Wednesday, 17 December 2014: 1:40 PM
Edward Garnero, Arizona State University, EarthScope National Office, School of Earth and Space Exploration, Tempe, AZ, United States, Allen K McNamara, Arizona State University, Tempe, AZ, United States and Sang-Heon Dan Shim, Arizona State University, School of Earth and Space Exploration, Tempe, AZ, United States
Abstract:
The term large low shear velocity province (LLSVP) represents large lowermost mantle regions of reduced shear velocities (Vs) relative to 1D reference models. There are two LLSVPs: one beneath the central Pacific Ocean, and one beneath the southern Atlantic Ocean and Africa. While LLSVP existence has been well known for several decades, more recently evidence from forward modeling has brought to light relatively sharp margins of the LLSVPs, i.e., the transition from low-to-“normal” Vs occurs over a short lateral distance (probably < ~100 km). This finding is further supported by the strongest lateral dVs gradients in tomography coinciding with locations of sharp LLSVP sides in high-resolution studies. Surface hotspot and large igneous province origination locations mostly map above the present day LLSVP edges. Combined with geochemical arguments that a deep mantle long-lived (possibly primordial) reservoir exists, and geodynamics experiments that demonstrate a dense basal reservoir would be swept by convection to reside beneath upwellings and plumes, a strong argument can be made for dense, chemically distinct material explaining LLSVPs. This presentation will present additional seismic information that needs to be considered for a self-consistent geodynamic and mineralogical framework. For example, there does not appear to be consistency between Vp and Vs reductions defining LLSVPs; however, this comparison is complicated by lowermost mantle Vp models exhibiting greater divergence from each other than Vs models. LLSVP forward modeling usually involves a trade-off between dVs within the LLSVP and LLSVP height/shape; thus continued mapping of heterogeneity within LLSVP is critical. ULVZs might relate to LLSVP chemistry, temperature, and evolution, and thus will be discussed. The chemistry that can explain large and old thermochemical piles is as of yet unconstrained; other mineralogical considerations include understanding the possible role of the post-perovskite phase transition within and outside LLSVPs (which may affect Vs differently from Vp), and the evolution of pile chemistry over time, since geodynamics work demonstrates how mantle material (including deeply subducted MORB) can become downward entrained into piles.
