InterRAS Source Term Inversion Dataset

This dataset serves as the supplementary raw research data for the paper "Source term inversion network based on improved Transformer". It was simulated and generated using the International Radiological Assessment System (InterRAS), aiming to provide data and computational support for the inversion of radioactive releases (source terms) and nuclear emergency responses during severe nuclear accidents. The dataset is provided in .csv format and comprises 18 feature dimensions, constructing a complex non-linear mapping relationship from "meteorological parameters and dose rates" to "nuclide release rates."Specifically, the data features consist of three main modules:Meteorological and Environmental Parameters: Includes Time (h), Release height, Atmospheric stability, Wind speed, Weather conditions, and Mixing layer height, which characterize the spatio-temporal dispersion conditions of the radioactive plume.Radiation Field Monitoring Data: Contains continuous gamma dose rate sequences at five different downwind spatial distance nodes (1km, 2km, 5km, 25km, and 50km).Source Term Target Labels: Comprises the initial release rates of seven typical airborne radionuclides (Kr-88, Te-132, I-131, Xe-133, Cs-137, Ba-140, and Ce-144).This dataset is suitable for training data-driven models in the interdisciplinary field of artificial intelligence and nuclear physics, analyzing the spatio-temporal evolution mechanisms of complex radiation fields, and evaluating the testing and generalization capabilities of source term inversion algorithms in nuclear emergencies.