Researchers led by Prof. LI Nuo from the Xinjiang Institute of Ecology and Geography of the Chinese Academy of Sciences have developed a method to simulate the concentrations of unmeasured geochemical elements in rock and stream sediment samples. 

Published in Ore Geology Reviews, their work uses machine learning to address the challenges posed by limited geochemical data.

Geochemical data play a crucial role in various scientific domains, serving multiple purposes such as basic geological research, mineral exploration, environmental assessments, and monitoring efforts. However, geochemical datasets are often limited by various factors, posing significant challenges for data analysis and application. The high cost of elemental analysis frequently constrains many geochemical projects to selectively examine only a small subset of elements, thus limiting the understanding of broader geochemical characteristics.

To address this issue, the researchers applied the Random Forest machine learning model, which enables the simulation of missing or unmeasured geochemical elements. This innovative approach uncovers the complex relationships between different elements in nature, providing a more comprehensive view of geochemical processes.

"Machine learning can enhance our capacity to extract valuable information from the extensive geochemical datasets already available," said ZHOU Shuguang, first author of the study.

This study provides a viable solution for overcoming gaps in geochemical data, offering valuable insights for fields such as geology, environmental science, and soil science.