Applying process-based models to the Borssele scenario

Abstract

The objective of this paper is to consider the implications of employing process-based models on predictions for radionuclide activity concentrations in grass and cow milk. The FDMT (Food Chain and Dose Module for Terrestrial Pathways as used in the JRODOS and ARGOS decision support systems) model has been transferred to a modelling platform enabling sub-models to be modified and replaced. Primarily, this has involved invoking process-based models for 137 Cs and 90 Sr that account for soil chemistry in simulating bioavailability and plant transfer. The implementation of such models can lead to quite dramatic differences in predicted activity concentrations of radionuclides in grass and milk compared to a default FDMT set-up for time periods later than a few weeks post deposition. Considering transfer within a spatial context, by combining information from the outputs of process-based models with illustrative soil maps, leads to the observation that the most elevated 137 Cs and 90 Sr concentrations in grass and milk might not necessarily occur in areas where deposition is highest. Not accounting for soil type when modelling food chain transfer might lead to the sub-optimal allocation of resources or misidentification of the most vulnerable areas in the long-term after an accidental release.