PM2.5 source apportionment identified with total and soluble elements in positive matrix factorization.

Source apportionments of urban aerosols identified with positive matrix factorization (PMF) are sensitive to input variables. So far, total elements were frequently included as effective factors in PMF-based source apportionment. We investigated the advances to involve soluble parts of elements in the source apportionment with four data sets of PM2.5 composition observed at a coastal city (Qingdao) in northern China: water-soluble ions plus organic and elemental carbon (IC set), the IC set plus total elements (ICTE set), the IC set plus soluble elements (ICSE set), and the IC set plus both total elements and soluble elements (ICAE set). The apportionments of six sources, including secondary sulfate, secondary nitrate, secondary oxalate, sea salt, biomass burning and dust, were identified with the IC set. In comparison, pollutants from vehicle + coal combustion, ship emissions, waste incineration and industrial activities were also identified with the ICTE, ICSE, or ICAE sets. We found that the PMF solutions of the ICAE set could distinguish aged and fresh dust, and identify fly ash and aged pollutants from industrial sources. The profiles and corresponding time series of vehicle + coal combustion, secondary aerosols, ship emissions, sea salt, and biomass burning emissions identified with the four data sets were very similar, while discrepancies were encountered for waste incineration, dust, and industrial sources. These results indicate the benefits and potentials with total and soluble elements involved in PMF-based source apportionments.