耿红,宣莹莹,蔡夏童,张媛,周欢,张志红.太原市2014年春节期间常规大气污染物浓度变化及聚类分析[J].环境科学学报,2015,35(4):965-974
太原市2014年春节期间常规大气污染物浓度变化及聚类分析
- Mass concentration variation and cluster analysis of urban air pollutants in Taiyuan, Shanxi Province during Chinese New Year of 2014
- 基金项目：国家自然科学基金(No.21177078);环保部公益项目(No.201309009);山西省回国留学人员科研项目(No.2013-012);教育部留学回国人员科研启动基金(No.2013-192)
- 宣莹莹
- 山西大学资源与环境工程研究所, 国家环境保护煤炭废弃物资源化高效利用技术重点实验室, 太原 030006
- 蔡夏童
- 山西大学环境与资源学院, 太原 030006
- 张媛
- 1. 山西大学环境科学研究所, 太原 030006;2. 山西大学环境与资源学院, 太原 030006
- 周欢
- 1. 山西大学环境科学研究所, 太原 030006;2. 山西大学环境与资源学院, 太原 030006
- 张志红
- 山西医科大学公共卫生学院, 太原 030001
- 摘要：为了探寻太原市春节期间不同监测站点各常规大气污染物的质量浓度变化规律及相互之间的关系,记录和收集了太原市上兰、南寨、涧河、尖草坪、桃园、坞城、小店、金胜、晋源9个监测点2014年农历小年至元宵节(2014-01-23—2014-02-14)期间的大气PM2.5、PM10、CO、NO2、O3、SO2小时浓度值以及相应的气温、气压、湿度、风级、能见度等气象数据,采用相关分析、小波分析、单因子污染指数评价和系统聚类等方法进行研究,发现:①该时段内就太原市总体而言,PM2.5超标倍数最大,其次是PM10、SO2、CO、NO2,O3污染最小.②农历小年、除夕、正月初八、元宵节大气PM2.5和PM10的浓度迅速增加,与自然气象因子基本无关,说明烟花爆竹的集中燃放对大气颗粒物尤其是细颗粒物产生较大影响.③SO2、NO2、CO与PM2.5和PM10浓度变化的波动趋势相似、主周期相同,反映了部分PM2.5和PM10与SO2、NO2、CO有共同的来源; O3的波动趋势及主周期与上述污染物完全不同,显示出它来源的特殊性.④按PM2.5聚类,南寨、涧河、尖草坪、桃园4个点聚为一类,小店和坞城2个点聚为一类,金胜和晋源聚为一类,位于太原市最北端作为清洁对照的上兰监测点自成一类,与它们的地理位置有较好的相符性,同时,聚类分析结果与各监测点的单因子污染指数评价结果相一致.本文提示小波分析与聚类分析相结合可以较好地反映城市大气污染物浓度变化的时间与空间分布规律.
- Abstract：In order to investigate the fluctuation pattern of mass concentrations of conventional air pollutants (i.e. ambient PM2.5, PM10, CO, NO2, O3, and SO2) at different monitoring stations in urban Taiyuan, Shanxi Province, China during the Chinese New Year Spring Festival, we recorded and analyzed the data on the air pollutants' hourly concentrations together with temperature, air pressure, relative humidity, wind speed and visibility from Jan. 23 to Feb. 14, 2014 at nine regular monitoring sites, from north to south as Shanglan (site 1#), Nanzhai (site 2#), Jianhe (site 3#), Jiancaoping (site 4#), Taoyuan (site 5#), Wucheng (site 6#), Xiaodian (site 7#), Jinsheng (site 8#), and Jinyuan (site 9#). The results were obtained by using the wavelet analysis, single factor pollution index assessment, partial correlations analysis and cluster analysis. It was shown that ambient PM2.5 was most affected by fireworks, followed by PM10, SO2, CO, NO2, and O3. On the days of the 23rd of the twelfth lunar month (Jan 23), New Year's Eve (Jan 30), the 8th (Feb 7) and 15th (the Lantern Festival) of the first lunar month, the mass concentrations of ambient PM2.5 and PM10 increased dramatically without significant relationship with meteorological factors. Wavelet analyses demonstrated that the concentration fluctuation patterns and dominant periods of PM2.5 and PM10 were similar to those of SO2, NO2, and CO concentrations, implying that they shared the same sources. The quite different fluctuation pattern of O3 indicated that O3 had a different source. Cluster analyses based on square Euclidean Distance were in a good agreement with the assessment from single factor pollution index for all pollutants. For PM2.5, the 4 sites of 2#, 3#, 4#, and 5#, the 2 sites of 6# and 7#, the 2 sites of 8# and 9#, and the site 1# (as the clean control) were categorized, respectively, which were perfectly corresponded with their locations. In conclusion, the wavelet analysis with cluster analysis is an effective method for studying the temporal and spatial distributions of urban air pollutants.
摘要点击次数： 2022 全文下载次数： 2570
