Geophysical Constraints on Partial Melting 


A. Mantle Electromagnetic and Tomography experiment

-51 ocean-bottom seismometers were deployed across the EPR Nov 1995-May 1996

-Seismometers arranged in 2 linear arrays 800 km long

-This experiment is in the middle of longest, straightest, fastest-spreading (14.5 cm/yr) section of MOR

B. Goals - to determine:

-Pattern of upwelling beneath a ridge

-Geometry of the region of partial melting

-Melt fraction in that region

-Distribution and interconnectedness of melt

C. Models of upwelling and decompression melting


-Drag at bottom of plates induces a broad zone of upwelling/melting : melt migrates horizontally to ridge axis (against buoyancy, which drives it upward); implies linear upwelling along ridge axis

-Dynamic: buoyancy of melt + reduction in mantle density by removal of melt + lower viscosity in upwelling zone => narrow zone of upwelling & melting (<10 km across); melt movement is mostly vertical; most melting is beneath ridge axis; implies 3D upwelling, concentrated in discrete centers, coupled with lateral movement of melt along ridge

-Are ridges passive features that tap a homogenized asthenosphere, or are they linked with deeper mantle structures & whole-mantle convection?


A. Earlier observations

-Ridge is distinctly asymmetric: shallower to west (i.e., Pacific Plate is subsiding slower than Nazca); more seamounts to west; WRT hot spot reference frame, Pacific Plate is moving west twice as fast as Nazca Plate is moving east => EPR is migrating west; overlapping spreading centers propagate along ridge and transfer seafloor from Pacific to Nazca Plate => half-spreading rate is faster toward the east

-All of this suggests underlying asymmetry of temperature distribution and melt production

B. Bathymetry and gravity

-Ridge axis is 5-20 km wide, stands 200-300 m above level of subsiding ocean floor

-Gravity anomaly: low gravity (mass deficit) area beneath ridge; may ~ narrow low-density melt zone reaching 10s of km into mantle; if upwelling is concentrated beneath anomaly, melt would have to be distributed along the axis, perhaps in an axial magma chamber; seismic evidence for this: reflections 0.8-2.0 km below ridge, <100 m thick, <1 km wide

III. Results

A. Asymmetry

-Not due to variation in crustal thickness => must be due to differences in mantle density (~ melting?)

-Low-velocity zone : zone is 100s of km wide ; not confined to narrow zone beneath ridge axis; center is displaced from ridge

--More seamounts & off-axis lava flows to west than to east.

-Recent volcanism occurs above areas with the lowest seismic velocities

B. Anisotropy

-Mantle is anisotropic: EQ waves travel faster in some directions than in others

-May be explained by: alignment of olivine crystals by mantle flow; or, alignment of melt-filled cracks along axis.

-Anisotropy and asymmetry may be caused by return flow from the Pacific Superswell

C. Contradictory results

-Lowest velocities (P, S, surface) all displaced from ridge axis => highest % melt is under Pacific Plate: this suggests dynamic upwelling that originates deep in the mantle; ridge is migrating west to stay above this zone of upwelling

-But, no anomalous velocities at 300-400 km depth => not deep upwelling => passive: isotopic anomalies are consistent with a deeper origin, or they could result from mantle heterogeneity (enriched pod)

Хостинг от uCoz