0585405 20250526121919.020240422235959.9 10.5194/egusphere-egu24-3075 1 s. E cav_un_epca*0585420MunichEuropean Geosciences Union2024 radioxenon observations nuclear test International Monitoring Stations cav_un_auth*0267768 Tichý Ondřej UTIA-B Adaptivní systémy Department of Adaptive Systems AS AS Department of Adaptive Systems Ústav teorie informace a automatizace AV ČR, v. v. i. cav_un_auth*0101207 Šmídl Václav UTIA-B Adaptivní systémy Department of Adaptive Systems AS AS Department of Adaptive Systems Ústav teorie informace a automatizace AV ČR, v. v. i. cav_un_auth*0387255 Mácha Václav UTIA-B Adaptivní systémy Adaptive Systems AS AS CZ Ústav teorie informace a automatizace AV ČR, v. v. i. cav_un_auth*0466450 Kuśmierczyk-Michulec J. AT cav_un_auth*0466451 Sommerer W. AT cav_un_auth*0466452 Tipka A. AT http://library.utia.cas.cz/separaty/2024/AS/tichy-0585405.pdf GA24-10400S GA ČR CZ cav_un_auth*0464279 The identification of a sample associated with a nuclear test is a challenging task for the CTBTO because of the presence of a noble gas background in the constant evolving atmosphere. This background is caused by nuclear power plants, nuclear research reactors, and medical isotope production facilities and contributes to samples collected by the noble gas systems of the International Monitoring Stations (IMS). Because of that background, standard linear inverse model applied to Xe-133 measurements is prone to substantial errors. To address this problem, we investigate possible methods for separation of the background signal and any signal from a nuclear explosion, which is further processed for estimation of the Xe-133 source term. We assume that the observed unknown point release of Xe-133 can be modeled as a linear model y=Mx, where y is the vector of observations, M is source-receptor sensitivity (SRS) matrix, and x is the temporal profile of the unknown release from a nuclear explosion, i.e. its source term. Since the signal in the observation vector is most probably mixed with civilian emitters, we test methods for separation of the contributions from the unknown signal and the background. We compare various approaches, ranging from simple model calibration, to simulated background term and their combinations with anomaly detection. The results are demonstrated on the data from the 1st Nuclear Explosion Signal Screening Open Inter-Comparison Exercise 2021 where advantages and disadvantages of studied methods are discussed and results are evaluated with the use of ground truth information on temporal and spatial location of the Xe-133 source. cav_un_auth*0466498 EGU General Assembly 2024 20240414 20240419 Vienna / Online https://www.egu24.eu/home.html AT BD 10000 10100 10102 2025 6 2 R 4 4r 4 20250526121909.1 4 20250526121919.0 PR RVO:67985556 https://hdl.handle.net/11104/0353494 S EGU24-3075 2024 Soubory v repozitáři: 0585405.pdf SCOPUS PUBMED WOS cav_un_epca*0585420 EGU General Assembly 2024 Abstracts Munich European Geosciences Union 2024