|
PARENT SESSION
1A - Environmental analytical methods Poster Hall 8:30 AM - Monday, 28 April 2003 Chair: Schäffer, A.1, 1 Co-chair: Bester, K.2, 2
(MOP/20) Principal Components Analysis of Dioxins from Rocky Mountain Arsenal and Denver CO.
Jones, Paul1, Giesy, John1, Newsted, John2, Henningsen, Gerry3, Williams, Laura3, Bicher, Carol4, Slocomb, John2, 1 Michigan State University, East Lansing, MI, USA2 Entrix Inc., Houston, TX, USA3 US-EPA, Denver, CO, USA4 Gannett Fleming Inc., Denver, CO, USA
ABSTRACT- Multivariate statistical analysis, particularly Principal Components Analysis (PCA), is often conducted with the aim of attributing contamination to specific, high concentration sources. In this study we used PCA to define sources of dioxins at low/realistic environmental concentrations originating from diffuse sources. PCA was conducted on concentrations of 2,3,7,8-substituted dioxins and dioxin-like PCBs in 220 soils collected from the Rocky Mountain Arsenal (RMA) and the Denver Front Range region. PCA demonstrated the presence of possible minor sources of dioxins on RMA. Samples from RMA were identified by a congener profile containing higher chlorinated PCDFs. In contrast samples from the reference areas were characterized by the presence of higher chlorinated PCDD congeners as would be expected for samples influenced by a range of general human activities. The presence of a PCDF signature for RMA samples does not necessarily indicate major on-site contamination. Indeed, the relatively diffuse nature of the sample clusters would argue strongly against the presence of a single large source. Instead, the predominance of the PCDF congeners probably indicates the mixed industrial activities that took place on the site or atmospheric deposition or a combination of the two. Being able to attribute sources/amounts of site-related dioxins as opposed to ubiquitous trace-level contamination is important for risk assessment for potent toxicants such as dioxins.
Key words: principal components analysis, dioxins, Soil, source identification
|