Masters student, Thomas Mcloughlin-Coleman, attended the AME Roundup conference in Vancouver January 22-25, 2024 where he took first place for the poster presentations. The AME Roundup is a conference on mineral exploration and development that centers on work on the west coast.
Thomas is working on determining pyrite provenance which can be difficult, especially when dealing with complex or cryptic textures, never mind detrital pyrite. In this study we demonstrate that by using a machine learning model, Random Forest, that has been trained on published trace element date from LA ICP MS spot analysis from 5 deposit types globally, we can determine the deposit type of a given pyrite correctly 90% of the time. In our case study, we were able to determine the provenance of the detrital pyrite in a paleoplacer deposit and infer an upstream source for the gold. 
Photo credit: Thomas Mcloughlin-Coleman
Poster credit: Thomas Mcloughlin-Coleman
Posted on: Thursday, February 1, 2024