MR12A-03
Elastic properties of silicate melts at high pressure and implications for low velocity anomalies in the crust and mantle

Monday, 14 December 2015: 10:50
301 (Moscone South)
Alisha N Clark, University of California Davis, Davis, CA, United States and Charles E Lesher, University of California Davis, Department of Earth & Planetary Sciences, Davis, CA, United States
Abstract:
Regions of low seismic velocities in the mantle and crust are commonly attributed to the presence of silicate melt or aqueous fluid. The elastic properties of silicate melts are typically modeled at high pressure using equations of state developed for crystalline materials. However, amorphous silicates spanning a wide range of composition and structure, i.e. SiO2 to MgSiO3, and including naturally occurring basalt compositions, exhibit a weak dependence of P-wave velocity on density in clear violation of Birch’s law, which governs the behavior of crystalline materials. This anomalous behavior is attributed to the high degree of flexibility of the silicate network on loading that may be a general property of naturally occurring silicate melts at crustal and upper mantle conditions. If this is the case, P-wave velocities for silicate melts will be significantly less pressure dependent than previously assumed, which in turn will enhance the effects of melt fraction on lowering aggregate mantle seismic velocities. Here we present VP calculated for partially molten mantle up to 20 GPa showing that melt fractions purported to explain VP reductions associated with the lithosphere-asthenosphere boundary may be overestimated by 15%, while those reported for the transition zone may be overestimated to an even greater extent. Moreover, we predict that d lnVS/d lnVP (RSP) should vary little across low velocities regions within the upper mantle due solely to the presence of melt, but will be strongly influenced by how melt is distributed, consistent the work of [1]. Finally, RSP is found to be relatively insensitive to type of fluid present, contrary to conventional wisdom, and thus caution is warranted in attributing changes in RSP to either silicate melt or aqueous fluids. The implications of these findings for interpreting low velocity anomalies beneath hotspots and arcs (e.g. Iceland and Japan) will be discussed.

[1] Takei, Y. (2002) JGR vol. 107