Abstract The geopolitical context advocates for monitoring nuclear activities within an international framework. In this context, the Comprehensive Nuclear‐Test‐Ban Treaty Organization has established measurement stations, known as the International Measurement System (IMS). This includes radioxenon air concentration measurements for atmospheric monitoring. However, civil infrastructures, mainly nuclear power plants and medical isotope production factories, continuously release radioxenon, making detection of nuclear tests more challenging. Emission and dispersion simulations can help quantify this atmospheric background. Here, the analysis of FLEXPART outputs allows to describe the distribution of xenon‐133 releases over the IMS, aiming to find relations between sources and stations. Variables quantifying source spread and overlap over IMS stations were defined. One third of the stations are affected by up to two sources. Conversely, each source reaches on average less than five stations. We built a graph representation of the IMS, modeling station similarities at the seasonal scale, using the aforementioned variables. We applied spectral clustering, and the 11 resulting regional clusters gather stations with comparable source distributions. They show a link between different configurations of source distribution and aggregated regional concentrations. Finally, by adapting linear models to different regional configurations observed on the network, an efficient source estimation was achieved with a mean squared error below , and only two stations with a maximum absolute error above . This work, in addition to quantifying the atmospheric background by analyzing the source distribution, provides a new representation of the IMS data and opportunities for developing graph‐based anomaly detection criteria.
Atiq et al. (Sat,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: