Analysis of geochemical basalts database as a simulant for uranium ore concentrates data: the Galaxy Serpent 3 Exercise

Itzhak Sedgi, Maor Assulin, Amiram Moyal, Boaz Astarchan, Arnon Rubinshtein, Eyal Elish

Research output: Contribution to journalArticlepeer-review

2 Scopus citations

Abstract

The Nuclear Forensics International Technical Working Group (NF-ITWG) conducted an exercise, Galaxy Serpent 3 (GS3), by the National Nuclear Forensics Libraries Task Group, during 2017. The main goal of the GS3 exercise was to illustrate the usefulness of a National Nuclear Forensic Library for assessing the potential consistency of unknown material with national holdings. Moreover, it offers the possibility to produce a realistic database resembling uranium ore concentrate (UOC) that can be used for algorithm development activities. A database was built based on basalt compositions from different tectonic settings: MORB, IAB, OIB and CRFB. Classes (Affinities) with subtly different chemical compositions. All data are open source and sample names can be traced to original data for error checking. Participants (29 teams) had to use existing techniques, or develop their own algorithms, to allow for dealing with the sparse database and managing incomplete data points. The exercise’s scenario introduced 3 simulated origin-unknown UOC samples, interdicted by authorities. Each team was asked to answer two main questions—“Are the 3 unknown samples consistent with each other?” and “are the 3 unknown samples consistent with the Mines in the database given?”. For the first time, the Israeli National Nuclear Forensic Laboratory (INNFL) participated in the GS exercise. A dedicated team was assigned the task of building capabilities needed and applying them to answer the raised questions. Methodology and tools developed and applied by the Israeli team are described. The results are presented and discussed, including new signatures allowing differentiation between unknown samples and mines.

Original languageEnglish
Article number806
JournalSN Applied Sciences
Volume1
Issue number8
DOIs
StatePublished - 1 Aug 2019
Externally publishedYes

Keywords

  • National Nuclear Forensics Libraries
  • Nuclear forensics
  • Principle component analysis
  • Ratio correlation matrix

ASJC Scopus subject areas

  • General Engineering
  • General Environmental Science
  • General Materials Science
  • General Physics and Astronomy
  • General Chemical Engineering
  • General Earth and Planetary Sciences

Fingerprint

Dive into the research topics of 'Analysis of geochemical basalts database as a simulant for uranium ore concentrates data: the Galaxy Serpent 3 Exercise'. Together they form a unique fingerprint.

Cite this