What is it about?
Unsupervised machine-learning can be used find patterns in mineral compositions from many eruptions of a single volcano. These patterns, when compared to other characteristics e.g. eruption explosivity, form links between chemical markers and eruptive behaviour. We then relate these chemical markers to physical conditions using thermodynamics modelling.
Featured Image
Photo by Peter Burdon on Unsplash
Why is it important?
We outline a method that correctly handles the mathematical properties of compositional data and leverages machine learning to find similarities in many mineral compositions. This is used to compare multiple eruptions from a single volcano and indentify likely causes for Villarrica's largest eruptions.
Read the Original
This page is a summary of: Insights Into Magma Storage Beneath a Frequently Erupting Arc Volcano (Villarrica, Chile) From Unsupervised Machine Learning Analysis of Mineral Compositions, Geochemistry Geophysics Geosystems, April 2022, American Geophysical Union (AGU),
DOI: 10.1029/2022gc010333.
You can read the full text:
Contributors
The following have contributed to this page
