Machine learning elucidates ubiquity of enhanced ozone air pollution in China linked to the spring festival effect

Baizhen Zhu,Jie Fang,Yunjiang Zhang,Jian Qiu,Kehong Chen,Kexin Zhang,Hongwei Liang,Han Yang,Yihua Ding,Xinlei Ge
DOI: https://doi.org/10.1016/j.apr.2024.102127
IF: 4.831
2024-03-31
Atmospheric Pollution Research
Abstract:Emissions' impact on ozone pollution is a focal point in atmospheric chemistry. The Chinese Spring Festival (CSF) effect offers a unique opportunity to study short-term emission reductions realistically. We conducted a comprehensive analysis using data from the national air quality observation platform and TROPOMI satellite observations of tropospheric formaldehyde (HCHO) and nitrogen dioxide (NO 2 ) column concentrations before and during the CSF from 2015 to 2022. Satellite observations during the Spring Festival revealed a 13% reduction in HCHO and a 9% reduction in NO 2 compared to the pre-CSF period. To comprehend the CSF effect on ozone precursor emissions, a random forest model accounting for meteorological effects was employed. CSF-induced emission reductions led to NO 2 concentration decreases of −24.1%, −17.5%, −9.9%, and −12.6% in the Yangtze River Delta (YRD), Pearl River Delta (PRD), Beijing-Tianjin-Hebei (BTH), and Sichuan Basin (SCB), respectively. Conversely, BTH and YRD regions experienced substantial emission-driven ozone concentration increases of 6.7% and 4.9%. Assessing regional ozone generation sensitivity during the CSF, the BTH, YRD, and SCB fell within the VOC-limited regime. Our study underscores the CSF effect's significant impact on ozone pollution. Given the non-linear relationship between ozone and its precursors, it emphasizes the necessity of adopting synergistic abatement strategies to mitigate the escalating ozone pollution.
environmental sciences
What problem does this paper attempt to address?