Poseidon would be flattered

(A little diversion this week from Santiaguito topics; I don't want to feel like I'm beating the subject to death. Plus I wanted to try out this ResearchBlogging thing...)

Every week the volcanology folks here try to get together for a paper discussion of some sort, and this semester's theme is basaltic volcanism. Not exactly my area of expertise, but I decided to volunteer to lead the discussion last week with a paper that recently came out in Geology: "'Poseidic' explosive eruptions at Loihi Seamount, Hawaii" by C.I. Schipper, J.D.L. White, B.F. Houghton, M. Shimizu, and R.B. Stewart. 

First, a little background on Loihi. This seamount, which lies about 35 km off the southeast coast of the Big Island, is the youngest volcano in the Hawaiian chain, although it will be at least another 10,000 years before it actually reaches the surface and starts developing black sand beaches for tourists to toast themselves on. Loihi has a summit caldera with cones and pit craters, and its eruption style varies from explosive (the subject of this article) to effusive (pillow lavas, which are mentioned in the discussion about vesicularity). When I first came to grad school, I would have thought of a submarine eruption as passive extrusion of pillow lavas, similar to the videos you can see on NOAA's Ocean Explorer website. This isn't always the case, however. While it's been fairly obvious that volcanoes just reaching the ocean's surface can erupt in an explosive and spectacular fashion, it hasn't always been clear that underwater volcanoes are capable of the same thing.

Schipper et al. review the traditional thinking on the process of submarine explosive eruptions: something like a Strombolian bubble burst, where slugs of gas exsolve from the magma and pop at its interface with the atmosphere (or seawater, in this case), throwing out little blobs of spatter and ash. This is a hard interpretation to test, however, if you can't observe the volcano erupting, or get hold of the eruptive products. Schipper et al. set out to do just that at Loihi, and made some interesting observations & interpretations. Their main conclusion is that the particular eruptive products they were examining were produced by thermohydraulic explosions (a kind of magma-water interaction), and not explosive fragmentation (magma being blown apart by rapid gas expansion). 

To begin with, the lapilli (lava chunks smaller than 6.4 cm but larger than 2 mm) were quite vesicular - up to 40% - unlike the pillow lavas that make up the surface they were found on. The vesicles were also small, mostly spherical and not interconnected. Interconnected vesicles in a lava implies that there was a network through which gas could flow (sometimes called "open-system" degassing, where gas is moving independently of the melt); the lack of connectivity in the Loihi lapilli implies that there was "closed-system" degassing, where gases exsolve and form bubbles in the melt, but remain "coupled" to the melt and move with it. Schipper et al. suggested that the magma moved too quickly to allow coalescence, and that the bubbles formed in the conduit rather than at the magma-water interface. 

Next to examine the issue of fragmentation....One mechanism that is often invoked for breaking up an erupting melt is explosive fragmentation, where gas in the melt becomes overpressurized relative to the outside atmosphere (as in when a gassy magma reaches the top of a conduit) and explodes. Schipper et al. note that this wouldn't work at Loihi, however, because at the 7.3 MPa pressures found 700 meters below sea level, the CO2 in the vesicles would be a supercritical liquid, and not capable of expanding fast enough to fragment the melt. (In addition, they suggest that the low viscosity of the basaltic melt would also allow rapid bubble expansion without fragmentation.) So how did these particles get fragmented? The authors point to a kind of hydromagmatic process called thermohydraulic fragmentation, which has to do with the interaction of hot material and water. In this process, water coming into contact with hot melt flash-boils and expands enormously, and the force of the steam expansion deforms and fragments the melt. 

The authors suggest that thermohydraulic fragmentation provides enough initial energy that the overpressured-gas fragmentation can then take over above the level of the vent. This is an interesting idea; if it's true, how much of the fragmentation contribution comes from each process? The smallest particles that Schipper et al. examined  (see Figure 4a-b, Schipper et al. 2010, at left) showed signs of having been produced by thermohydraulic processes (fewer interconnected vesicles, fracture surfaces that weren't bounded by vesicle walls, etc.), but the larger lapilli seemed to have been broken up following expulsion from the vent, which implies that thermohydraulic processes may be limited to the region below the vent. Additionally, there's the question of how representative the samples were; the study retrieved them some years after the eruption took place, and there's no telling how much was carried off by currents or other reworking processes. (Plus it can't be easy to sample something that's 3,000 feet down, since you have to use a submersible to get there. I would have liked to see the authors these as limitations, however.)

The figure at right (Figure 3d, Schipper et al. 2010) provides a good summary of what the authors think is going on: quickly-moving melt exsolves gas and develops vesicles that are unable to coalesce, so they remain spherical and isolated in the melt. When the melt reaches a point somewhere below the vent, it interacts with seawater in thermohydraulic fragmentation (samples represented by white circles in the figure), which then provides the energy to help initiate explosive fragmentation (the dark gray circles, which represent lapilli and bombs in their samples that were not formed by hydromagmatic processes).

The discussion group pretty much all agreed that they liked this paper for both writing style and conclusions, but had some issues with introducing a new bit of terminology to describe explosive eruptions. I have to agree; the term 'Poseidic' does have a basis for distinguishing these eruptions from subaerial ones (based on being underwater, having coupled volatile exsolution, and a particular combination of textures in its tephra), but you could reasonably also call this a 'submarine violent Strombolian' or 'submarine subPlinian' eruption. Still, distinguishing the style as an end-member is a wise choice, given the range of behaviors that volcanoes can exhibit, and the name might just stick. (Loihian is a suggested alternative, although that would come with spelling and pronunciation issues AND would be inaccurate in a few thousand years.)

Schipper, C., White, J., Houghton, B., Shimizu, N., & Stewart, R. (2010). "Poseidic" explosive eruptions at Loihi Seamount, Hawaii Geology, 38 (4), 291-294 DOI: 10.1130/G30351.1