Conduit processes during the February 11, 2010 Vulcanian eruption of Soufrière Hills, Montserrat

Abstract

We analyzed pumice from the February 11, 2010 Vulcanian explosion that immediately followed a large dome collapse at Soufrière Hills volcano. We obtained pre-explosive values of porosity, pressure, and depth by combining textural analyses and glass water content determinations. Our data suggest that the February 2010 explosion evacuated the upper 3 km of the conduit from the dense magma (≤10 vol.% porosity) it contained. The low porosity distribution in the volcanic conduit implies that the magma rising from the reservoir had time to extensively degas during emplacement. We use a conduit flow model to characterize the effects of permeability on ascent conditions. Model input parameters were fitted so as to match our pre-explosive porosity data, which yielded first-order constraints on conduit radius, mass flux, outgassing efficiency, and permeability. This parametric study suggests that efficient lateral gas escape is necessary to explain the low pre-explosive porosities. Steady-state solutions fitting the observed range of dome extrusion rate in the month preceding the February 11 event suggest permeabilities <10-13 m 2 deeper than 500 m, which are values typical of crack-supported permeability. Conversely, solutions with parameters consistent with bubble-supported permeability imply transient flow conditions prior to the February 11 event. The transient conditions imply that our data represent a snapshot of the porosity distribution within the conduit that does not preclude the temporary presence of much higher porosities in the conduit. Such unsteady conduit flow conditions are consistent with the irregular but active dome growth in the month prior to the February 11 event.

The role of oxides in the shallow vesiculation of ascending magmas

“…New and re-analyzed samples are listed in Table S1 in the Supplementary Information. Published data include products emitted by Vulcanian events on February 11, 2010 (23 samples, Burgisser et al, 2019) and during August-October 1997(8 samples, Giachetti et al, 2010 at Soufrière Hills, and during the 1980 blast at Mount St. Helens (Cashman, 1988).…”

“…All our analyses involve 4 magnification J o u r n a l P r e -p r o o f levels, except those of Merapi products and those of the February 2010 Soufrière Hills explosive products that only involve two magnification levels. It has been shown in Drignon et al (2016) and Burgisser et al (2019) that cumulated number densities in these samples are unaffected by the artifacts introduced by this information gap because the smallest and most numerous objects are captured by the highest SEM magnification.…”

“…Two of them are composed of deformed vesicles of equivalent size >300 μm across and circularity <0.2. Type L1 comprises large angular voids between crystal fragments that are interpreted to have been formed in response to the decompression accompanying the Vulcanian explosion (Giachetti et al, 2010;Burgisser et al, 2019). Type L2 is composed of the remaining vesicles of >300 μm, which are considered as pre-explosive vesicles.…”

“…Bubble nucleation, however, is sensitive to the number density of nucleation sites rather than to bulk volume or mass fraction of a given mineral. There are, to our knowledge, natural data on oxide number densities (OND) at explosive volcanoes for only one sample from the 1980 blast at Mount St. Helens, USA (Cashman, 1988), a few samples of Vulcanian explosions at Soufrière Hills, Montserrat Burgisser et al, 2019), Shinmoedake, Japan (Mujin and Nakamura, 2014), and at Kilian volcano, France (Colombier et al, 2017). Extending the scope from explosive eruptions to the reservoir-related sample from the Bishop Tuff (Gualda and Anderson, 2007) and to the fully degassed Tokachi-Ishizawa obsidian lava (Sano et al, 2015) does not vastly extend the amount of available data.…”

“…Events from Soufrière Hills, Kilian, and Merapi volcanoes have been included in this study also because the pre-explosive conditions of these sample series (i.e. the pressure and the porosity of each sample in the conduit prior to explosion) have been determined by prior studies Drignon et al, 2016;Colombier et al, 2017;Burgisser et al, 2019;Colombier et al, submitted).…”