Analysis of the photochemical characteristics and sensitivity of ozone pollution in Xi'an
Xiyan Guan,Keding Lu,Ningning Zhang,Xin Li,Xuefei Ma,Xinping Yang,Huabin Dong,Shiyi Chen,Junji Cao,Limin Zeng,Yuanhang Zhang
DOI: https://doi.org/10.1360/TB-2021-0436
2021-01-01
Chinese Science Bulletin
Abstract:The Fenwei Plain is an important agricultural production base and transportation hub in China with a residential population of over 55 million. With rapid industrialization and urbanization, this area has experienced serious photochemical pollution in the summer season. In this study, a field campaign was conducted in summer 2019 with measurements of ozone and photochemical precursors in Xi'an, a megacity in the Fenwei Plain. An explicit case was performed to explore O-3-NOx-VOCs sensitivity and the local ozone budget using an observation-based model. Results revealed that precursors significantly promoted ozone production. The positive matrix factorization (PMF) model combined with the relative incremental reactivity (RIR) method was applied to identify the key VOC sources that affect ozone formation. An ozone pollution episode lasting 11 d was selected and divided into three periods: Initial (July 8-11), polluted (July 12-15), and sweep (July 16-18) periods. It was found that strong solar radiation intensity, high temperature, low relative humidity, and low pressure prevailed during the polluted period, which favored ozone production. The maximum O-3 concentrations at the upwind CB site and urban DHS site were found to be 125 ppbv during the polluted period, whereas at the QL site, the maximum O-3 concentrations were much lower (98 ppbv). The maximum local ozone production rate (P(O-x)) of the three sites during the polluted period was significantly higher than that in the initial and sweep periods. The P(O-x) was 17.7, 33.9, and 14.9 ppbv/h for the CB, DHS, and QL sites, respectively. The in situ Q. budget demonstrated that ozone pollution at the CB and DHS sites was mainly attributed to local photochemical production, whereas the regional transportation effect could not be ignored at the QL site. The O-3-NOx-VOCs sensitivity analysis revealed that ozone production regimes were transitional, VOCs-limited. and NOx-limited at the CB, DHS, and QL sites, respectively, and the regimes stayed unchanged for all the three sites during the entire ozone pollution episode. However. during the initial period at the DHS site, NOx reduction led to a negative effect on ozone control. The ozone formation regime obtained from the emission rate box model was consistent with that of the RIR method. From the proportion of grouped VOCs, the alkane and OVOCs were the dominant species in the mixing ratio (78%-86%), whereas the contribution of aromatics to OFP (16%-24%) was second only to that of OVOCs. Finally, by combining the PMF model and RIR method, ozone production was found to be more sensitive to precursors reduction in the initial period than in the polluted period at the three sites, which indicated that pollutants control measures for ozone pollution mitigation will be more effective in the initial than in polluted periods. Therefore, during an ozone pollution episode, the time to take action for ozone pollution control is not during the polluted period when the ozone concentrations peak, but should be brought forward, when controlling the concentration of precursors has the highest efficiency in reducing ozone production.