<p>Polychlorinated dibenzo-p-dioxins (PCDD) and polychlorinated dibenzo-furans (PCDFs) belong to the most toxic compounds known to science and they are defined as Persistent Organic Pollutants (POPs) under the Stock-holm Convention. The general human exposure to PCDD/Fs is primarily through dietary intake. The importance of contaminated sites as secondary PCDD/F sources (i.e. sources that once received its contamination from a primary source) are getting increased attention. To be able to assess the risks for human exposure at PCDD/F contaminated sites, the environmen-tal distribution of PCDD/Fs and the potential mobilization between dif-ferent environmental media (e.g. food chain transfer) must be known. The primary aim of the work presented in this thesis was to investigate human exposure pathways associated with PCDD/F contaminated sites by combining field measurements and modeling. Site specific field measurements were made at a PCDD/F contaminated site in Sweden and multimedia modeling scenarios were evaluated against site specific data and national reference data.</p><p>The results show that the congener distributions in exposure media affected by contaminated soil may differ significantly from the distribu-tions found in media from reference locations. Higher chlorinated PCDD/Fs may be transferred into food chains where they contribute to a large fraction of the toxic equivalent concentration (TEQ). Ingestion of locally produced animal food may be an important exposure pathway al-ready at low or moderate PCDD/F soil concentrations. However, the con-gener composition of the source is critical for the exposure. The signifi-cance of the individual exposure routes varies depending on e.g. the spatial distribution and magnitude of the soil contamination, the pro-perties of the exposure media and the human behavior. Multimedia mo-deling can be used in risk assessments as long as model algorithms and model parameters are representative for the superhydrophobic properties of PCDD/Fs. However, selection of physical-chemical PCDD/F properties is a challenge due to large variation in reported values. For some proper-ties, data is scattered or completely lacking</p>

Polychlorinated dibenzo-p-dioxins (PCDD) and polychlorinated dibenzo-furans (PCDFs) belong to the most toxic compounds known to science and they are defined as Persistent Organic Pollutants (POPs) under the Stock-holm Convention. The general human exposure to PCDD/Fs is primarily through dietary intake. The importance of contaminated sites as secondary PCDD/F sources (i.e. sources that once received its contamination from a primary source) are getting increased attention. To be able to assess the risks for human exposure at PCDD/F contaminated sites, the environmen-tal distribution of PCDD/Fs and the potential mobilization between dif-ferent environmental media (e.g. food chain transfer) must be known. The primary aim of the work presented in this thesis was to investigate human exposure pathways associated with PCDD/F contaminated sites by combining field measurements and modeling. Site specific field measurements were made at a PCDD/F contaminated site in Sweden and multimedia modeling scenarios were evaluated against site specific data and national reference data. The results show that the congener distributions in exposure media affected by contaminated soil may differ significantly from the distribu-tions found in media from reference locations. Higher chlorinated PCDD/Fs may be transferred into food chains where they contribute to a large fraction of the toxic equivalent concentration (TEQ). Ingestion of locally produced animal food may be an important exposure pathway al-ready at low or moderate PCDD/F soil concentrations. However, the con-gener composition of the source is critical for the exposure. The signifi-cance of the individual exposure routes varies depending on e.g. the spatial distribution and magnitude of the soil contamination, the pro-perties of the exposure media and the human behavior. Multimedia mo-deling can be used in risk assessments as long as model algorithms and model parameters are representative for the superhydrophobic properties of PCDD/Fs. However, selection of physical-chemical PCDD/F properties is a challenge due to large variation in reported values. For some proper-ties, data is scattered or completely lacking