A Big Data Architecture for Learning-Based Source-Term Estimation

After the detection of a radioactive substance of unknown origin in the atmosphere, the source location is estimated via inverse modelling. Depending on various factors, such as the spatial resolution desired, traditional inverse modelling can be computationally time-consuming and therefore its application can be problematic when timing is critical. In a complementary presentation (S. Andronopoulos et al., “Towards Inverse Source Term Estimation using Big Data Technologies”), we discuss a data- scientific approach to source term estimation, which allows us to perform the bulk of the processing prior to such an event taking place, therefore allowing for rapid estimation. This poster presents the big data platform that enabled the implementation of this work and the design and rationale behind our software architecture.