HR: 08:50h
AN: S11A-03    [PDF]
TI: Origin of the Low Velocity Zone in Oceanic and Continental Regions
AU: * Stixrude, L
EM: stixrude@umich.edu
AF: University of Michigan, Department of Geological Sciences, Ann Arbor, MI 48109-1063 United States
AU: Lithgow-Bertelloni, C
EM: crlb@umich.edu
AF: University of Michigan, Department of Geological Sciences, Ann Arbor, MI 48109-1063 United States
AB: By virtue of its intermediate position between the upper thermal boundary layer and the nearly adiabatic interior, the low velocity zone holds important clues to the relationship between plate tectonics and the earth's internal dynamics and to the origin of the crust. Forty years after its discovery, the origin of this region is still debated. Some have argued that the presence of small amounts of partial melt are needed to explain the extremely low shear wave velocities observed, while others have argued that no partial melt is required by seismological observations. The question is complicated by the importance of attenuation and anisotropy in the low velocity zone, and, perhaps most importantly, the limited information we have had until recently of the elastic properties and phase equilibria of relevant mantle assemblages. We apply a new method to the construction of one-dimensional models of the low velocity zone in a number of tectonic settings, which allow direct comparison with seismological observations. The method is based on a thermodynamic formulation that allows us to predict phase equilibria, density, and compressional and shear wave velocities of multi-component, multi-phase mantle assemblages self-consistently. Such an approach is important because many phase transformations occur over the upper few hundred km of the mantle including plagioclase=spinel=garnet, and the gradual pyroxene to garnet transition, and because the phases involved have very different elastic properties. To extrapolate experimental data that is often still limited, we use an anisotropic generalization of the Mie-Gruneisen-Debye theory modified after Davies (1974), which is expected to capture much better the behavior of elastic constants at high pressure-temperature conditions than polynomial expansions that have sometimes been used in the past. We find that no partial melt is required to account for the low velocity zone in oceanic regions. Along half-space cooling geotherms, the low velocity zone is pronounced even in the elastic limit: the low velocity zone becomes more pronounced and moves to shallwer depths as age decreases. These age dependent features are seen in many models of the low velocity zone. The only feature of the low velocity zone that is not captured by the elastic limit is the magnitude of the velocity anomaly. However, the lowest velocities are readily accounted for by plausible values of attenuation and attendant dispersion without invoking partial melt. One-dimensional shear and compressional velocity profiles that include plausible Q models agree well with published one-dimensional seismological models over a range of lithospheric age. We will report on extensions of our work to continental areas and to the more direct comparison with seismological observations via the computation of phase velocities.
DE: 3909 Elasticity and anelasticity
DE: 3939 Physical thermodynamics
DE: 7218 Lithosphere and upper mantle
DE: 8124 Earth's interior--composition and state (old 8105)
SC: Seismology [S]
MN: 2003 Fall Meeting