Source apportionment of PM10, PM2.5 and PM0.1 by positive matrix factorization analysis over three stations in Taiwan
- 作者：G.BalakrishnaiahHan WeiChun-Nan LiuAmit AwasthiChuen-Jinn Tsai
- 会议时间：2012-03-01
- 关键词：Positive matrix factorization ; Enrichment factor analysis ; conditional probability function analysis
- 作者单位：Institute of Environmental Engineering,Taiwan Chiao Tung University,Taiwan
- 母体文献：第九届海峡两岸气溶胶技术研讨会暨第二届海峡两岸环境保护双门论坛论文集
- 会议名称：第九届海峡两岸气溶胶技术研讨会暨第二届海峡两岸环境保护双门论坛
- 会议地点：厦门
- 主办单位：中国颗粒学会
- 语种：chi
摘要
Air pollution associated with atmospheric particulate matter (PM10,PM2.5 and PM0.1) is a serious problem in Taiwan.To provide a better understanding of source contributions,PM10,PM2.5 and PM0.1 samples were collected simultaneously over a 24-h period during May 2011 to November 2011 at three different air monitors (Zhongshan,Shinjung,Jhudong) in Taiwan.Samples were analyzed for 17 metallic trace elements (Al,Fe,Na,Mg,K,Ca,Ba,Ti,Mn,Cu,Zn,Pb,V and Cr),the ionic compounds (F-,Cl-,? 3 NO,? 4 SO,+ 4 NH ).The average PM10 concentrations were found to be 37.46±10.73 μg/m3 in Zhongshan,39.45±11.58 μg/m3 in Shinjung and 32.15±12.02 μg/m3 in Jhudong,the PM2.5 concentrations were 24.44±8.01 μg/m3,21.82±7.50 μg/m3 and 12.91±5.67 μg/m3,respectively and PM0.1 mass concentrations were 1.37±0.57 μg/m3,1.42±0.56 μg/m3 and 0.7±0.40 μg/m3 respectively for the three stations.Positive Matrix Factorization (EPA-PMF3.0) was applied to these data sets to identify the possible sources.A total of five source types were identified to be important which include: soil dust,vehicle exhaust,marine aerosols,industrial pollution and secondary aerosols.Conditional probability function (CPF) analysis was performed to identify local source as well as remote source regions affecting the monitoring stations.Results of our study suggest contending relationships between anthropogenic and natural source profiles over three monitoring stations which depend on PM types and locations of the receptors.