20 Critical Minerals Potential in the Land of Enchantment: Past Production and Future Potential

Global demand for critical minerals is increasing. These non-fuel minerals are essential for clean energy and high-tech industries. New Mexico's future critical minerals potential is being re-evaluated through USGS Earth MRI surveys and DOE-funded efforts.

Tuesday, 19 May 2026
8:30 AM - 12:00 PM
Alvarado B

Expert-guided Machine Learning for Characterization and Extraction of Critical Minerals

Paper number: 20-2
Program: Technical Sessions
Time: 8:55 AM – 9:15 AM

Authors

Abstract

I would like to present a new machine learning tool called Critical Mineral Explorer (CMEX). A low-concentration critical minerals (CM) layer with a more favorable mineral association may be preferable to a highly concentrated layer with a more challenging mineral association because the mineral association directly impacts the processing methods required to extract the CMs effectively. Geologists and geochemists possess specialized knowledge about the preferred mineral associations for CMs that allows them to identify specific signatures within a dataset that correspond to these preferred mineral associations. By leveraging this knowledge, they can guide the exploration and assessment process, focusing on areas with the most promising mineralogy for extraction. To facilitate the identification of these preferred mineral associations, we have developed a specialized algorithm called sparse non-negative matrix factorization (sNMF). This algorithm takes two matrices as inputs: one matrix contains the actual data, such as geochemical or geophysical measurements, while the other is a sparse matrix that represents the desired mineral association signatures. The sparse matrix serves as a template or guide for the algorithm. It directs the algorithm to search for patterns and signatures within the data matrix that closely resemble the sparse matrix. By constraining the algorithm's search space, the sparse matrix helps to identify regions within the dataset that exhibit characteristics similar to the preferred mineral associations. Through this process, the algorithm identifies the expected signatures of the preferred mineral associations within the dataset. It also determines the corresponding CM concentrations associated with these signatures, providing the potential abundance of CMs in specific locations. The output of the algorithm includes the identified signatures of high concentrations CMs and corresponding locations. The locations enable geologists and mining professionals to prioritize areas for further exploration and assessment. By combining the knowledge of preferred mineral associations with advanced algorithmic techniques, this approach enhances the efficiency and effectiveness of CM exploration. sNMF has been encoded into CMEX and will be demonstrated on a New Mexico Dataset.

Geological Society of America Abstracts with Programs. Vol. 58, No. 4

doi: 10.1130/abs/2026RM-410534

© Copyright 2026 The Geological Society of America (GSA), all rights reserved. Permission is hereby granted to the author(s) of this abstract to reproduce and distribute it freely, for noncommercial purposes. Permission is hereby granted to any individual scientist to download a single copy of this electronic file and reproduce up to 20 paper copies for noncommercial purposes advancing science and education, including classroom use, providing all reproductions include the complete content shown here, including the author information. All other forms of reproduction and/or transmittal are prohibited without written permission from GSA Copyright Permissions.