T. Koyaguchi et Aw. Woods, ON THE FORMATION OF ERUPTION COLUMNS FOLLOWING EXPLOSIVE MIXING OF MAGMA AND SURFACE-WATER, J GEO R-SOL, 101(B3), 1996, pp. 5561-5574
When magma vents into the sea or a crater lake, the ensuing magma-wate
r interaction can affect the style of eruption dramatically. If the ma
ss of surface water incorporated into the erupting material is small,
(< 15% of the total mass), then typically, this water vaporizes, and t
he density and temperature of the erupting mixture decreases. As a res
ult, the minimum eruption velocity for which a Plinian-style eruption
column may develop decreases. If a larger mass of cold surface water i
s added to the mixture, then part of this water may not vaporize, the
initial mixture has the saturation temperature, and the initial densit
y increases again. For sufficiently large masses of surface water mixe
d into the erupting magma, the ascending mixture cannot become buoyant
. Instead, relatively cold, wet and dense ash hows spread laterally fr
om a collapsing fountain. With a small or moderate eruption rate, < 10
(8) kg/s, the height of rise of a buoyant column does not vary signifi
cantly with the surface water content. However, for very large eruptio
n rates, > 10(8) kg/s, the height progressively decreases with surface
water content. This occurs when the magma and surface water begin to
constitute a significant fraction of the mass at the top of the column
, so that an increasing fraction of the initial magmatic thermal energ
y is converted to the surface water rather than the entrained air. The
transitions in eruption style which result from changes in the mass o
f surface water mixing with the magma may account for observations of
both buoyant plumes and wet surge during the eruptions of Taal in 1965
and Miyake-jima in 1983 and for the changes in the eruptive activity
at Surtsey in 1963-1964 as the access of seawater to the vent became m
ore restricted. We also present calculations which suggest that the ac
cretionary lapilli, which are often found in wet how deposits, may res
ult from condensation of vapor in both the cold, wet collapsing founta
ins and in the flows themselves.