The COVID-19 pandemic promoted strict restrictions to human activities in China, which led to an unexpected increase in ozone (O3) regarding to nitrogen oxides (NOx) and volatile organic compounds (VOCs) co-abatement in urban China. However, providing a quantitative assessment of the photochemistry that leads to O3 increase is still challenging. Here, we evaluated changes in O3 arising from photochemical production with precursors (NOX and VOCS) in industrial regions in Shanghai during the COVID-19 lockdowns by using machine learning models and box models. The changes of air pollutants (O3, NOX, VOCs) during the COVID-19 lockdowns were analyzed by deweathering and detrending machine learning models with regard to meteorological and emission effects. After accounting for effects of meteorological variability, we find increase in O3 concentration (49.5%). Except for meteorological effects, model results of detrending the business-as-usual changes indicate much smaller reduction (-0.6%), highlighting the O3 increase attributable to complex photochemistry mechanism and the upward trends of O3 due to clear air policy in Shanghai. We then used box models to assess the photochemistry mechanism and identify key factors that control O3 production during lockdowns. It was found that empirical evidence for a link between efficient radical propagation and the optimized O3 production efficiency of NOX under the VOC-limited conditions. Simulations with box models also indicate that priority should be given to controlling industrial emissions and vehicle exhaust while the VOCs and NOX should be managed at a proper ratio in order to control O3 in winter. While lockdown is not a condition that could ever be continued indefinitely, findings of this study offer theoretical support for formulating refined O3 management in industrial regions in Shanghai, especially in winter.
Keywords: COVID-19 confinement; Machine learning; Ozone; Photochemical box model; Volatile organic compounds.
© 2023 Elsevier Ltd. All rights reserved.