One of the most puzzling aspects of the magma/water mixing process is its unpredictability. In some cases, the mixing of magma and water results in passive, low-level release of steam. In other cases it produces devastating explosions. The range of consequences is presumably a result of different conditions under which mixing occurs, but those conditions are difficult to identify. In this study, clast vesicularity is used to distinguish between two fundamentally different types of magma/water mixing conditions: those involving static or slowly rising magma that occupies a vent during eruptive pauses, and those involving magma that is rapidly rising during the eruption. The study is based on extensive investigations of volcanic tephra at Kilauea Volcano, Hawaii.
At Kilauea volcano, an early, lithic-poor phase of a large, basaltic phreatomagmatic eruption in 1790 (the Keanakakoi eruption, unit II, in the nomenclature of Decker and Christiansen[1]) produced tephra that was unusual among phreatomagmatic deposits for its high vesicularity. The vesicularity of more than 1800 clasts larger than 5 mm in diameter, analyzed using the methods of Houghton and Wilson[2], averages 73+12%. Vesicularities less than 40% were measured in less than 2% of the clasts. The absence of dense juvenile debris indicates that the lapilli do not just represent the inflated interiors of clasts whose margins had quenched and separated prior to vesiculation. The magma apparently reached a fairly high, uniform degree of vesicularity before it contacted water.
In magmatic eruptions at Kilauea, such high vesicularities are reached primarily during lava fountaining, when magma is ascending rapidly up the conduit[3]. Slow magma ascent, typical of effusive lava-flow eruptions, produces moderately vesiculated magmas (~40-75%, occasionally up to 85%[4]). Magma that has been exposed to the atmosphere for several minutes or more, as would occupy a vent during pauses in an eruption, contains 20-40% vesicularity or less[5].
The tephra is also distinctive for its abundance of very fine vesicles. The number of bubbles per cubic centimeter of melt (number density) ranges from about 10^5 to 10^7, with mean bubble diameters .03-.07 mm (M. Mangan, USGS, HVO, written commun.). At Kilauea, vesicle number densities greater than 10^4 are characteristic only of high lava-fountain eruptions, where magma ascent and bubble nucleation are very rapid. Recent lava-fountain eruptions at Kilauea have produced vesicle number densities of 10^4-10^5 and mean vesicle diameters of .1-.2 mm[3]. Effusive lavas have vesicle number densities around 103 and mean vesicle diameters of 0.2-0.4 mm[4].
The high vesicularity and fine vesicle textures of the Keanakakoi tephra indicate that magma was rising rapidly up the conduit at the time it was quenched by groundwater. The hydrologic conditions that must be met in order for water to mix with rapidly ascending, erupting magma are much more restrictive than those that would allow water influx during eruptive pauses. During eruptive pauses (Fig. 1a), water may enter the vent whenever the magma level drops below the water table. The water table need not be shallow for influx to occur, as long as the magma level drops below it. During active eruptions (Fig. 1b), water would enter the conduit only if the water pressure, at a given depth, were higher than the magma pressure at that depth. Numerical modelling using magma and conduit conditions typical at Kilauea[6] has shown this would occur only if the water table were less than ~120 meters deep--400 meters shallower than its current depth.
In previous studies, phreatomagmatic deposits with uniformly well-vesiculated juvenile fragments and low lithic content have been recognized primarily in silicic phreatoplinian ashes (e.g. the Hatepe Ash of Self and Sparks[7] and Houghton and Wilson [2]). Like the Keanakakoi ash, phreatoplinian ashes also result from mixing of water with rapidly ascending magma. Most phreatoplinian ashes have erupted through lakes or in shallow marine environments[7]. In view of the high vesicularity of these deposits, one would expect unusually large energy release when they mix with water. This seems to be supported by the fact that deposits of individual phreatoplinian eruptions are unusually widespread[7].
More typical basaltic phreatomagmatic deposits (the Surtseyan deposits of Self and Sparks[7]) contain clasts of significantly lower (though variable) vesicularity than those of the Keanakakoi[2, 8]. If the vesicularity of Kilauean eruptions can be used as a model, the lower vesicularity would imply that these magmas were sitting statically in the vent, or were rising slowly, at the time of mixing. This is consistent with observations of historic phreatomagmtic eruptions (e.g. at Surtsey[9]), which are typified by discrete explosions separated by eruptive pauses. These eruptions have involved both groundwater (e.g. at Ukinrek Maars[10]) and surface water (e.g. Surtsey). The low vesicularity of this magma means that other fragmentation processes must supplement vesiculation in order to maximize the efficiency of the eruption. If magma is not fragmented by other mechanisms, explosions will not take place, or they will be comparatively small.
Return to:
[Volcano-Groundwater Interaction Project] ...
[Mastin Publications Menu] ...
