Natural clustering of pyrite with implications for its formational environment

Abstract

The modern mineral classification system is largely based on unique combinations of idealized major element composition and crystal structure, and therefore emphasis on the diverse modes of origin for each mineral is generally lacking in this classification system. However, mineral modes of formation provide insights into Earth's co-evolving geosphere and biosphere, and also have the potential to illustrate otherwise obscure aspects of planetary evolution. Here we use a case study of pyrite to demonstrate the power of using machine learning to divide the single type of mineral into different natural clusters. A variety of deposit types of pyrite (e.g., iron oxide copper-gold, orogenic Au, porphyry Cu, sedimentary exhalative, volcanic-hosted massive sulfide deposits, and barren sedimentary pyrite) are used to evaluate the clustering process. An array of trace elements determined by LA-ICP-MS are treated as predictors. Different clustering algorithms (e.g., K-means, PAM, Hierarchical clustering, mclust and DBSCAN) are employed in this study and their dissimilarities in model outputs are highlighted. We also compare model-based natural clusters with the original depositional environment of pyrite, and discuss the implications of our natural clustering.