| Title: | Predicting ozone formation in petrochemical industrialized Lanzhou city by interpretable ensemble machine learning |
| Author(s): | Wang L; Zhao Y; Shi J; Ma J; Liu X; Han D; Gao H; Huang T; |
| Address: | "Collaborative Innovation Center for Western Ecological Safety, Lanzhou University, Lanzhou, 730000, China. Key Laboratory for Environmental Pollution Prediction and Control, Gansu Province, College of Earth and Environmental Sciences, Lanzhou University, Lanzhou, 730000, China. Electronic address: zhaoyuan@lzu.edu.cn. Laboratory for Earth Surface Processes, College of Urban and Environmental Sciences, Peking University, Beijing, 100871, China. Key Laboratory for Semi-Arid Climate Change of the Ministry of Education, College of Atmospheric Sciences, Lanzhou University, Lanzhou, 730000, China. Key Laboratory for Environmental Pollution Prediction and Control, Gansu Province, College of Earth and Environmental Sciences, Lanzhou University, Lanzhou, 730000, China" |
| DOI: | 10.1016/j.envpol.2022.120798 |
| ISSN/ISBN: | 1873-6424 (Electronic) 0269-7491 (Linking) |
| Abstract: | "Ground-level ozone (O(3)) formation depends on meteorology, precursor emissions, and atmospheric chemistry. Understanding the key drivers behind the O(3) formation and developing an accurate and efficient method for timely assessing the O(3)-VOCs-NOx relationships applicable in different O(3) pollution events are essential. Here, we developed a novel machine learning ensemble model coupled with a Shapley additive explanation algorithm to predict the O(3) formation regime and derive O(3) formation sensitivity curves. The algorithm was tested for O(3) events during the COVID-19 lockdown, a sandstorm event, and a heavy O(3) pollution episode (maximum hourly O(3) concentration >200 mug/m(3)) from 2019 to 2021. We show that increasing O(3) concentrations during the COVID-19 lockdown and the heavy O(3) pollution event were mainly caused by the photochemistry subject to local air quality and meteorological conditions. Influenced by the sandstorm weather, low O(3) levels were mainly attributable to weak sunlight and low precursor levels. O(3) formation sensitivity curves demonstrate that O(3) formation in the study area was in a VOCs-sensitive regime. The VOCs-specific O(3) sensitivity curves can also help make hybrid and timely strategies for O(3) abatement. The results demonstrate that machine learning driven by observational data has the potential to be a very useful tool in predicting and interpreting O(3) formation" |
| Keywords: | Humans *Ozone/analysis *Air Pollutants/analysis Environmental Monitoring/methods *covid-19 Communicable Disease Control *Air Pollution/analysis Machine Learning China *Volatile Organic Compounds/analysis Ensemble machine learning Ground-level ozone (O(3)); |
| Notes: | "MedlineWang, Li Zhao, Yuan Shi, Jinsen Ma, Jianmin Liu, Xiaoyue Han, Dongliang Gao, Hong Huang, Tao eng England 2022/12/05 Environ Pollut. 2023 Feb 1; 318:120798. doi: 10.1016/j.envpol.2022.120798. Epub 2022 Dec 1" |
