Quantitative identification of sources of dioxin-like polychlorinated biphenyls in sediments by a factor analysis model and a chemical mass balance model combined with Monte Carlo techniques.

The major sources of dioxin-like polychlorinated biphenyls (PCBs) in two sediment cores from Tokyo Bay and Lake Shinji (both in Japan) were identified and their source contributions estimated using two receptor models. The first was a nonnegative constrained factor analysis (FA) model, and the second was a nonnegative constrained chemical mass balance model combined with Monte Carlo techniques (CMB-MC) to take into account the variability and uncertainty in both PCB congener profiles of sources and environmental samples. According to the FA model, variations in the concentrations of dioxin-like PCBs in each sediment core were accounted for almost entirely by two factors, which were considered to correspond to Kanechlors (KCs; Japanese PCB products) and incineration. The CMB-MC model investigated the trends of the burdens from four types of KCs and incineration to the concentrations of dioxin-like PCBs in each sediment core. The results for both sediment cores obtained by both models indicated that the burden from KCs increased gradually beginning in the 1950s, peaked around 1970, and declined thereafter, whereas the burden from incineration increased gradually from the 1950s to the early 1990s. The estimated contribution from incineration to the toxic equivalent concentration of dioxin-like PCBs was comparable to that from KCs.