96
Views
31
CrossRef citations to date
0
Altmetric
Original Articles

Formation of particulate MSA: deductions from size distribution measurements in the Finnish Arctic

, , &
Pages 159-171 | Received 05 Jan 1996, Accepted 10 Oct 1996, Published online: 18 Jan 2017
 

Abstract

Formation of particulate MSA was studied by conducting a field campaign in the Finnish Arctic. Based on continuous mass size distributions extracted from impactor measurements, 4 MSA modes were identified: an accumulation mode centred between 0.3-0.5 μm of particle aerodynamic diameter, an Aitken mode below 0.1 μm, and 2 supermicron modes that peaked at 2-3 and 7-10 μm, respectively. The lower supermicron mode resulted primarily from the reaction of gaseous MSA with sea salt particles, and the upper mode probably from its reaction with soil-derived particles. From 70 to 90% of the MSA was found in the accumulation mode, where it was distributed very similarly to ammonium. The overall MSA to nss-SO42- ratio, R, ranged from 2 to 34%, with most of this variation resulting from different degrees of anthropogenic perturbation in the measured air masses. When comparing different-size particles, R was clearly the highest in the Aitken mode, suggesting that MSA and sulfate contribute with comparable magnitudes to nuclei condensational growth at high latitudes during the summer. The evident variation of R with particle size, together with potential enhanced MSA production in air influenced by pollution, demonstrate further that one needs to be extremely careful when using observed MSA to nss- SO42- ratios for estimating the contribution of biogenic sources to total particulate sulfate in different environments. Strong indications on the saturation of MSA over accumulation-mode particles were found. The saturation effect was hypothesized to explain, in part, the observed large partitioning of MSA in the supermicron mode in warm and very acidic aerosol systems. The distribution of submicron MSA, including potential saturation, was shown to be affected significantly by cloud processing.