The story behind this project starts with an international cooperation between two universities that are both famous for their major in meteorology. In 2020, I was a doctoral candidate at the Nanjing University of Information Science and Technology (NUIST) while studying at the University of Reading (UoR) as a visiting PhD student. The study was funded and supervised by Prof. Xinlei Ge at NUIST and Prof. Michaela Hegglin at UoR. Due to the project's start coinciding with the beginning of the COVID-19 pandemic, we selected the COVID-19 period in 2020 as a unique opportunity to explore the drivers of air pollutants and related health risk changes.

The pandemic of COVID-19 has now been lasting for almost three years. Over this period, there is increasing knowledge for the public to understand the virus itself, while research on its impact has also become more comprehensive and in-depth. In our current study, we contribute to this discussion by providing a more detailed evaluation of different drivers of air quality and related health risks during this period. The COVID-19 restrictions have worldwide expressed themselves in drastic reductions in industrial and transport activities, have led to a distinct reduction of primary pollutant emissions as shown and extensively quantified in many research studies. However, another factor to be known when assessing the impact of emissions is the prevailing meteorological conditions This aspect of air pollution was neglected in most of these research studies. To complicate the story further, the Chinese New Year (CNY) and the Clean Air Plan (CAP) in China also contributed to the observed reductions. To understand the relative roles of these confounding factors on air quality and related health risks, their respective influence had to be isolated.

In this study, the stational observation dataset and the Copernicus Atmosphere Monitoring Service Reanalysis (CAMSRA) dataset with a fixed emissions inventory (which represents a counterfactual to the world having experienced observed emissions reductions) were used to explore the different anthropogenic drivers. The Gradient Boosting Machine (GBM) model, driven by surface meteorological variables, was executed to investigate the meteorological drivers. Finally, a HAQI model was applied to calculate the relative health risks.

Our study reveals that the CNY reduces NO₂ concentrations on average by 26.7% each year, while the COVID lockdown measures have led to an additional 11.6% reduction in 2020, and the CAP to a reduction in NO₂ by 15.7%. On the other hand, meteorological conditions led to increases in NO₂ of 7.8%. Neglecting the CAP and meteorological drivers thus leads to an overestimate and underestimate of the effect of the COVID lockdown on NO₂ reductions, respectively. For O₃ the opposite behavior is found, with changes of +23.3%, +21.0%, +4.7%, and -0.9% for CNY, COVID lockdown, CAP, and meteorology effects, respectively. The effect of ozone increase was more than compensated for by the decreases in PM and NO2 in the HAQI calculation. As a result, the total effects of these drivers show a drastic reduction in multi-pollutant related health risks across China, with meteorology affecting particularly the Northeast of China adversely. Importantly, the CAP’s contribution highlights the effectiveness of the Chinese government’s air quality regulations on NO₂ reduction.

I am grateful to my co-authors: Michaela I. Hegglin, Yuanfei Luo, Yue Yuan, Bing Wang, Johannes Flemming, Junfeng Wang, Yunjiang Zhang, Mindong Chen, Qiang Yang and Xinlei Ge, for their detailed guidance and ample support. This study is funded by the National Natural Science Foundation of China (Grant Nos. 42021004 and 21976093).

Reference:

1. Liu F, et al. Abrupt decline in tropospheric nitrogen dioxide over China after the outbreak of COVID-19. Science Advances 6, eabc2992 (2020).
2. Miyazaki K, et al. Global tropospheric ozone responses to reduced NOx emissions linked to the COVID-19 worldwide lockdowns. Science Advances 7, eabf7460 (2021).
3. Rybarczyk Y, Zalakeviciute R. Assessing the COVID-19 Impact on Air Quality: A Machine Learning Approach. Geophys Res Lett 48, e2020GL091202 (2021).