Affordable Access

deepdyve-link
Publisher Website

Resolution of ionic components of late winter arctic aerosols

Authors
  • Shao-Meng, Li
  • Winchester, John W.
Type
Published Article
Journal
Atmospheric Environment (1967)
Publisher
Elsevier
Publication Date
Jan 01, 1989
Volume
23
Issue
11
Pages
2387–2399
Identifiers
DOI: 10.1016/0004-6981(89)90251-5
Source
Elsevier
Keywords
License
Unknown

Abstract

A total of 69 size fractionated aerosol samples was taken day and night, 16 March–6 May 1986, in Barrow, Alaska, during the AGASP-II project in order to identify components of the Arctic haze and their likely sources. Samples were taken sequentially, 12 ionic species in the samples were measured, and multivariate principal component analysis was used to resolve different components. The samples were analyzed by ion chromatography for SO 2− 4, Cl −, Br −, NO − 3, methanesulfonate, formate, acetate, propionate, pyruvate, K +, Na + and NH + 4. The temporal variabilities in measured concentrations were fit to a linear physical mixing model by principal component analysis of the data. Coarse and fine aerosol components exhibited both similarities and differences. Most of the coarse concentration variability is described by three components that explain 80% of total variance in the data set: sea salt contaminated with non-sea salt (nss) SO 2− 4 and nss K +, a carboxylic acid anion component and a component containing most of Br − 3, NO 3 − and methanesulfonate. Ion balances between cations and anions in all three components are close to neutrality, consistent with these components being internally mixed aerosols. The large amounts of nss K + in the first coarse component suggest a combustion origin at lower latitudes. The second component consists of carboxylic acid anions and some sea salt, and may represent reaction between the acids and sea salt. The third component of Br −, NO − 3 and methanesulfonate could be the product of gas phase reactions of marine organic compounds. Some independent aerosol occurrences, consisting of seven coarse samples of highest acetate and sea salt concentrations, is identified by the combination of linear regression and absolute principal component analysis (APCA) and does not belong to the communality of the rest of the data. In the fine fraction four components explain 86% of total variance in the data set, with compositions that differ somewhat from their coarse fraction counterparts. The first component is mainly a mixture of carboxylic acid anions with additional NH + 4 and Na +, which resembles the second coarse component in composition. The second is a contaminated sea salt component similar in composition to the contaminated sea salt in the coarse fraction. However, it contains smaller amounts of nss SO 2− 4 and is not as deficient in Cl −. The third component is defined and contains most of the Br −, NO − 3 and methanesulfonate, and has a similar composition to the third coarse component. The fourth component is mainly nss SO 2− 4 plus significant amounts of NO − 3 and NH + 4.

Report this publication

Statistics

Seen <100 times