Use of Receptor Models to Evaluate Sources of PAHs in Sediments

Abstract

Receptor models are mathematical procedures for resolving one or more of these parameters in a mixed chemical system: (1) the number of sources, (2) their chemical characteristics, and (3) the relative contribution of each source in environmental samples. These models are being used increasingly to evaluate sources of polycyclic aromatic hydrocarbons (PAHs) in sediments. As with any mathematical model, understanding the underlying assumptions is critical in interpreting the output. Three assumptions that raise particular challenges when applying receptor models to evaluate multiple sources of pyrogenic PAHs are (1) identification of all important sources, (2) stability of source profiles, and (3) linear independence of each profile. Variability within source types, and similarities among the PAH profiles of different sources, create uncertainties that must be considered when evaluating the results of receptor models. Various procedures for evaluating uncertainties have been applied in the literature, but validation and standardization of such methods are often lacking. Using a case study, this article demonstrates how a more detailed evaluation of model output can produce conclusions that differ from those initially published. While not eliminating uncertainty, we recommend a multiple-lines-of-evidence approach that includes both mixing and unmixing receptor models, along with other environmental forensic techniques.