T216 Obsidian from Magma to Artifact: Geological and Archaeological Perspectives

Tuesday, 20 October 2009
1:30 PM - 5:30 PM
E146

WHAT CAN 3D X-RAY TOMOGRAPHY OF A PUMICE CLAST'S BUBBLE NETWORK AND CHEMICAL COMPOSITION TELL US ABOUT VOLCANIC ERUPTION DYNAMICS?

Paper number: T216-3
Program: Technical Sessions
Time: 1:50 PM – 2:05 PM

Authors

Abstract

Volcanic ejecta provide a plethora of insights into volcano eruption dynamics. Physical characteristics, such as pumice texture, crystal/bubble content, and magma rheology, all contribute to eruption explosivity by modifying magma permeability. Explosive eruptions typically occur when there is a buildup of gas pressure in a volcano conduit due to high magma viscosity and low magma permeability. Conversely, effusive eruptions tend to occur when gas can easily leave the conduit magma due to high magma permeabilities and/or low viscosity. The ability for magma to degas depends on the efficiency at which percolating bubble networks are produced. However, percolating bubble networks tend to form only under certain conditions. For example, bubbles are less likely to coalesce if magma viscosity is high. However, two bubbles may coalesce more easily by reducing the bubble surface tension produced by elongation. Deformation of a bubble from it's preferred spherical shape occurs when an adequately strong shear stress is applied to the bubble, where the degree of deformation is determined by shear stress and viscosity in the surrounding magma.

We will explore the effects of magma viscosity, and therefore chemical composition, on bubble network formation and permeability by examining physical characteristics of volcanic pumice. We determine the chemical composition of the erupted magma using a JOEL electron microprobe. Using X-ray tomography, we obtain 3D images of pumice clasts, isolate percolating bubble clusters, identify anisotropy in the bubble network, and numerically determine the sample's permeability using lattice-Boltzmann simulations. Several other physical characteristics are also numerically determined, including the specific surface area, pore fraction, and tortuosity of the bubble network. Therefore, known eruption dynamics can be compared to estimations based on characterizations of bubble network behavior relative to magma chemistry and related viscosity.