Applying multivariate statistics to discriminate uranium ore concentrate geolocations using (radio)chemical data in support of nuclear forensic investigations. (October 2016)

• Record Type: Journal Article
• Authors: Reading, David G.; Croudace, Ian W.; Warwick, Phillip E.; Cigliana, Kassie A.
• Is Part Of: Journal of environmental radioactivity. Volume 162/163(2016:Oct.)
• Page Start: 172
• Page End: 181
• Publication Date: 2016-10
• Subjects: Radioactivity -- Periodicals; Radiation, Background -- Periodicals; Radioecology -- Periodicals; Radioactive pollution -- Periodicals; Environmental Pollutants -- Periodicals; Radioactive Pollutants -- Periodicals; Radioactivity -- Periodicals; Radioécologie -- Périodiques; Pollution radioactive -- Périodiques; Fond de rayonnement -- Périodiques; 539.752
• DOI: 10.1016/j.jenvrad.2016.05.013
• Languages: English
• ISSNs: 0265-931X

Abstract

Abstract: The application of Principal Components Analysis (PCA) to U and Th series gamma spectrometry data provides a discriminatory tool to help determine the provenance of illicitly recovered uranium ore concentrates (UOCs). The PCA is applied to a database of radiometric signatures from 19 historic UOCs from Australia, Canada, and the USA representing many uranium geological deposits. In this study a key process to obtain accurate radiometric data (gamma and alpha) is to digest the U-ores and UOCs using a lithium tetraborate fusion. Six UOCs from the same sample set were analysed 'blind' and compared against the database to identify their geolocation. These UOCs were all accurately linked to their correct geolocations which can aid the forensic laboratory in determining which further analytical techniques should be used to improve the confidence of the particular location. Highlights: Principal Components Analysis is used as geolocating tool for nuclear forensics. High quality radiometric signatures for uranium ore concentrates obtained using a rapid and effective dissolution technique. "Unknown" samples are statistically compared against a database of known signatures.