Simulation of a Nuclear Cloud's Propagation Following a Nuclear Accident

Radioactive particles leaked from a point source or a reactor are generally carried around by the wind, they are mostly concentrated in the downwind direction of their point sources when the winds blow in a particular direction continuously. For example, the chances of people being exposed to radiation particles are expected to be greater on the downwind direction of a source of nuclear contamination compared to the opposite case. If this principle of winds carrying radiation particles is the greatest most influential factor that determines the characteristics of contaminant dispersion under most weather conditions, it can be concluded that weather and wind patterns can be used as a medium to predict radiation fallout. The risk of exposure thus could be approximated using wind properties and by considering other factors that might pose as variables in the simulation. However, if this assumption is true remains unclear as the dispersion patterns and weather and wind patterns are both quite complex, making it difficult to find significant correlation factors. In this study, various machine learning-based prediction algorithms are used to clarify the dispersion patterns using the available wind patterns.