V23E-08
Alternative hypothese for the origin of the ultra-low velocity zones

Tuesday, 15 December 2015: 15:25
307 (Moscone South)
Jie Li1, Jiachao Liu1, Rostislav Hrubiak2 and Jesse Smith3, (1)University of Michigan Ann Arbor, Department of Earth and Environmental Sciences, Ann Arbor, MI, United States, (2)CeSMEC-Mechanical & Materials, Miami, FL, United States, (3)HPCAT, Carnegie Institution of Washington, Argonne, IL, United States
Abstract:
Decades ago seismic tomography revealed peculiar patches known as the ultra-low-velocity zones (ULVZ), located just above the Earth's core-mantle boundary. Typically 5 to 40 km in height, and 100 km in lateral extent, the ULVZs are denser than the surrounding mantle by about 10%, and slower in shear and compressional wave velocities by roughly 30% and 10%, respectively. Elucidating the origin of ULVZs is a key issue in understanding mantle heterogeneity and unraveling the history of chemical differentiation in deep Earth. Existing models invoking partial melting of silicate lithology require specific solidus temperatures to produce isolated melt regions, and they may not be able to account for the large density excess.The scenarios involving iron-rich post-bridgmanite or ferropericlase have been questioned by recent theoretical results showing that iron enrichment in crystalline phases cannot simultaneously explain the observed density and velocity anomalies. Thus the nature of the ULVZs has remained enigmatic.

Here we show that the eutectic melting curve of the iron-carbon binary system intersects with the present-day geotherm near the base of the Earth's mantle, suggesting the presence of metallic melt in the D" layer. Using an established model for solid-liquid compositions and approximate values of density and elastic parameters, we found that introducing a suitable fraction of metallic melt could match all the seismic observations, and the fraction depends on the wetting behavior of the melt. We propose a number of alternative hypotheses for the origin of the ultra-low velocity zones, including short-lived ULVZs with metallic melt that are supplied by subducted slabs, long-lasting ULVZs involving both metallic and silicate melts, and ULVZs containing iron-rich solids and residual metallic melt. These hypotheses can be tested by future studies on surface tension, element partitioning, elastic properties, and dynamic modeling.