What the rock record tells us about ocean crustal faulting

Abstract:

Imaging and monitoring efforts along mid-ocean ridges, transform faults, and forebulges of subduction zones raise some apparently contradictory points about faulting in the oceanic lithosphere. Faults are conduits to fluid flow, but also reside within broader flow fields that extend beyond the fault planes themselves. Fault-zone properties appear to govern slip dynamics, yet seismicity and deformation also occur far off the map-trace of faults. Geologic observations of fast-spread oceanic crust exposed in seafloor escarpments and analogs in ophiolites offer some insight. (i) The upper crust — the sheeted dikes and deepest lavas — is highly altered, both within fault zones through which hydrothermal fluids flowed, but also in surrounding, low-permeability basaltic rock. (ii) The upper crust is highly fractured over areas tens-of-meters wide, despite relatively low (meters-to-tens of meters) fault displacements. (iii) Detailed structural mapping of such sections finds that fractures become more densely spaced toward fault zones, achieving a critical spacing of ~1 cm, below which the basaltic material is comminuted to breccia and gouge. (iv) Breccia and gouge also have characteristic grain sizes of ~1 cm, below which grain-size distributions steepen. (v) In experiments conducted in the Penn State Rock Mechanics Lab, altered basaltic fault rocks tended to be moderately weak and velocity strengthening. Putting i-v together paints a picture wherein gouge-filled faults indeed localize both fluid flow and stable sliding, but also reside within damage zones which host most of the active seismicity and within which penetrative fluid flow also occurs. Though these observations predominantly come from windows into subaxial processes, this damaged and altered crust is spread off axis where it is deformed and reactivated by transform and bend faults. Moreover, new faults and fractures formed off axis may well follow the same patterns of localization as subaxial faults, minus the flow of very high temperature fluids. An outstanding question, though, is how the lower gabbroic crust and upper mantle behave with respect to the upper crust, the former thought to control a great deal of transform seismicity, and the latter to be hydrated during bend faulting.