Probing Seismically Melting Induced Mantle Heterogeneities in Thermal-chemical Convection Models

Abstract:

Two regions at the base of the Earth's mantle (the Large Low-Shear Velocity Provinces) pose a fundamental problem in understanding large-scale mantle dynamics and history. Are they dense piles of (possibly primordial) material separated from mantle circulation, or large-scale thermal features which are part of global mantle convection? Or some combination of the two?

We use our numerical 3D spherical mantle convection code to perform simulations of the Earths mantle dynamical evolution. We drive the surface velocity of the model according to 200 Ma plate motion reconstructions, to arrive at Earth-like structures in the mantle at present day. Variations in bulk chemistry will be tracked in two ways: 1) by starting the calculations with a (primordial) dense layer at the base of the mantle, and 2) by tracking basalt fraction which is fractionated upon melting close to the surface. The resulting distribution of chemical heterogeneity and temperature will be converted to seismic velocities. This will be done with a thermodynamical database (Stixrude & Lithgow-Bertelloni, GJI, 2005, 2011), allowing us to compare the model with previous observations of triplications and waveform complexity near the margins of the LLSVPs. These observations have been taken as proof that strong chemical variations are present; our simulations can be used to show whether this is true, or if purely thermal convection can also cause these features. We simulate finite-frequency, 3D seismograms at ~5 s period and compare these with previous studies.