Skip to content

Hydrothermal mineralisation prior to gas-driven eruptions: mineral seal formation constrained using flow-through experiments

Geoff Kilgour 1, Lucjan Sajkowski1, Bruce Mountain1, Bruce Christenson1, Frankie Haywood2, Paul Jarvis1, Eric Sonnenthal3, Brian Tattitch4

  • Affiliations:  1GNS Science, New Zealand; 2School of Earth Sciences, University of Bristol, UK; 3Lawrence Berkeley National Lab, USA; 4College of Earth, Ocean, and Atmospheric Sciences, Oregon State University, USA 

  • Presentation type: Talk

  • Presentation time: Thursday 11:15 - 11:30, Room S160

  • Programme No: 3.14.4

  • Theme 3 > Session 14

Abstract

Gas-driven volcanic eruptions (phreatic eruptions) occur regularly, most often associated with andesite volcanoes with active hydrothermal systems, and are difficult to forecast. Forecasting their eruption onset is one of the key challenges in volcanology, with some exciting progress in automated analysis of unrest signals, coupled with greater understanding of subsurface processes. Hydrothermal minerals are known to form a blockage beneath a vent area, yet the kinetics, tensile strength, and efficiency of that blockage is unconstrained. Here we use a large temperature gradient (~25 to 200 °C) flow-through reactor filled with andesite granules and percolate crater lake fluid through the particles to simulate downwelling fluids being swept into fumarolic conduits. Our results show that within weeks, a mineral seal had formed, restricting fluid flow and significant pressurisation to failure. Hydrothermal minerals present in our experiments are dominantly alunite and anhydrite, with minor sulfur, and rare monazite. This mineral assemblage formed at ~150-180 °C and matches with hydrothermal phases present in ejecta from past gas-driven eruptions. Analysis of the reacted effluent fluid show clear trends during hydrothermal seal formation that we will analyse further to constrain reaction rates of the rock and fluid in the experiments. We expect to re-define geochemical changes occurring in the crater lake, along with the full suite of monitoring data to fundamentally improve gas-driven eruption forecasting.