Elucidate long-term changes of ozone in Shanghai based on an integrated machine learning method

Highlights

● A novel integrated machine learning method to analyze O₃ changes is proposed.

● Various factors affecting long-term changes of O₃ in Shanghai are quantified.

● Meteorological, photochemical, and regional background O₃ are well separated.

Abstract

Surface ozone (O₃) is influenced by regional background and local photochemical formation under favorable meteorological conditions. Understanding the contribution of these factors to changes in O₃ is crucial to address the issue of O₃ pollution. In this study, we propose a novel integrated method that combines random forest, principal component analysis, and Shapley additive explanations to distinguish observed O₃ into meteorologically affected ozone (O_{3_MET}), chemically formed from local emissions (O_{3_LC}), and regional background ozone (O_{3_RBG}). Applied to three typical stations in Shanghai during the warm season from 2013 to 2021, the results indicate that O_{3_RBG} in Shanghai was 48.8 ± 0.3 ppb, accounting for 79.6%–89.4% at different sites, with an overall declining trend of 0.018 ppb/yr. O_{3_LC} at urban and regional sites ranged from 5.9–9.0 ppb and 8.9–14.6 ppb, respectively, which were significantly higher than the contributions of 2.5–7.4 ppb at an upwind background site. O_{3_MET} can be categorized into those affecting O₃ photochemical generation and those changing O₃ dispersion conditions, with absolute contributions to O₃ ranging from 13.4–19.0 ppb and 13.1–13.7 ppb, respectively. We found that the O₃ rebound in 2017, compared to 2013, was primarily influenced by unfavorable O₃ dispersion conditions and unbalanced emission reductions; while the O₃ decline in 2021, compared to 2017, was primarily influenced by overall favorable meteorological conditions and further emissions reduction. These findings highlight the challenge of understanding O₃ change due to meteorology and regional background, emphasizing the need for systematic interpretation of the different components of O₃.

Graphical abstract

Keywords

Ozone / Integrated method / Machine learning

References

[1] Alduchov O A, Eskridge R E. (1996). Improved Magnus form approximation of saturation vapor pressure. Journal of Applied Meteorology, 35(4): 601–609

[2] Austin E, Xiang J, Gould T R, Shirai J H, Yun S, Yost M G, Larson T V, Seto E. (2021). Distinct ultrafine particle profiles associated with aircraft and roadway traffic. Environmental Science & Technology, 55(5): 2847–2858

[3] Berlin S R, Langford A O, Estes M, Dong M, Parrish D D. (2013). Magnitude, decadal changes, and impact of regional background ozone transported into the greater Houston, Texas, area. Environmental Science & Technology, 47(24): 13985–13992

[4] Camalier L, Cox W, Dolwick P. (2007). The effects of meteorology on ozone in urban areas and their use in assessing ozone trends. Atmospheric Environment, 41(33): 7127–7137

[5] Chen L, Zhu J, Liao H, Yang Y, Yue X. (2020). Meteorological influences on PM_2.5 and O₃ trends and associated health burden since China’s clean air actions. Science of the Total Environment, 744: 140837

[6] Chen W, Guenther A B, Shao M, Yuan B, Jia S, Mao J, Yan F, Krishnan P, Wang X. (2022). Assessment of background ozone concentrations in China and implications for using region-specific volatile organic compounds emission abatement to mitigate air pollution. Environmental Pollution, 305: 119254

[7] Dai Q, Hou L, Liu B, Zhang Y, Song C, Shi Z, Hopke P K, Feng Y. (2021). Spring festival and COVID-19 lockdown: disentangling PM sources in major Chinese cities. Geophysical Research Letters, 48(11): e2021GL093403

[8] Dang R, Liao H, Fu Y. (2021). Quantifying the anthropogenic and meteorological influences on summertime surface ozone in China over 2012–2017. Science of the Total Environment, 754: 142394

[9] Ding J, Dai Q, Fan W, Lu M, Zhang Y, Han S, Feng Y. (2023). Impacts of meteorology and precursor emission change on O₃ variation in Tianjin, China from 2015 to 2021. Journal of Environmental Sciences-China, 126: 506–516

[10] Gao D, Xie M, Liu J, Wang T, Ma C, Bai H, Chen X, Li M, Zhuang B, Li S. (2021). Ozone variability induced by synoptic weather patterns in warm seasons of 2014–2018 over the Yangtze River Delta region, China. Atmospheric Chemistry and Physics, 21(8): 5847–5864

[11] González-Pardo J, Ceballos-Santos S, Manzanas R, Santibáñez M, Fernández-Olmo I. (2022). Estimating changes in air pollutant levels due to COVID-19 lockdown measures based on a business-as-usual prediction scenario using data mining models: a case-study for urban traffic sites in Spain. Science of the Total Environment, 823: 153786

[12] Grange S K, Carslaw D C. (2019). Using meteorological normalisation to detect interventions in air quality time series. Science of the Total Environment, 653: 578–588

[13] Grange S K, Carslaw D C, Lewis A C, Boleti E, Hueglin C. (2018). Random forest meteorological normalisation models for Swiss PM₁₀ trend analysis. Atmospheric Chemistry and Physics, 18(9): 6223–6239

[14] Henneman L R F, Holmes H A, Mulholland J A, Russell A G. (2015). Meteorological detrending of primary and secondary pollutant concentrations: method application and evaluation using long-term (2000–2012) data in Atlanta. Atmospheric Environment, 119: 201–210

[15] Hirsch R M, Slack J R, Smith R A. (1982). Techniques of trend analysis for monthly water quality data. Water Resources Research, 18(1): 107–121

[16] Hou L, Dai Q, Song C, Liu B, Guo F, Dai T, Li L, Liu B, Bi X, Zhang Y. . (2022). Revealing drivers of haze pollution by explainable machine learning. Environmental Science & Technology Letters, 9(2): 112–119

[17] Hu C, Kang P, Jaffe D A, Li C, Zhang X, Wu K, Zhou M. (2021). Understanding the impact of meteorology on ozone in 334 cities of China. Atmospheric Environment, 248: 118221

[18] Huang Y, Chen J, Duan Q, Feng Y, Luo R, Wang W, Liu F, Bi S, Lee J. (2022). A fast antibiotic detection method for simplified pretreatment through spectra-based machine learning. Frontiers of Environmental Science & Engineering, 16(3): 38

[19] Joliffe I, Morgan B. (1992). Principal component analysis and exploratory factor analysis. Statistical Methods in Medical Research, 1(1): 69–95

[20] Langford A O, Senff C J, Banta R M, Hardesty R M, Alvarez R J II, Sandberg S P, Darby L S. (2009). Regional and local background ozone in Houston during Texas air quality study 2006. Journal of Geophysical Research, 114(D7): D00F12

[21] Lelieveld J, Evans J S, Fnais M, Giannadaki D, Pozzer A. (2015). The contribution of outdoor air pollution sources to premature mortality on a global scale. Nature, 525(7569): 367–371

[22] Li C, Zhu Q, Jin X, Cohen R C. (2022). Elucidating contributions of anthropogenic volatile organic compounds and particulate matter to ozone trends over China. Environmental Science & Technology, 56(18): 12906–12916

[23] Li K, Jacob D J, Liao H, Qiu Y, Shen L, Zhai S, Bates K H, Sulprizio M P, Song S, Lu X. . (2021a). Ozone pollution in the North China Plain spreading into the late-winter haze season. Proceedings of the National Academy of Sciences of the United States of America, 118(10): e2015797118

[24] Li K, Jacob D J, Liao H, Shen L, Zhang Q, Bates K H. (2019). Anthropogenic drivers of 2013–2017 trends in summer surface ozone in China. Proceedings of the National Academy of Sciences of the United States of America, 116(2): 422–427

[25] Li K, Jacob D J, Shen L, Lu X, De Smedt I, Liao H. (2020). Increases in surface ozone pollution in China from 2013 to 2019: anthropogenic and meteorological influences. Atmospheric Chemistry and Physics, 20(19): 11423–11433

[26] Li X B, Fan G, Lou S, Yuan B, Wang X, Shao M. (2021b). Transport and boundary layer interaction contribution to extremely high surface ozone levels in eastern China. Environmental Pollution, 268: 115804

[27] Lin C, Lau A K H, Fung J C H, Song Y, Li Y, Tao M, Lu X, Ma J, Lao X Q. (2021). Removing the effects of meteorological factors on changes in nitrogen dioxide and ozone concentrations in China from 2013 to 2020. Science of the Total Environment, 793: 148575

[28] Liu Y, Wang T. (2020a). Worsening urban ozone pollution in China from 2013 to 2017 – Part 1: The complex and varying roles of meteorology. Atmospheric Chemistry and Physics, 20(11): 6305–6321

[29] Liu Y, Wang T. (2020b). Worsening urban ozone pollution in China from 2013 to 2017 – Part 2: The effects of emission changes and implications for multi-pollutant control. Atmospheric Chemistry and Physics, 20(11): 6323–6337

[30] Lovrić M, Pavlović K, Vuković M, Grange S K, Haberl M, Kern R. (2021). Understanding the true effects of the COVID-19 lockdown on air pollution by means of machine learning. Environmental Pollution, 274: 115900

[31] Lu W, Huo W, Gulina H, Pan C. (2022). Development of machine learning multi-city model for municipal solid waste generation prediction. Frontiers of Environmental Science & Engineering, 16(9): 119

[32] Lu X, Zhang L, Chen Y, Zhou M, Zheng B, Li K, Liu Y, Lin J, Fu T M, Zhang Q. (2019). Exploring 2016–2017 surface ozone pollution over China: source contributions and meteorological influences. Atmospheric Chemistry and Physics, 19(12): 8339–8361

[33] Lundberg S M, Erion G, Chen H, Degrave A, Prutkin J M, Nair B, Katz R, Himmelfarb J, Bansal N, Lee S I. (2020). From local explanations to global understanding with explainable AI for trees. Nature Machine Intelligence, 2(1): 56–67

[34] LundbergS MLee S I (2017). A Unified Approach to Interpreting Model Predictions. In: Proceedings of the 31st International Conference on Neural Information Processing Systems, Long Beach, California, USA. Red Hook, NY: Curran Associates Inc, 4768–4777

[35] Mousavinezhad S, Choi Y, Pouyaei A, Ghahremanloo M, Nelson D L. (2021). A comprehensive investigation of surface ozone pollution in China, 2015–2019: Separating the contributions from meteorology and precursor emissions. Atmospheric Research, 257: 105599

[36] Ni R, Lin J, Yan Y, Lin W. (2018). Foreign and domestic contributions to springtime ozone over China. Atmospheric Chemistry and Physics, 18(15): 11447–11469

[37] Pathakoti M, Santhoshi T, Aarathi M, Mahalakshmi D V, Kanchana A L, Srinivasulu J, Shekhar R S S, Soni V K, Sai S M V R, Raja P. (2021). Assessment of spatio-temporal climatological trends of ozone over the Indian region using machine learning. Spatial Statistics, 43: 100513

[38] Qu L, Liu S, Ma L, Zhang Z, Du J, Zhou Y, Meng F. (2020). Evaluating the meteorological normalized PM_2.5 trend (2014–2019) in the “2+26” region of China using an ensemble learning technique. Environmental Pollution, 266: 115346

[39] Sahu S K, Liu S, Liu S, Ding D, Xing J. (2021). Ozone pollution in China: Background and transboundary contributions to ozone concentration & related health effects across the country. Science of the Total Environment, 761: 144131

[40] Sen P K. (1968). Estimates of the regression coefficient based on Kendall’s tau. Journal of the American Statistical Association, 63(324): 1379–1389

[41] Shi Z, Song C, Liu B, Lu G, Xu J, Van Vu T, Elliott R J R, Li W, Bloss W J, Harrison R M. (2021). Abrupt but smaller than expected changes in surface air quality attributable to COVID-19 lockdowns. Science Advances, 7(3): eabd6696

[42] Song C, Becagli S, Beddows D C S, Brean J, Browse J, Dai Q, Dall’osto M, Ferracci V, Harrison R M, Harris N. . (2022). Understanding sources and drivers of size-resolved aerosol in the high Arctic Islands of Svalbard using a receptor model coupled with machine learning. Environmental Science & Technology, 56(16): 11189–11198

[43] Stein A F, Draxler R R, Rolph G D, Stunder B J B, Cohen M D, Ngan F. (2015). NOAA’s HYSPLIT atmospheric transport and dispersion modeling system. Bulletin of the American Meteorological Society, 96(12): 2059–2077

[44] Suciu L G, Griffin R J, Masiello C A. (2017). Regional background O₃ and NO_x in the Houston–Galveston–Brazoria (TX) region: a decadal-scale perspective. Atmospheric Chemistry and Physics, 17(11): 6565–6581

[45] Tang G, Liu Y, Huang X, Wang Y, Hu B, Zhang Y, Song T, Li X, Wu S, Li Q. . (2021). Aggravated ozone pollution in the strong free convection boundary layer. Science of the Total Environment, 788: 147740

[46] Tang M X, Huang X F, Sun T L, Cheng Y, Luo Y, Chen Z, Lin X Y, Cao L M, Zhai Y H, He L Y. (2022). Decisive role of ozone formation control in winter PM_2.5 mitigation in Shenzhen, China. Environmental Pollution, 301: 119027

[47] Vu T V, Shi Z, Cheng J, Zhang Q, He K, Wang S, Harrison R M. (2019). Assessing the impact of clean air action on air quality trends in Beijing using a machine learning technique. Atmospheric Chemistry and Physics, 19(17): 11303–11314

[48] Wang F T, Zhang K, Xue J, Huang L, Wang Y J, Chen H, Wang S Y, Fu J S, Li L. (2022a). Understanding regional background ozone by multiple methods: a case study in the Shandong region, China, 2018–2020. Journal of Geophysical Research. Atmospheres, 127(22): e2022JD036809

[49] Wang P, Guo H, Hu J, Kota S H, Ying Q, Zhang H. (2019). Responses of PM_2.5 and O₃ concentrations to changes of meteorology and emissions in China. Science of the Total Environment, 662: 297–306

[50] Wang T, Xue L, Brimblecombe P, Lam Y F, Li L, Zhang L. (2017). Ozone pollution in China: a review of concentrations, meteorological influences, chemical precursors, and effects. Science of the Total Environment, 575: 1582–1596

[51] Wang W, Parrish D D, Wang S, Bao F, Ni R, Li X, Yang S, Wang H, Cheng Y, Su H. (2022b). Long-term trend of ozone pollution in China during 2014–2020: distinct seasonal and spatial characteristics and ozone sensitivity. Atmospheric Chemistry and Physics, 22(13): 8935–8949

[52] Wang Y, Wen Y, Wang Y, Zhang S, Zhang K M, Zheng H, Xing J, Wu Y, Hao J. (2020). Four-month changes in air quality during and after the COVID-19 lockdown in six megacities in China. Environmental Science & Technology Letters, 7(11): 802–808

[53] Wren S N, Mihele C M, Lu G, Jiang Z, Wen D, Hayden K, Mittermeier R L, Staebler R M, Cober S G, Brook J R. (2020). Improving insights on air pollutant mixtures and their origins by enhancing local monitoring in an area of intensive resource development. Environmental Science & Technology, 54(23): 14936–14945

[54] Wu Q, Li T, Zhang S, Fu J, Seyler B C, Zhou Z, Deng X, Wang B, Zhan Y. (2022). Evaluation of NO_x emissions before, during, and after the COVID-19 lockdowns in China: A comparison of meteorological normalization methods. Atmospheric Environment, 278: 119083

[55] Xu J, Tie X, Gao W, Lin Y, Fu Q. (2019). Measurement and model analyses of the ozone variation during 2006 to 2015 and its response to emission change in megacity Shanghai, China. Atmospheric Chemistry and Physics, 19(14): 9017–9035

[56] Yang L, Luo H, Yuan Z, Zheng J, Huang Z, Li C, Lin X, Louie P K K, Chen D, Bian Y. (2019). Quantitative impacts of meteorology and precursor emission changes on the long-term trend of ambient ozone over the Pearl River Delta, China, and implications for ozone control strategy. Atmospheric Chemistry and Physics, 19(20): 12901–12916

[57] Yang P, Yang H, Sardans J, Tong C, Zhao G, Peñuelas J, Li L, Zhang Y, Tan L, Chun K P. . (2020). Large spatial variations in diffusive CH₄ fluxes from a subtropical coastal reservoir affected by sewage discharge in Southeast China. Environmental Science & Technology, 54(22): 14192–14203

[58] Yin H, Lu X, Sun Y, Li K, Gao M, Zheng B, Liu C. (2021). Unprecedented decline in summertime surface ozone over eastern China in 2020 comparably attributable to anthropogenic emission reductions and meteorology. Environmental Research Letters, 16(12): 124069

[59] Zhang K, Liu Z, Zhang X, Li Q, Jensen A, Tan W, Huang L, Wang Y, De Gouw J, Li L. (2022). Insights into the significant increase of ozone during COVID-19 in a typical urban city of China. Atmospheric Chemistry and Physics, 22(7): 4853–4866

[60] Zhang Y, Vu T V, Sun J, He J, Shen X, Lin W, Zhang X, Zhong J, Gao W, Wang Y. . (2020). Significant changes in chemistry of fine particles in wintertime Beijing from 2007 to 2017: impact of clean air actions. Environmental Science & Technology, 54(3): 1344–1352

[61] Zhou W, Lei L, Du A, Zhang Z, Li Y, Yang Y, Tang G, Chen C, Xu W, Sun J. . (2022). Unexpected increases of severe haze pollution during the post COVID-19 period: effects of emissions, meteorology, and secondary production. Journal of Geophysical Research: Atmospheres, 127(3): e2021JD035710

[62] Zhu Q, Gu A, Li D, Zhang T, Xiang L, He M. (2021). Online recognition of drainage type based on UV-vis spectra and derivative neural network algorithm. Frontiers of Environmental Science & Engineering, 15(6): 136

Authorship Contribution

JX performed the data analysis and prepared the manuscript with contributions from all co-authors. FT Wang conducted the PCA analysis. KZ conducted the MK and ML analysis. DJ, YSD, YWL and QYF conducted the field observations. HHZ, EY, YJW, LH and TL helped to interpret the data analysis results. JSF reviewed the manuscript. LL formulated the research goals, led data analysis and discussions, edited and reviewed the manuscript. All authors contributed to data interpretations and discussions.

Competing Interests

The authors declare no conflict of interest.

Data and Code Availability

Data are available upon request to the corresponding authors. The IM model is available through a GitHub repository: https://github.com/Xuejin66/An-integrated-machine-learning-approach-that-can-elucidate-factors-influencing-long-term-ozone-chang.

Acknowledgements

This study is financially supported by the Shanghai Municipal Bureau of Ecology and Environment (China) ([2022]37), National Natural Science Foundation of China (NOs. 42075144 and 42005112) and Key Research and Development Project of Shanghai Science and Technology Commission, China (No.20dz1204000). We appreciate State Ecology and Environment Scientific Observation and Research Station for the Yangtze River Delta at Dianshan Lake (SEED) (China) for supporting the study.

Electronic Supplementary Material

Supplementary material is available in the online version of this article at https://doi.org/10.1007/s11783-023-1738-5 and is accessible for authorized users.

RIGHTS & PERMISSIONS

2023 Higher Education Press