With the implementation of VOCs and NOx control measures, the O3-VOCs-NOx sensitivity and VOCs sources have changed, and therefore we still need to follow their variations in the future to timely adjust O3 control strategies during the 14th Five-Year Plan.
Keywords: Nanjing; O3-VOCs-NOx sensitivity; VOCs source apportionment; observation-based model (OBM); positive matrix factorization (PMF) model.