基于机器学习评估烟花爆竹燃放对春节期间PM_(2.5)浓度的影响  

Evaluation on impact of fireworks on PM_(2.5) concentration during the Spring Festival based on machine learning

作　　者：李津慧 胡启后 李启华 季祥光 薛劲凯 林培泽 LI Jinhui;HU Qihou;LI Qihua;JI Xiangguang;XUE Jingkai;LIN Peize(Institutes of Physical Science and Information Technology,Anhui University,Hefei 230601,China;Key Laboratory of Environment Optics and Technology,Anhui Institute of Optics and Fine Mechanics,Hefei Institutes of Physical Science,Chinese Academy of Sciences,Hefei 230031,China;School of Environmental Science and Optoelectronic Technology,University of Science and Technology of China,Hefei 230026,China)

机构地区：[1]安徽大学物质科学与信息技术研究院,合肥230601 [2]中国科学院合肥物质科学研究院,安徽光学精密机械研究所中国科学院环境光学与技术重点实验室,合肥230031 [3]中国科学技术大学环境科学与光电技术学院,合肥230026

出　　处：《环境工程》2025年第3期90-102,共13页Environmental Engineering

基　　金：安徽省重点研究与开发计划项目“PM2.5和O3无盲区垂直廓线协同探测技术与装备”(2023t07020015)。

摘　　要：自《大气污染防治行动计划》颁布以来,我国实施了严格的污染防治措施,并取得了显著成果。然而,其中的烟花爆竹禁放政策自执行起就饱受争议。尤其是随着我国大部分城市PM_(2.5)年均浓度达到GB 3095—2012《环境空气质量标准》二级标准,亟需量化烟花爆竹禁放在空气质量改善中所起的作用。基于我国环境监测总站空气质量监测网络的观测资料和ERA5气象数据,通过构建随机森林模型,解耦气象因素和人为排放对颗粒物浓度的贡献,评估烟花爆竹禁放与否对春节期间、冬季和全年颗粒物浓度的影响。从年际变化来看,2016—2023年全国各城市PM_(2.5)浓度呈显著下降趋势,年均下降速率达到了4.2%~7.5%。对比全年、冬季、春节期间3个时间尺度上人为减排引起的PM_(2.5)浓度变化速率,以北京、上海为例,全年平均下降9.58%、7.36%,冬季年均下降10.07%、7.94%,春节期间下降9.14%、6.23%,基本呈冬季>全年>春节期间的规律,即同一城市全年和冬季人为排放引起浓度变化明显,但春节期间变化微弱。对比2016—2022年和2019—2023年春节期间人为减排引起的PM_(2.5)浓度变化,同一城市2016—2022年人为减排效果明显,变化速率较快,以北京、广州、合肥PM_(2.5)浓度下降幅度居前列,分别下降9.14%、8.57%、6.71%;2019—2023年减排速率缓慢,特别是广州、成都、长沙三地PM_(2.5)浓度出现了13.69%、11.42%、5.77%的增长。从短期分析烟花爆竹排放变化对空气治理的影响,选取限放—禁放和禁放—限放,即2017与2018、2022与2023两个阶段研究春节期间气象和人为排放对颗粒物浓度变化的贡献,结果表明,2018年全面禁放之后排放贡献的PM_(2.5)浓度下降明显,北京、广州、成都、长沙、合肥分别下降14.87,4.54,26.17,15.67,27.61μg/m^(3);2023年解禁以后排放贡献显著上升,尤其广州、成都、长沙、郑州、滨州、东营的排放贡献分别�Since the enactment of the Air Pollution Prevention and Control Action Plan, China has implemented strict pollution control measures and achieved significant results. However, the ban on fireworks has been controversial since its inception. With the annual PM_(2.5) concentration in most cities reaching the level Ⅱ of Ambient Air Quality Standard(GB 3095 —2012), it is urgent to quantify the role of the fireworks ban in improving air quality. This paper utilized observational data from China′s National Environmental Monitoring Center and ERA5 meteorological data to construct a random forest model, decoupling the contributions of meteorological factors and human emissions to particulate matter concentrations. It assessed the impact of the fireworks ban on PM_(2.5) concentrations during the Spring Festival, winter, and the entire year. From an interannual perspective, the PM_(2.5) concentrations in cities across China significantly decreased from 2016to 2023, with an annual average decline rate of 4.2% to 7.5%. Comparing the PM_(2.5) concentration change rates due to human emissions in three time scales:annual, winter, and the Spring Festival,using Beijing and Shanghai as examples, the annual average decline rates were 9.58% and 7.36%, winter′s annual average decline rates were 10.07% and 7.94%, and the Spring Festival decline rates were 9.14% and 6.23%. This indicated a pattern of winter > annual > the Spring Festival, meaning that in the same city, the concentration changes due to human emissions were significant throughout the year and in winter, but minimal during the Spring Festival. Comparing the PM_(2.5) concentration changes due to human emissions in the Spring Festival from 2016 to 2022, and 2019 to 2023, the reduction effects were evident in the same city from 2016 to 2022, with faster change rates. Notably, PM_(2.5) concentrations in Beijing, Guangzhou, and Hefei decreased by 9.14%, 8.57%, and 6.71%, respectively. The emission reduction rates were slowed down from 2019 to 2023. Moreover,some cities

关 键 词：机器学习 随机森林 大气污染 烟花爆竹燃放 

分 类 号：F42[经济管理—产业经济]