基于随机森林模型的武汉市城区大气PM_(2.5)来源解析  

Source Analysis of Ambient PM_(2.5) in Wuhan City Based on Random Forest Model

作　　者：张志豪 陈楠 祝波 陶卉婷 成海容[1] ZHANG Zhi-hao;CHEN Nan;ZHU Bo;TAO Hui-ting;CHENG Hai-rong(School of Resource and Environment Science,Wuhan University,Wuhan 430072,China;Eco-Environment Monitoring Centre of Hubei Province,Wuhan 430074,China)

机构地区：[1]武汉大学资源与环境科学学院,武汉430072 [2]湖北省生态环境监测中心站,武汉430074

出　　处：《环境科学》2022年第3期1151-1158,共8页Environmental Science

基　　金：国家重点研发计划项目(2019YFB2102902)。

摘　　要：基于2019年12月~2020年11月期间武汉市城区大气PM_(2.5)及其主要化学组分(碳质组分、水溶性离子和元素组分)的在线监测数据,分析武汉城区大气PM_(2.5)的污染特征,并利用主成分分析方法和随机森林模型,对PM_(2.5)进行来源解析.结果表明,武汉市大气ρ(PM_(2.5))冬季最高,为(61.33±35.32)μg·m^(-3),而夏季最低,为(17.87±10.06)μg·m^(-3).其中碳质组分以有机碳为主,年均值为(7.27±3.51)μg·m^(-3),离子组分中ρ(NO_(3)-)、ρ(SO_(4)^(2-))和ρ(NH_(4)^(+))最高,年均值分别为(11.55±3.86)、(7.55±1.53)和(7.34±1.99)μg·m^(-3),元素组分中ρ(K)、ρ(Fe)和ρ(Ca)最高,年均值分别为(752.80±183.98)、(542.34±142.55)和(459.70±141.99)ng·m^(-3).通过主成分分析因子提取和随机森林定量分析,得到5类主要污染源,其在春、夏、秋、冬这4个季节贡献率结果分别如下:燃煤与二次源(46%、39%、41%、52%)、机动车排放源(22%、28%、27%、21%)、工业排放源(14%、18%、17%、13%)、扬尘源(10%、8%、6%、6%)和生物质燃烧源(8%、7%、9%、8%).最后对随机森林模型进行评价,发现4个季节模拟效果R^(2)均达到了0.85以上,处于较高水平,其中冬季(R^(2)=0.974)模型拟合效果最好,春季(R^(2)=0.936)与秋季(R^(2)=0.937)效果次之,夏季(R^(2)=0.866)表现相对较弱.Based on the online monitoring data of fine particle(PM_(2.5))mass concentration,carbonaceous components,ionic constituents,and elemental components in an urban site of Wuhan from December 2019 to November 2020,the chemical characteristics of PM_(2.5)were analyzed.In addition,seasonal source apportionment of PM_(2.5)was conducted using the principal component analysis(PCA)method and random forest(RF)algorithm model.The results indicated thatρ(PM_(2.5))was the highest in winter[(61.33±35.32)μg·m^(-3)]and the lowest in summer[(17.87±10.06)μg·m^(-3)].Furthermore,organic carbon(OC),with a concentration of(7.27±3.51)μg·m^(-3),accounted for the major proportion compared with that of elemental carbon(EC)in the carbonaceous component of PM_(2.5).NO_(3)-,SO_(4)^(2-),and NH_(4)^(+)had the highest proportion in ionic components,with concentrations of(11.55±3.86),(7.55±1.53),and(7.34±1.99)μg·m^(-3),respectively.K,Fe,and Ca were the main elements in elemental components,with concentrations of(752.80±183.98),(542.34±142.55),and(459.70±141.99)ng·m^(-3),respectively.Relying on main factor extraction by PCA and quantitative analysis by RF,five emission sources were ultimately confirmed.The seasonal concentration distribution of these emission sources was as follows:coal burning and secondary sources(46%,39%,41%,and 52%for spring,summer,autumn,and winter,respectively)made the highest contribution to PM_(2.5),followed by vehicle emission sources(22%,28%,27%,and 21%),industrial emission sources(14%,18%,17%,and 13%),dust sources(10%,8%,6%,and 6%),and biomass burning sources(8%,7%,9%,and 8%).The valuation of the RF model was evaluated using multiple indicators,including RMSE,MSE,and R^(2).The evaluation results showed that the model for winter had the best performance(R^(2)=0.974,RMSE=3.795μg·m^(-3),MAE=2.801μg·m^(-3)),the models for spring(R^(2)=0.936,RMSE=3.512μg·m^(-3),MAE=2.503μg·m^(-3))and autumn(R^(2)=0.937,RMSE=4.114μg·m^(-3),MAE=3.034μg·m^(-3))performed with moderate-fitting goodness,and the summ

关 键 词：PM_(2.5) 主成分分析(PCA) 随机森林(RF) 来源解析 污染特征 

分 类 号：X513[环境科学与工程—环境工程]