RESUMEN
Since the Chernobyl Nuclear Power Plant accident, decision support systems (DSS) for supporting response of the decision makers in emergencies have been developed and refined. Data available from real accidents are used to validate these systems, thus demonstrating their real capabilities and finally to improve them. This article presents the findings of the simulation exercises using JRODOS DSS performed in Spain after the first days of the accident in the Fukushima Daiichi Nuclear Power Plant. The investigation was carried out in two phases. The first phase is considered the early phase of the accident when few details of the real emissions are known (operational modeling). The second phase demonstrates how real measurements could be used (reconstructive modeling) to improve model predictions. Only major releases to the atmosphere, occurring during the first two weeks, were taken into account. Validation of the model was performed by direct comparison of the modeled results with real measurements.