Analysis of the Global Radioxenon Background with Atmospheric Transport Modelling for Nuclear Explosion Monitoring

Radioxenon plays an important role in the monitoring of nuclear explosions, particularly in the detection of nuclear tests. Due to its inert character as a noble gas it is a likely candidate to escape from the cavity of an underground nuclear explosions. In any case, once in the atmosphere it is subject to prevailing winds and can be transported over vast distances, while the dispersion in the atmosphere leads to a dilution of the concentration. Beta-gamma coincidence and high-resolution gamma techniques allow to detect concentration as small as 1 mBq=m3 and below. A worldwide monitoring system is currently under construction by the Comprehensive Nuclear-Test-Ban Treaty Organization and already over 80% operational. However, besides nuclear explosions also other sources of radioxenon exist. Nuclear ssion processes involving heavy nuclei produce a certain yield of radioxenon isotopes. The strongest atmospheric emitters are nuclear power plants and medical isotope production facilities. Atmospheric Transport Modelling (ATM) has the potential to support the analysis of radioxenon detections. It is a relatively new tool that has come available in recent years due to improvements in computation power and data storage capabilities. The Lagrangian particle dispersion model Flexpart can simulate the sensitivity between sources and receptors on a global scale. The importance of ATM in radioxenon monitoring of nuclear explosions has increased since the existence and strength of background sources are known. The role of ATM in monitoring of nuclear explosions against this radioxenon background are examined in this work. The results can be divided in three parts. First, the general radioxenon background was analysed. The background for all 39 IMS noble gas stations was simulated for the course of one year and the time series of each station follows a certain distribution. The simulated data are compared with experimental data for stations with available data. Regionally dominant emitters of radioxenon and the impact of the availability of known, time resolved source term - opposed to a yearly average - on the ATM prediction capability is validated. Also, using the Fukushima nuclear accident as an example it is demonstrated that it is possible to reconstruct a source term of known location. The problem of reconstruction is reduced to solving an overdetermined set of linear equations. Secondly, possible roles of ATM in the automatic categorisation are suggested. ATM can be used to train categorisation schemes that inherently do not apply ATM. On the other hand ATM itself can be incorporated in such a scheme. Third, the network coverage of the IMS noble gas component is de ned in order to quantify its capability to detect nuclear explosions worldwide. Di erent emission and background scenarios are used to determine the ratio of detected and undetected hypothetical events, and recommendations to improve this value are given