HR: 13:55h
AN: V32F-02 [PDF]
TI: Melting due to Buoyant Migration of Water in the Hot Mantle Wedge Above a Subducting Plate
AU: Rilling, J L
EM: Jennifer_Rilling@brown.edu
AF: Brown University, Department of Geological Sciences
Box 1846, Providence, RI 02912 United States
AU: * Cagnioncle, A M
EM: amandine@brown.edu
AF: Brown University, Department of Geological Sciences
Box 1846, Providence, RI 02912 United States
AU: Elkins-Tanton, L T
EM: Linda_Elkins_Tanton@brown.edu
AF: Brown University, Department of Geological Sciences
Box 1846, Providence, RI 02912 United States
AU: Parmentier, E M
EM: EM_Parmentier@brown.edu
AF: Brown University, Department of Geological Sciences
Box 1846, Providence, RI 02912 United States
AB:
Water-bearing magma generated at convergent plate boundaries is thought to be due to the release of water carried to depth by
the subducting crust and mantle. The objective of our study is to create models of the buoyant upward migration of
water-rich fluids from the slab, through the hot mantle wedge, and consequent mantle melting.
Fluid in the models is assumed to migrate buoyantly along mineral grain edges with a prescribed melt fraction-grain
size-permeability relationship (e.g. Wark et al., 2003). Pressure gradients in the solid mantle flow should be important to
melt migration only if the mantle viscosity exceeds about 10$^{18}$ Pa-s. The model results thus far neglect the effect of
solid deformation on melt permeability, including the possible effects of the dependence of grain size on stress and the
anisotropic permeability created by solid deformation (e.g. Kohlstedt and Zimmerman, 1996). The volume of melt generated by
the interaction of water with the mantle is parameterized using results from MELTS (Ghiorso and Sack, 1995). The volume of
water released from the slab between depths of 80 to 150 km is derived from the estimates of Schmidt and Poli (1998).
The models consider a range of slab velocities and grain sizes using solid flow and temperature distributions from an earlier
model with temperature-dependent viscosity (Kelemen et al., 2003). Melt distribution in the wedge is strongly dependent on
both grain size and slab velocity. For a grain size of 2 mm at slab velocities of 2-6 cm/yr, fluid rises from the slab to
generate melt. At higher plate velocities or smaller grain sizes, water released from the slab is carried downward into the
deeper mantle, in which case melting is not triggered. As a consequence, melt flux at the top of the mantle wedge also varies
significantly with convergence rate. At slow to intermediate rates, calculated melt fluxes are comparable to values
observed in island arcs.
Thus, in the absence of other effects, significant differences in volcanic flux between fast and slow convergence rates would
be expected. In appropriate nondimensional form, the ratio of slab velocity to the square of the grain size controls the
distribution of fluids in the wedge. In nature, the absence of such a strong convergence rate dependence on magmatic flux
might indicate that grain size in the mantle wedge varies with plate velocity. Thus, future work should consider the effect
of variable grain size on melt permeability. Accounting for hydrous mineral stability limits may also introduce a
convergence rate and plate age dependence to predicted volcanic flux.
DE: 3210 Modeling
DE: 3699 General or miscellaneous
DE: 5114 Permeability and porosity
DE: 8120 Dynamics of lithosphere and mantle--general
DE: 8434 Magma migration
SC: Volcanology, Geochemistry, Petrology [V]
MN: 2003 Fall Meeting