The effective and efficient off-site management of a nuclear accident can be aided by the exploitation of suitable decision support systems, such as RODOS. The RODOS system is an operational tool that offers comprehensive decision support throughout all phases of a nuclear accident covering all distance ranges. The modelling system can provide description of the transport and dispersion of radionuclides in the atmosphere, aquatic systems and food chain. For this aim, predictions are made on suitable parameters including air concentration, deposition, concentration in foods, external dose rates and concentrations in water systems. However, since the model is based on general assumptions that are often chosen for larger areas, the model's predictions may not appropriately reflect local or regional conditions after an accident. A solution to this would be to update the model with data coming from available real monitoring observations. In an effort to improve the model's predictive capabilities, the DAONEM project developed and realised a methodology for exploiting these data. Employing the so-called data assimilation techniques, uncertainties related to model-based dose and consequence assessments are significantly reduced, resulting into more accurate predictions. The developed data assimilation tools are grounded on the specialised Bayes theorem and Kalman filtering and distinguished according to the different radioactivity pathways in the biosphere. The RODOS system was enhanced through full data assimilation capabilities concerning the atmospheric phase of a nuclear accident, the food chain and dose assessment module, and the hydrological model chain. The common approach used for all various pathways offers advantages of harmonisation and coherence in the propagation of uncertainties between model chains. It is anticipated that the optimised RODOS system will gain a wider acceptance in Europe and fulfil expectations of radiological advisors for accurate, timely and efficient emergency management.