HR: 16:45h
AN: V22H-04    [PDF]
TI: On the significance of composition gradients across dunite-harzburgite interfaces in ophiolites.
AU: * Morgan, Z T
EM: Zachary_T_Morgan@brown.edu
AF: Brown University, Dept. of Geo. Sci., Providence, RI 02912
AU: Liang, Y
EM: Yan_Liang@brown.edu
AF: Brown University, Dept. of Geo. Sci., Providence, RI 02912
AU: Kelemen, P B
EM: peterk@whoi.edu
AF: WHOI, 360 Woods Hole Rd., Woods Hole, MA 02543
AU: Parmentier, E M
EM: EM_Parmentier@brown.edu
AF: Brown University, Dept. of Geo. Sci., Providence, RI 02912
AB: It has been suggested that dunite dikes or veins found in harzburgite hosts in the mantle sections of ophiolites are high porosity channels through which basaltic magmas were extracted from their source regions. The formation of such channels may involve pervasive melt flow and reactive dissolution that should leave geochemical fingerprints in the dunite and the harzburgite. Indeed, variations in mineral compositions across dunite-harzburgite contacts have been documented in several ophiolites. This study focuses on the distances over which mineral compositions vary in the dunite and harzburgite, referred to as (composition) boundary layers. Field observations suggest that the boundary layers whether in harzburgite or dunite are typically less than one third of the dunite width. The data also suggest a correlation between boundary layer thickness and dunite width. In order to understand the origin and significance of the concentration gradients we carried out a numerical study that simulates melt flow, diffusion and crystal-melt exchange in a 2D periodically distributed porous dunite and harzburgite matrix. Melt velocities were calculated from a prescribed pressure distribution and permeability structure according to Darcy's Law. Transport properties used in the calculations are similar to those used by Morgan and Liang (2003). Development of composition boundary layers in the dunite and harzburgite depends on a number of factors including, (1) porosity, (2) melt flow rates in the dunite and harzburgite, (3) flow directions, (4) time, (5) crystal-melt exchange rate, (6) rate of dunite-harzburgite interface migration, (7) deformation, and (8) subsolidus reequilibration. Here, we focus on our simulations for the first three. Where steady-state melt flow is parallel to the dunite-harzburgite interface, the thickness of the boundary layers is proportional to the square root of the melt velocity. Since the velocity in the dunite channel is generally larger than in the harzburgite matrix, the boundary layer thickness in the harzburgite is predicted to be larger than in the dunite. When there is a significant component of melt flow across the dunite-harzburgite interface, the relative boundary layer thickness between the two rock units is different. For instance, the boundary layer thickness in the harzburgite becomes smaller than in the dunite when there is a significant component of melt flux flowing from the harzburgite into the dunite. The permeability has a pronounced effect on the boundary layer thickness. The width of the boundary layer in the harzburgite is comparable to or smaller than that in the dunite if the porosity of the harzburgite is much smaller than that of the dunite. In summary, composition gradients developed around the dunite-harzburgite contact are expected be sensitive to the permeability structures of the dunite and harzburgite, melt flow rates, and flow directions. Limited geochemical field observations indicate that the dunite boundary layer is generally thicker than the harzburgite boundary layer for narrow dunites ($<$ 10s cm), whereas the harzburgite boundary layer is commonly comparable to or slightly thicker than in the dunite for wider dunites ($>$ 1 m), though exceptions are common. With more detailed field, laboratory, and numerical studies we hope to be able to infer the rate and time scales of melt extraction and possibly permeability structures of the mantle from the composition gradients measured at dunite-harzburgite contacts.
DE: 1749 Volcanology, geochemistry, and petrology
DE: 3600 MINERALOGY AND PETROLOGY (replaces
DE: 3699 General or miscellaneous
DE: 8434 Magma migration
SC: Volcanology, Geochemistry, Petrology [V]
MN: 2003 Fall Meeting