Testing and optimizing two factor-analysis techniques on aerosol at Narragansett, Rhode Island

Elemental data for aerosol at Narragansett, RI, USA, were used to compare t he source-identification power of positive matrix factorization (PMF), a ne w Variant of factor analysis, with that of conventional factor analysis (CF A) and to investigate how much each technique can be "tuned" for best resul ts. The techniques generally yielded similar results. Although both were de graded by weak elements and gave factors that always differed somewhat from known sources, they nonetheless provided substantial insight into sources of elements. PMF: was harder to use than CTA but resolved crustal and marin e components up to an order of magnitude better. Best results were generall y obtained when the data were log-transformed, when missing data were repla ced by means, and when various numbers of factors were tried and their resu lts carefully evaluated for physical reasonableness. But the most important consideration was found to be the choice of elements, which outweighed all differences between techniques. Therefore, to maximize the source-identifi cation power of factor analysis, the two most important steps appear to be selecting the optimum set of elements and selecting the basic technique, in that order. (C) 1999 Elsevier Science Ltd. All rights reserved.