苏州市区大气监测点位调整中的相关性和聚类分析  

Correlation analysis and clustering analysis to the adjustment of atmospheric monitoring sites in urban area of Suzhou

作　　者：吴也正[1] 

机构地区：[1]苏州市环境监测中心,江苏苏州215004

出　　处：《干旱环境监测》2016年第2期49-53,87,共6页Arid Environmental Monitoring

基　　金：苏州市科技支撑计划项目(SS201524)

摘　　要：根据苏州市8个国控大气监测站点2013—2015年的PM_(10)、PM_(2.5)、O_3、SO_2监测数据,计算了各站点之间监测数据的相关系数;通过分析相关系数矩阵,发现苏州市大气国控点位的监测数据相关性均较高,点位设置存在一定的冗余,国控监测点位的优化调整十分必要。利用最长距离法对各站点的监测数据进行聚类分析,从而得出中心城区的点位调整方案。点位调整前后4种污染物监测数据四季平均值的相对误差均小于5%;正负偏差的二项检验结果表明,调整后未对监测结果产生显著的正/负影响。因此,经过优化的大气监测网络具有较好的客观性和代表性。Based on the concentration data of pollutants（ PM_（10）,PM_（2. 5）,ozone and sulfur dioxide） monitored by the 8 atmospheric monitoring sites in the urban area of Suzhou from 2013 to 2015,the sites＇ correlation coefficient matrices of the pollutants were calculated,and the monitoring data between sites were found highly correlated,which means it was very necessary to adjust the atmospheric monitoring sites. By the clustering analysis of monitoring data of all the sites with complete linkage method,we obtained the adjustment scheme. The fractional errors of all the seasonal average concentrations of the pollutants before and after the adjustment was all within 5%. To investigate whether the variations of the monitoring data caused by the adjustment was significant positive or negative,a binomial test was applied,and the test result indicated the total variation caused by the adjustment was neither significant positive nor significant negative. Above all,after the adjustment,the new atmospheric monitoring sites are objective and representative.

关 键 词：大气监测点位调整 相关性分析 聚类分析 二项检验 苏州市区 

分 类 号：X830.1[环境科学与工程—环境工程]