Revealing Drivers of Haze Pollution by Explainable Machine Learning

被引:107
作者
Hou, Linlu [1 ,2 ]
Dai, Qili [1 ,2 ,6 ]
Song, Congbo [3 ]
Liu, Bowen [4 ]
Guo, Fangzhou [5 ]
Dai, Tianjiao [1 ,2 ]
Li, Linxuan [1 ,2 ]
Liu, Baoshuang [1 ,2 ,6 ]
Bi, Xiaohui [1 ,2 ,6 ]
Zhang, Yufen [1 ,2 ,6 ]
Feng, Yinchang [1 ,2 ,6 ]
机构
[1] Nankai Univ, State Environm Protect Key Lab Urban Ambient Air, Coll Environm Sci & Engn, Tianjin 300350, Peoples R China
[2] CMA NKU Cooperat Lab Atmospher Environm Hlth Res, Tianjin 300350, Peoples R China
[3] Univ Birmingham, Sch Geog Earth & Environm Sci, Birmingham B15 2TT, W Midlands, England
[4] Univ Birmingham, Dept Econ, Birmingham B15 2TT, W Midlands, England
[5] Rice Univ, Dept Civil & Environm Engn, Houston, TX 77005 USA
[6] Nankai Univ, Ctr Urban Transport Emiss Res, Coll Environm Sci & Engn, Tianjin 300350, Peoples R China
来源
ENVIRONMENTAL SCIENCE & TECHNOLOGY LETTERS | 2022年 / 9卷 / 02期
基金
国家重点研发计划; 中国国家自然科学基金;
关键词
METEOROLOGICAL NORMALIZATION; ANTHROPOGENIC EMISSIONS; PM2.5; TRENDS; IMPACT; AEROSOL; CHINA; PREDICTIONS; PARTICLES; SULFATE; MODELS;
D O I
10.1021/acs.estlett.1c00865
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
Many places on earth still suffer from a high level of atmospheric fine particulate matter (PM2.5) pollution. Formation of a particulate pollution event or haze episode (HE) involves many factors, including meteorology, emissions, and chemistry. Understanding the direct causes of and key drivers behind the HE is thus essential. Traditionally, this is done via chemical transport models. However, substantial uncertainties are introduced into the model estimation when there are significant changes in the emissions inventory due to interventions (e.g., the COVID-19 lockdown). Here we applied a Random Forest model coupled with a Shapley additive explanation algorithm, a post hoc explanation technique, to investigate the roles of major meteorological factors, primary emissions, and chemistry in five severe HEs that occurred before or during the COVID-19 lockdown in China. We discovered that, in addition to the high level of primary emissions, PM2.5 in these haze episodes was largely driven by meteorological effects (with average contributions of 30-65 mu g m(-3) for the five HEs), followed by chemistry (similar to 15-30 mu g m(-3)). Photochemistry was likely the major pathway of formation of nitrate, while air humidity was the predominant factor in forming sulfate. Our results highlight that the machine learning driven by data has the potential to be a complementary tool in predicting and interpreting air pollution.
引用
收藏
页码:112 / 119
页数:8
相关论文
共 57 条
[1]   Empirical comparison of tree ensemble variable importance measures [J].
Auret, Lidia ;
Aldrich, Chris .
CHEMOMETRICS AND INTELLIGENT LABORATORY SYSTEMS, 2011, 105 (02) :157-170
[2]   On Pixel-Wise Explanations for Non-Linear Classifier Decisions by Layer-Wise Relevance Propagation [J].
Bach, Sebastian ;
Binder, Alexander ;
Montavon, Gregoire ;
Klauschen, Frederick ;
Mueller, Klaus-Robert ;
Samek, Wojciech .
PLOS ONE, 2015, 10 (07)
[3]   Global estimates of mortality associated with long-term exposure to outdoor fine particulate matter [J].
Burnett, Richard ;
Chen, Hong ;
Szyszkowicz, Mieczyslaw ;
Fann, Neal ;
Hubbell, Bryan ;
Pope, C. Arden ;
Apte, Joshua S. ;
Brauer, Michael ;
Cohen, Aaron ;
Weichenthal, Scott ;
Coggins, Jay ;
Di, Qian ;
Brunekreef, Bert ;
Frostad, Joseph ;
Lim, Stephen S. ;
Kan, Haidong ;
Walker, Katherine D. ;
Thurston, George D. ;
Hayes, Richard B. ;
Lim, Chris C. ;
Turner, Michelle C. ;
Jerrett, Michael ;
Krewski, Daniel ;
Gapstur, Susan M. ;
Diver, W. Ryan ;
Ostro, Bart ;
Goldberg, Debbie ;
Crouse, Daniel L. ;
Martin, Randall V. ;
Peters, Paul ;
Pinault, Lauren ;
Tjepkema, Michael ;
van Donkelaar, Aaron ;
Villeneuve, Paul J. ;
Miller, Anthony B. ;
Yin, Peng ;
Zhou, Maigeng ;
Wang, Lijun ;
Janssen, Nicole A. H. ;
Marra, Marten ;
Atkinson, Richard W. ;
Tsang, Hilda ;
Thuan Quoc Thach ;
Cannon, John B. ;
Allen, Ryan T. ;
Hart, Jaime E. ;
Laden, Francine ;
Cesaroni, Giulia ;
Forastiere, Francesco ;
Weinmayr, Gudrun .
PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA, 2018, 115 (38) :9592-9597
[4]   Field Determination of Nitrate Formation Pathway in Winter Beijing [J].
Chen, Xiaorui ;
Wang, Haichao ;
Lu, Keding ;
Li, Chunmeng ;
Zhai, Tianyu ;
Tan, Zhaofeng ;
Ma, Xuefei ;
Yang, Xinping ;
Liu, Yuhan ;
Chen, Shiyi ;
Dong, Huabin ;
Li, Xin ;
Wu, Zhijun ;
Hu, Min ;
Zeng, Limin ;
Zhang, Yuanhang .
ENVIRONMENTAL SCIENCE & TECHNOLOGY, 2020, 54 (15) :9243-9253
[5]   The control of anthropogenic emissions contributed to 80% of the decrease in PM2.5 concentrations in Beijing from 2013 to 2017 [J].
Chen, Ziyue ;
Chen, Danlu ;
Kwan, Mei-Po ;
Chen, Bin ;
Gao, Bingbo ;
Zhuang, Yan ;
Li, Ruiyuan ;
Xu, Bing .
ATMOSPHERIC CHEMISTRY AND PHYSICS, 2019, 19 (21) :13519-13533
[6]   Spring Festival and COVID-19 Lockdown: Disentangling PM Sources in Major Chinese Cities [J].
Dai, Qili ;
Hou, Linlu ;
Liu, Bowen ;
Zhang, Yufen ;
Song, Congbo ;
Shi, Zongbo ;
Hopke, Philip K. ;
Feng, Yinchang .
GEOPHYSICAL RESEARCH LETTERS, 2021, 48 (11)
[7]   Dispersion Normalized PMF Provides Insights into the Significant Changes in Source Contributions to PM2.5 after the COVID-19 Outbreak [J].
Dai, Qili ;
Liu, Baoshuang ;
Bi, Xiaohui ;
Wu, Jianhui ;
Liang, Danni ;
Zhang, Yufen ;
Feng, Yinchang ;
Hopke, Philip K. .
ENVIRONMENTAL SCIENCE & TECHNOLOGY, 2020, 54 (16) :9917-9927
[8]   Seasonal differences in formation processes of oxidized organic aerosol near Houston, TX [J].
Dai, Qili ;
Schulze, Benjamin C. ;
Bi, Xiaohui ;
Bui, Alexander A. T. ;
Guo, Fangzhou ;
Wallace, Henry W. ;
Sanchez, Nancy P. ;
Flynn, James H. ;
Lefer, Barry L. ;
Feng, Yinchang ;
Griffin, Robert J. .
ATMOSPHERIC CHEMISTRY AND PHYSICS, 2019, 19 (14) :9641-9661
[9]   Residential coal combustion as a source of primary sulfate in Xi'an, China [J].
Dai, Qili ;
Bi, Xiaohui ;
Song, Wenbin ;
Li, Tingkun ;
Liu, Baoshuang ;
Ding, Jing ;
Xu, Jiao ;
Song, Congbo ;
Yang, Naiwang ;
Schulze, Benjamin C. ;
Zhang, Yufen ;
Feng, Yinchang ;
Hopke, Philip K. .
ATMOSPHERIC ENVIRONMENT, 2019, 196 :66-76
[10]   Impact of meteorological condition changes on air quality and particulate chemical composition during the COVID-19 lockdown [J].
Ding, Jing ;
Dai, Qili ;
Li, Yafei ;
Han, Suqin ;
Zhang, Yufen ;
Feng, Yinchang .
JOURNAL OF ENVIRONMENTAL SCIENCES, 2021, 109 :45-56