WATER IN THE OCEANIC CRUST; LESSENS FROM ABOVE AND BELOW

Ophiolite studies have facilitated the interpretation of geophysical data and geological observations collected over the world's oceans. The "ophiolite model" of Eldridge Moores and others is the standard against which models of seismic velocity, density, and other geophysical parameters are tested. However, the pervasive alteration of ophiolitic rocks still has not been completely reconciled with evidence for hydrothermal circulation observed on the mid-ocean ridges and flanks.

From sea floor observations we know that high temperature hydrothermal venting occurs many places along mid-ocean ridges, regardless of the spreading rate and morphology of the ridge. As the crust ages, these spectacular manifestations of hydrothermal circulation are cut off from their heat source, either by distance or by sealing of cracks in the crust. However, there is substantial geological and geophysical evidence that hydrothermal cooling of the ocean lithosphere must continue for tens of millions of years. It is generally believed that this cooling is accommodated by diffuse flow, primarily in the upper crust.

Submersible observations illustrate that much of the hydrothermal flow is accommodated by fault zones that provide the necessary high permeabilities. On the Gorda Ridge, we observe venting some 3 kilometers away from the axis, and some 300 meters up the rift valley wall. The wall is clearly faulted and it is likely that the faulting focuses the flow. Similar observations have been made on other ridge segments. These faults almost certainly extend into the mid- to lower crust, at least on the ridge flanks, and thus, they may permit water penetration to deeper crustal levels.

Seafloor observations do not constrain the depth of hydrothermal circulation, and we must examine geophysical data for indirect confirmation of circulation models. There is evidence of long-term evolution of seismic velocities in the lower oceanic crust. This evolution cannot be the result of magmatic processes, and they indicate the penetration of water into the mid- to lower crust must occur. It is likely that this flow, and the resulting crustal alteration, is accommodated by the same faults that are observed at the seafloor.