Variability of dose predictions for cesium-137 and radium-226 using the PRISM method

The uncertainty associated with dose predictions for cesium-137 and radium-226 in a specific ecosystem has been studied. The method used is a systematic method for determining the effect of parameter uncertainties on model prediction called PRISM. The ecosystems studied are different types of lakes where the following transport processes are included: runoff of water in the lake, irrigation, transport in soil, in groundwater and in sediment. The ecosystems are modelled by the compartment principle, using the BIOPATH-code. Seven different internal exposure pathways are included. The total dose commitment for both nuclides varies about two orders of magnitude. For cesium-137 the total dose and the uncertainty are dominated by the consumption of fish. The most important factor to the total uncertainty is the concentration factor water-fish. For radium-226 the largest contributions to the total dose are the exposure pathways, fish, milk and drinking-water. Half of the uncertainty lies in the milk dose. This uncertainty is dominated by the distribution factor for milk. (orig.)