120-15 Beyond Pearson: Nonparametric Correlation Coefficients Quantify New Elemental Associations in a Porphyry Copper Geochemical Database

Abstract:

Porphyry copper deposits (PCDs) and their associated mineralized environments, such as skarns and epithermal deposits, are the world’s largest sources of copper and molybdenum, major sources of gold and silver, and large contributors of critical elements such as tellurium and selenium. These high-tonnage, multi-commodity deposits make PCDs an attractive exploration target offering opportunities for coproduct production, byproduct recovery, and waste valorization of critical elements. To increase exploration success, mining companies rely on porphyry indicator minerals (PIMs), porphyry vectoring and fertility tools (PVFTs), and pathfinder elements. However, the correlation coefficients that underpin such tools have introduced a bias that has gone largely unexamined. This study represents the first application of Chatterjee’s correlation coefficient in tandem with Spearman's correlation coefficient to a large-scale porphyry-based geochemical data set. Both Chatterjee and Spearman correlation coefficients are nonparametric, rank-based, and characterize broader classes of associations without requiring distributional assumptions. Applied to a geochemical database, this approach reveals elemental associations that differ meaningfully from Pearson’s correlation coefficient, which is commonly used for geochemical vectoring and exploration targeting. Few elements in this dataset meet the assumptions required for classical Pearson-based inference. Where Spearman and Chatterjee’s coefficients diverge from Pearson’s, they frequently identify stronger or additional copper associations, including common byproduct critical minerals rhenium and selenium, which are recovered during porphyry copper processing methods. These results identify candidate pathfinder elements alongside the established suite and indicate that correlation coefficient choice carries direct consequences for both critical mineral-focused and copper-focused exploration targeting. We advocate broader implementation of Chatterjee’s and Spearman’s correlation coefficients, alongside rigorous evaluation of distributional assumptions, as standard practice in geochemical analysis.

Geological Society of America Abstracts with Programs. Vol. 58, No. 5, 2026

doi: 10.1130/abs/2026AM-21527

© Copyright 2026 The Geological Society of America (GSA), all rights reserved.