A note on partition coefficient distributions

Much interest has been generated in assessing the possible risks of contamination for humans and marine life resulting from the dumping of nuclear waste in Arctic Seas by the former Soviet Union. Models are being used to predict the transport of radionuclides released from the dumping grounds. A key parameter in these models is the partition coefficient representing the uptake potential of marine sediments and seawater for radioactive contaminants. Partition coefficients are dependent upon the independent variables of sediment concentration and the sediment/water radionuclide distribution coefficients. Modelers must use estimated average values for the independent variables because data for these variables are lacking for the Arctic environment. In this note, we illustrate the differences between assuming that partition coefficients can be computed solely from the average values of the independent variables vs. exact probability distributions, and illustrate the technique for one of the radioactive contaminants,¹³⁷Cs. In general, errors in the sediment partition coefficient can be as high as 67% using average values; for¹³⁷Cs. the error is about 37%.