In the aftermath of a large-scale nuclear emergency, remediation of affected farmland is key to return to normalcy. Due to the technical complexity of this endeavour and the socio-economic constraints, prioritization of farmland for remediation, selection and efficiency monitoring of remediation actions are essential. Therefore, sufficiently accurate, cost-effective and preferably simple approaches must be found. This can be alleviated with well-designed Information Technology (IT) response systems. The general objective of this PhD proposal is to strengthen the scientific basis for such IT-systems and in particular for multiple-criteria prioritization of affected farmland and of technical options for remediation. First, innovative predictive models of the fate of nuclear contamination will be set up through artificial intelligence techniques trained using spatially explicit data about historical and simulated accidents, agricultural practices, environmental and agro-economic conditions and state-of-the-art knowledge of radio-ecological behaviour. Next, the outcome of the predictive models will be fed into algorithms coming from operations research for testing and finetuning of the latter, eventually leading to novel robust decision-making procedures for optimizing remediation efforts in response to large-scale nuclear emergencies affecting food and agriculture.
This work will be carried out in close collaboration with the SWMCN Laboratory of the Joint FAO/IAEA Division.