| Summary: | A size-segregated chemical composition of atmospheric aerosols was investigated in May 2004 at the SMEAR II station, southern Finland. Aerosols were collected using two 12-stage low pressure impactors (SDI) and two virtual impactors (VI). The samples were analyzed for mass, inorganic ions and organic (OC) and elemental carbon (EC). By comparing the gravimetric mass and the results from the chemical analyses, a chemical mass closure was constructed. In addition to the impactors an Electrical Low Pressure Impactor (ELPI), Differential Mobility Particle Sizer (DMPS) and Aerodynamic Particle Sizer (APS) were used to measure the mass size distribution continuously. The chemical composition of fine particles (particle diameter %X 1 um) was very similar over the whole measurement campaign with 40% of mass composed of ammonium sulfate, 35% of OC and 5% of EC. In the submicron range the chemical mass closure of the collected samples was reached within a few percent on average. The chemical mass to gravimetric mass ratio was 0.98 ± 0.10 and 1.05 ± 0.13 (average ± S.D.) for the VI and SDI, respectively. Also, quite a good agreement was obtained between the mass size distributions measured with the ELPI and that measured with the DMPS-APS combination. When the total mass concentration of the fine particles was calculated, the mass concentration of the ELPI was found to be larger than that of the SDI and VI (ELPI/VI ratio 1.11 ± 0.13). This may be due to the semivolatile components lost in impactors. For the SDI and DMPS-APS the concentration of the fine particles was smaller than that of the VI with the SDI/VI and DMPS-APS/VI ratios of 0.70 ± 0.11 and 0.92 ± 0.08, respectively. For the DMPS and APS the mass concentration was calculated from the number concentration by estimating the particle density. The particle density was assessed in two ways; from the chemical composition of the particles (composite density) and by comparing the mass obtained from the DMPS-APS combination with the VI mass concentration (gravimetric ...
|
|---|