Machine learning could revolutionize mining exploration

Machine learning could revolutionize mining exploration

The Morenci mine in Arizona is one of the world’s largest suppliers of copper and other sought-after minerals. As the demand for rare earth elements and metals increases to power global technology, new techniques are needed to find the next big porphyry copper deposits. Credit: Stephanie Salisbury/Wikimedia, CC BY 2.0

21st century technologies, including those essential to a low-carbon future, rely on rare earths and metals. Many of these sought-after minerals reside in porphyry copper deposits that contain hundreds of millions of metric tons of ore. In addition to copper, these deposits are a source of significant amounts of gold, molybdenum and rhenium. However, the mining industry has identified and mined most of the world’s large accessible porphyry deposits. Despite increasing investment in mineral exploration, the rate of discovery of mineral deposits is declining.

In a study recently published in the Journal of Geophysical Research: Solid Earth, Zou et al. present two new machine learning techniques to identify novel, deeply buried porphyry copper deposits by characterizing magma fertility. Fertile magma refers to magmas that can form porphyry deposits. Barren magmas, on the other hand, are not likely to develop rich ores. The authors aimed to improve traditional geochemical indicators plagued by high rates of false positives.

The authors developed two algorithms, random forest and deep neural network. They formulated the models using a global zircon chemistry dataset, which is used to assess porphyry copper deposits in magma. The authors focused the models on 15 trace elements. They validated the models with independent datasets from two well-characterized porphyry copper deposits in south-central British Columbia, Canada, and Tibet, China.

Both models gave a classification accuracy of 90% or better. The random forest model had a false positive rate of 10%, while the deep neural network model had a false positive rate of 15%. By comparison, traditional metrics report false positives at a rate of 23% to 66%.

Europium, yttrium, neodymium, cerium and other elements emerged as significant indicators of magma fertility. Model performance indicates that the algorithms can distinguish between fertile and barren magmas using trace element ratios. Notably, model performance was not affected by regional differences or geological context between the Canadian and Chinese assessment datasets.

As the demand for rare earth elements, minerals and metals increases, new techniques are needed to discover previously unknown deposits. According to the researchers, the results highlight the promise of machine learning as a robust, accurate and efficient approach to identifying and locating porphyry copper resources.

Study reveals petrogenesis of porphyry copper deposits in southern Tibet

More information:
Shaohao Zou et al, Application of machine learning to the characterization of magma fertility in porphyry copper deposits, Journal of Geophysical Research: Solid Earth (2022). DOI: 10.1029/2022JB024584

Provided by American Geophysical Union

This story is republished courtesy of Eos, hosted by the American Geophysical Union. Read the original story here.

Quote: Machine Learning Could Revolutionize Mineral Exploration (2022, August 26) Retrieved August 27, 2022 from

This document is subject to copyright. Except for fair use for purposes of private study or research, no part may be reproduced without written permission. The content is provided for information only.

#Machine #learning #revolutionize #mining #exploration

Leave a Comment