科研之友
微信
新浪微博
Facebook
分享链接
摘要
Source samples including crustal dust, cement dust, coal combustion were sampled and ambient samples of PM2.5 and PM10 were synchronously collected in Hefei from April to December 2014. The size distributions of the markers in the measured source profiles were incorporated into ME-2 solution to develop a new method, called the SDABB model (an advanced ABB three-way factor analysis model incorporating size distribution information). The performance of this model was investigated using three-way synthetic and ambient dataset. For the synthetic tests, the size distributions of markers estimated by the SDABB model were more consistent with true condition. The AAEs between estimated and observed contributions of the SDABB ranged from 15.2% to 29.0% for PM10 and 19.9%-31.6% for PM2.5, which is lower than those of PMF2. For the ambient PM, six source categories were identified by SDABB for both sizes, although the profiles were different. The source contributions were sulphate (3333% and 24.53%), nitrate and SOC (22.33% and 18.16%), coal combustion (19.01% and 18.23%), vehicular exhaust (12.99% and 12.07%), crustal dust (10.69% and 19.40%) and cement dust (1.65% and 5.39%) for PM2.5 and PM10 respectively. In addition, the estimated ratios of Al, Si, Ti and Fe in CRD were 0.76, 0.84, 1.10 and 0.85; those of Al and Si in CC were 0.42 and 0.66; Ca and Si in CD were 0.95 and 1.10; NO3- and NH(4)(+)in nitrate were 1.11 and 1.01; and SO42- and NH4+ in sulphate were 0.96 and 1.16. These modeled ratios were consistent with the measured ratios. The size distribution of contributions also came close to reality. Thus, the advanced SDABB three-way model can better capture the characteristics of sources between sizes by effectively incorporating the size distributions of the markers as physical constraints.
- 出版日期2018-11
- 单位南开大学; 安徽省环境监测中心站
