HR: 14:40h
AN: V53F-05 INVITED     [Abstracts]
TI: Water Recycling, Lower Mantle Slab Subduction, and Viscous Layering of the Deep Mantle
AU: * Williams, Q
EM: quentw@es.ucsc.edu
AF: Dept. of Earth Sciences, UC Santa Cruz, 1156 High St., Santa Cruz, CA 95064 United States
AU: McNamara, A
EM: Allen.McNamara@asu.edu
AF: Dept. of Geological Sciences, Arizona State U., Box 871404, Tempe, AZ 85287 United States
AU: Garnero, E
V53F-05 AF: Dept. of Geological Sciences, Arizona State U., Box 871404, Tempe, AZ 85287 United States
AB: We explore the causes and possible consequences of a water/hydrogen-depleted layer in the lowermost ~1000 km of Earth`s mantle. At least three distinct, non-exclusive mechanisms exist that could generate such a layer: (1) descending melts could extract water from the deep mantle, and possibly sequester it within D``; (2) hydrogen could be stripped from deep mantle material during core formation, through formation of iron hydrides; and (3) the accreting planet could have nearly completely degassed, with the terrestrial water budget being accreted in a late hydrous veneer. In the latter two instances, the water budget of the mantle, and particularly the deep mantle, must entirely be generated by injection of water into the interior from the near surface. Our hypothesis is thus that the lower portion of Earth`s mantle might be (or have been) essentially dry, in contrast to the possible presence of 10's to 100's of ppm water in the overlying material. The principal geophysical effect of a water-depleted zone likely involves a marked increase in viscosity: for reference, such a decrease in water content produces about a 2-order of magnitude increase in the viscosity of upper mantle material. Fluid dynamic simulations show that a layer with a 2-order of magnitude viscosity increase in the bottom 1000 km of Earth`s mantle produces a substantial impediment to subduction, with subducted material laterally spreading out above this viscous layer. This behavior is compatible with tomographic images showing a lack of slab continuity into the deepest mantle, and the viscosity contrast thus produces a barrier to water ingress into the deep viscous layer, allowing it to remain anhydrous for extended time periods. Notably, the boundary between the viscous layer and overlying mantle and slab material undergoes substantial deflections, and because of the chemical similarity of the layers, should be seismically undetectable. Our results provide a straightforward mechanism through which geochemically primordial material could be retained for extended periods, and produce a means for impeding heat flow through the deep mantle. The natural inference for water recycling is that hydration of the mantle has been driven by subduction, and that any leakage between an anhydrous viscous layer and the overlying mantle should produce a deepening of the viscous boundary with time. Thus, a continuous progression from pseudo-two-layer to one-layer mantle convection, hinging upon the degree to which water is present at depth, could be proceeding through time with water content exercising the ultimate control on mantle mixing. In short, water-depletion-induced viscous layering may be the simplest mechanism for inducing a time-dependent ``stealth`` layer within the deep mantle.
DE: 1025 Composition of the mantle
DE: 1031 Subduction zone processes (3060, 3613, 8170, 8413)
DE: 8124 Earth's interior: composition and state (1212, 7207, 7208, 8105)
DE: 8147 Planetary interiors (5430, 5724, 6024)
SC: Volcanology, Geochemistry, Petrology [V]
MN: Fall Meeting 2005