天津市PM_(2.5)在线源解析成分特征谱库构建研究

Research on Characteristic Database Construction of PM_(2.5) On-line Source Apportionment in Tianjin

根据天津市颗粒物源解析结果,选取扬尘、燃煤、机动车尾气、工业生产、生物质燃烧及餐饮油烟等典型PM_(2.5)排放源,利用单颗粒气溶胶质谱仪开展天津市PM_(2.5)在线源解析成分特征谱库的构建研究。应用自适应共振理论神经网络分类算法(ART-2a)对采集的数据进行分析,确定了天津市PM_(2.5)的成分谱特征。通过对PM_(2.5)源谱离线与在线组分测试结果的相关性分析,得出各类污染物在线源解析源谱具有较好的代表性,能够定性反映各类污染源的基本信息,可作为在线源解析受体样品进行匹配的依据使用。

This paper describes the setting-up of a characteristic spectrum library of PM2.5 online source components in Tianjin by means of SPAMS（single particle aerosol mass spectrometer） based on the results of source apportionment of particulate matters in Tianjin, several typical PM2.5 emission sources, such as flying dust, coal-burning, motor vehicle exhaust, industrial production, biomass combustion, cooking fume of catering business, and etc. The analysis of the data acquired was determined by ART-2a, an adaptive and clustering algorithms method, and thus the spectral characteristics of the components of PM2.5 in Tianjin were determined. According to the results of correlation analysis of PM2.5 composition off-line and online measurements, it was concluded that being enough to reflect the basic information regarding various contaminants the online source apportionment spectra of various pollutants were of good representativeness, so it could be used for matching the online source analysis receptor samples.