基于集成学习方法的中国近地面臭氧浓度时空分布
Estimating ground-level ozone concentration in China using ensemble learning methods
- 2023年27卷第8期 页码：1792-1806
DOI： 10.11834/jrs.20231845
引用
扫 描 看 全 文
摘要
自2013年大气污染防治行动以来，PM,2.5,、PM,10,、SO,2,、NO,2,等空气污染物浓度都有不同程度下降，但臭氧污染仍有上升趋势，臭氧污染已成为制约中国空气质量持续改善的关键问题。地基站点可以提供空间上特定点的臭氧浓度，但无法获得近地面臭氧连续的空间分布。由于臭氧大量分布于平流层，遥感卫星反演的臭氧柱浓度产品仅能反映整层臭氧柱浓度，但整层臭氧柱浓度与近地面浓度无明显相关性，因而无法体现近地面臭氧浓度。本文综合地基监测数据、再分析资料、卫星产品，采用不同的模型方法，得到近地面臭氧浓度的时空分布，结果表明集成学习方法可以准确估算近地面臭氧在空间上的分布状况和在时间上的变化趋势。本文对比了梯度提升回归树（GBRT）、极端随机树（ERT）、极端梯度提升器（XGBoost）3种不同的集成学习方法在近地面臭氧污染估算的效果表现，3种集成学习方法在2019年—2020两年的十折交叉验证,R,2,都在0.89以上，极端梯度提升器（XGBoost）方法在RMSE、MAE指标上有最好的表现，2019年—2020年两年的平均RMSE、MAE分别为15.77 μg/m,3,、10.53 μg/m,3,，但基于极端随机树（ERT）方法获得的近地面臭氧空间分布更加连续自然。因此最终选择极端随机树（ERT）方法估算得到中国近地面臭氧浓度数据集，并在此基础上进行时空分析。由于中国政府实施积极的减排措施的及疫情影响，臭氧浓度多年来的上升趋势得到了逆转，2020年臭氧年均值为107.41±18.6 μg/m,3,，较去年平均值109.26±19.71 μg/m,3,，有所减少。时间上，每年的5—9月气温较高，光化学反应剧烈，因而臭氧高污染事件频繁发生。空间上，京津冀地区、长三角地区、珠三角地区、成渝地区等城市群显著高于周围其他区域，是臭氧污染防治的重点区域。
Abstract
Following the successful implementation of the Air Pollution Prevention and Control Action Plan (2013—2017) and the Three-Year Action Plan to Win the Blue Sky Defense War (2018—2020), the concentrations of five major pollutants (i.e., PM2.5, PM10, SO,2, NO,2, and CO), except for ozone, significantly dropped for most cities in China. The increasing ground-level ozone concentrations have been a key factor restricting the improvement in ambient air quality, especially during summer.,Compared with the measurements from ground-based monitoring sites, satellite remote sensing technology can obtain spatially continuous total column ozone. However, given that ozone is abundantly distributed in the stratosphere, ground-level ozone has a very low contribution to the total column ozone observed from space. Therefore, the satellite total column ozone product provides limited information for estimating ground-level ozone concentrations. In this study, by combining TROPOMI ozone precursor (NO,2, and HCHO) products, ERA5 meteorological parameters, and ground-based monitoring data, a machine learning model was developed to estimate the daily maximum 8-hour average ground-level ozone concentration over China from years 2019 to 2020.,By comparing the performance of three ensemble learning methods, namely, extreme gradient boosters (XGBoost), extreme random trees (ERT), and gradient boost regression tree (GBRT), the averaged overall 10-fold cross-validation ,R,2, of 2019 and 2020 are all larger than 0.89. Although the results estimated by XGBoost showed the best agreement between the model predictions and observations with an average RMSE and MAE of 15.77 μg/m,3, and 10.53 μg/m,3, respectively, the ERT method was eventually selected to model the daily maximum 8-hour average ground-level ozone concentration by considering the rationalization of spatial distribution.,Due to the proactive emission reduction measures implemented by the Chinese government and the impact of the COVID-19 pandemic, the rising trend of ozone concentration over the years has been reversed. The annual average value of ground-level ozone concentration in 2020 reached 107.41±18.6 μg/m,3, over China, which is 1.85 μg/m,3, less than that recorded in 2019 (109.26±19.71 μg/m,3,). Severe surface ozone pollution events frequently occur from May to September of every year because the high temperatures during these months can promote photochemical reactions. The estimated ground-level ozone concentrations in the Beijing-Tianjin-Hebei, Yangtze River Delta, Pearl River Delta, and Chengdu-Chongqing regions are significantly higher than those in their surrounding areas, making these regions the key areas for ozone pollution prevention and control.
关键词
遥感近地面臭氧集成学习极端随机树梯度回归提升树极端梯度提升器TROPOMI新冠疫情
Keywords
remote sensingground-level ozoneEnsemble LearningERTGBRTXGBoostTROPOMICOVID-19
references
Bell M L. 2006. The use of ambient air quality modeling to estimate individual and population exposure for human health research: a case study of ozone in the Northern Georgia Region of the United States. Environment International, 32(5): 586-593 [DOI: 10.1016/j.envint.2006.01.005http://dx.doi.org/10.1016/j.envint.2006.01.005]
Breiman L. 2001. Random forests. Machine Learning, 45(1): 5-32 [DOI: 10.1023/A:1010933404324http://dx.doi.org/10.1023/A:1010933404324]
Chen L F, Wang Y P, Zhang X X, Wang Z F, Tao J H, Wang L L and Zhang Y. 2019. Satellite remote sensing monitoring of ozone and its precursors for regional secondary pollution risk control. Environmental Monitoring and Forewarning, 11(5): 13-21
陈良富, 王雅鹏, 张欣欣, 王子峰, 陶金花, 王莉莉, 张莹. 2019. 面向区域二次污染风险控制的臭氧及其前体物卫星遥感监测. 环境监控与预警, 11(5): 13-21 [DOI: 10.3969/j.issn.16746732.2019.05.003http://dx.doi.org/10.3969/j.issn.16746732.2019.05.003]
Chen R J, Chen B H and Kan H D. 2010. Health impact assessment of surface ozone pollution in Shanghai. China Environmental Science, 30(5): 603-608
陈仁杰, 陈秉衡, 阚海东. 2010. 上海市近地面臭氧污染的健康影响评价. 中国环境科学, 30(5): 603-608
Chen T Q and Guestrin C. 2016. XGBoost: a scalable tree boosting system//Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. San Francisco: ACM: 785-794 [DOI: 10.1145/2939672.2939785http://dx.doi.org/10.1145/2939672.2939785]
Cheng L J. 2018. A Study on the Evolution of Ozone Pollution in China and Regional Management Methods. Beijing: China University of Geosciences
程麟钧. 2018. 中国臭氧污染特征及分区管理方法研究. 北京: 中国地质大学
Chin M, Jacob D J, Munger J W, Parrish D D and Doddridge B G. 1994. Relationship of ozone and carbon monoxide over North America. Journal of Geophysical Research: Atmospheres, 99(D7): 14565-14573 [DOI: doi.org/10.1029/94JD00907http://dx.doi.org/doi.org/10.1029/94JD00907]
Dueñas C, Fernández M C, Cañete S, Carretero J and Liger E. 2002. Assessment of ozone variations and meteorological effects in an urban area in the Mediterranean Coast. Science of the Total Environment, 299(1/3): 97-113 [DOI: 10.1016/S0048-9697(02)00251-6http://dx.doi.org/10.1016/S0048-9697(02)00251-6]
Duncan B N, Yoshida Y, Olson J R, Sillman S, Martin R V, Lamsal L, Hu Y T, Pickering K E, Retscher C, Allen D J and Crawford J H. 2010. Application of OMI observations to a space-based indicator of NOx and VOC controls on surface ozone formation. Atmospheric Environment, 44(18): 2213-2223 [DOI: 10.1016/j.atmosenv.2010.03.010http://dx.doi.org/10.1016/j.atmosenv.2010.03.010]
Fang T. 2020. Study on Estimation and Prediction of Near-Surface Ozone based on Neural Network. Shanghai: Shanghai Normal University
方韬. 2020. 基于神经网络的近地面臭氧估算和预测研究. 上海: 上海师范大学 [DOI: 10.27312/d.cnki.gshsu.2020.002120http://dx.doi.org/10.27312/d.cnki.gshsu.2020.002120]
Feister U and Balzer K. 1991. Surface ozone and meteorological predictors on a subregional scale. Atmospheric Environment. Part A. General Topics, 25(9): 1781-1790 [DOI: 10.1016/0960-1686(91)90262-6http://dx.doi.org/10.1016/0960-1686(91)90262-6]
Friedman J, Hastie T and Tibshirani R. 2000. Additive logistic regression: a statistical view of boosting (With discussion and a rejoinder by the authors). The Annals of Statistics, 28(2): 337-407 [DOI: 10.1214/aos/1016218223http://dx.doi.org/10.1214/aos/1016218223]
Gardner M W and Dorling S R. 2000. Statistical surface ozone models: an improved methodology to account for non-linear behaviour. Atmospheric Environment, 34(1): 21-34 [DOI: 10.1016/S1352-2310(99)00359-3http://dx.doi.org/10.1016/S1352-2310(99)00359-3]
Geurts P, Ernst D and Wehenkel L. 2006. Extremely randomized trees. Machine Learning, 63(1): 3-42 [DOI: 10.1007/s10994-006-6226-1http://dx.doi.org/10.1007/s10994-006-6226-1]
Haagen-Smit A J. 1952. Chemistry and physiology of Los Angeles smog. Industrial and Engineering Chemistry, 44(6): 1342-1346 [DOI: 10.1021/ie50510a045http://dx.doi.org/10.1021/ie50510a045]
Sillman S. 1995. The use of NOy, H2O2, and HNO3 as indicators for ozone‐NOx‐hydrocarbon sensitivity in urban locations. Journal of Geophysical Research, 100: (D7):14175-14188[DOI: 10.1029/94JD02953http://dx.doi.org/10.1029/94JD02953]
Song Z G, Bai Y, Wang D F, Li T and He X Q. 2021. Satellite retrieval of air pollution changes in central and Eastern China during COVID-19 lockdown based on a machine learning model. Remote Sensing, 13(13): 2525 [DOI: 10.3390/rs13132525http://dx.doi.org/10.3390/rs13132525]
Tao J H, Fan M, Gu J B and Chen L F. 2020. Satellite observations of the return-to-work over China during the period of COVID-19. Journal of Remote Sensing (Chinese), 24(7): 824-836
陶金花, 范萌, 顾坚斌, 陈良富. 2020. 新冠病毒疫情期间复工复产卫星遥感监测. 遥感学报, 24(7): 824-836 [DOI: 10.11834/jrs.20200098http://dx.doi.org/10.11834/jrs.20200098]
Wang C, Wang S, Yang B B, Zhang Z H, Wang L and Liu M. 2015. Study of the effect of meteorological conditions on the ambient air ozone concentrations in Shenyang. Environmental Monitoring in China, 31(3): 32-37
王闯, 王帅, 杨碧波, 张丽辉, 王磊, 刘闽. 2015. 气象条件对沈阳市环境空气臭氧浓度影响研究. 中国环境监测, 31(3): 32-37 [DOI: 10.19316/j.issn.1002-6002.2015.03.007http://dx.doi.org/10.19316/j.issn.1002-6002.2015.03.007]
Wei J, Li Z Q, Li K, Dickerson R R, Pinker R T, Wang J, Liu X, Sun L, Xue W H and Cribb M. 2022. Full-coverage mapping and spatiotemporal variations of ground-level ozone (O3) pollution from 2013 to 2020 across China. Remote Sensing of Environment, 270: 112775 [DOI: 10.1016/j.rse.2021.112775http://dx.doi.org/10.1016/j.rse.2021.112775]
Zhan Y, Luo Y Z, Deng X F, Grieneisen M L, Zhang M H and Di B F. 2018. Spatiotemporal prediction of daily ambient ozone levels across China using random forest for human exposure assessment. Environmental Pollution, 233: 464-473 [DOI: 10.1016/j.envpol.2017.10.029http://dx.doi.org/10.1016/j.envpol.2017.10.029]
Zhang Y, Wang W H, Zhang X Y and Lawrence F. 2013. Interannual variations of Arctic ozone and their relationship to the polar vortex. Journal of Remote Sensing, 17(3): 527-540
张艳, 王维和, 张兴赢, Lawrence F. 2013. 北极臭氧年际变化特征及其与极涡的关系. 遥感学报, 17(3): 527-540 [DOI: 10.11834/jrs.20132037http://dx.doi.org/10.11834/jrs.20132037]
文章被引用时，请邮件提醒。
提交
相关作者
暂无数据
相关机构
中国资源卫星应用中心
中国气象局中国遥感卫星辐射测量和定标重点开放实验室/国家卫星气象中心(国家空间天气监测预警中心)
许健民气象卫星创新中心
中国科学院大学
中国科学院合肥物质科学研究院 安徽光机所环境光学与技术重点实验室
- 地址：北京市海淀区北四环西路19号 《遥感学报》编辑部 邮编：100190
- 联系电话：010-58887052 Email：nrsb@aircas.ac.cn
- 技术支持由北京北大方正电子有限公司提供 京ICP备20021838号-9
京公网安备11010802024621 - 本系统建议在Chrome、 IE9+ 以上版本浏览器阅读本站内容，360浏览器请切换至极速模式
- Cookies帮助我们提供服务并提供个性化体验。使用本网站，即表示您同意我们使用Cookies
0
回到顶部