Data Assimilation For Assessing Radioactive Contamination In Feed And Foodstuffs
Free (open access)
K Richter, F Gering & H Müller
After a large-scale emission of radioactivity into the atmosphere in the case of a nuclear accident appropriate measures have to be taken in order to minimize the health risk of the population. Estimations of the contamination of feed and foodstuffs are of particular importance for this. The decision support system RODOS, which is installed in many European countries for assessing the consequences of nuclear emergencies, contains a food chain model to predict the activity concentration in feed and foodstuffs. The uncertain y of the food chain model parameters, for example root uptake factors and the time of harvest, as well as the uncertainty of the input parameters, namely the deposited activity on ground and plants contribute to the uncertainty of the model output. Results of a sensitivity analysis to identify the most important model parameters are presented. Since in a real emergency, measurements of activity concentrations in feed and foodstuffs are available, they can be used to improve the results of the food chain model. This data assimilation is done with a Kalman filter, using the results of the sensitivity analysis. The Kalman filter updates the predicted activity concentration every time new measurements are available taking into account both the uncertainty of the food chain model output and the uncertainty of the measurements. Measurements carried out at a particular site and point in time may influence the predicted activity concentration at other sites and other points in time. Since the food chain model works on a large-scale grid a suboptimal Kalman filter, the Ensemble Kalman Filter, is applied in order to reduce the data amount and the computational time. Examples of the temporal course and the spatial distribution of activity concentrations are shown.