Plate and Plume Flux: Constraints for paleomagnetic reference frames and interpretation of deep mantle seismic heterogeneity. (Invited)

Plate and plume flow are dominant modes of mantle convection, as pointed out by Geoff Davies early on. Driven, respectively, from a cold upper and a hot lower thermal boundary layer these modes are now sufficiently well imaged by seismic tomographers to exploit the thermal boundary layer concept as an effective tool in exploring two long standing geodynamic problems. One relates to the choice of an absolute reference frame in plate tectonic reconstructions. Several absolute reference frames have been proposed over the last decade, including those based on hotspot tracks displaying age progression and assuming either fixity or motion, as well as palaeomagnetically-based reference frames, a subduction reference frame and hybrid versions. Each reference frame implies a particular history of the location of subduction zones through time and thus the evolution of mantle heterogeneity via mixing of subducted slab material in the mantle. Here we compare five alternative absolute plate motion models in terms of their consequences for deep mantle structure. Taking global paleo-plate boundaries and plate velocities back to 140 Ma derived from the new plate tectonic reconstruction software GPlates and assimilating them into vigorous 3-D spherical mantle circulation models, we infer geodynamic mantle heterogeneity and compare it to seismic tomography for each absolute rotation model. We also focus on the challenging problem of interpreting deep mantle seismic heterogeneity in terms of thermal and compositional variations. Using published thermodynamically self-consistent mantle mineralogy models in the pyrolite composition, we find strong plume flux from the CMB, with a high temperature contrast (on the order of 1000 K) across the lower thermal boundary layer is entirely sufficient to explain elastic heterogeneity in the deep mantle for a number of quantitative measures. A high excess temperatures of +1000--1500 K for plumes in the lowermost mantle is particularly important in understanding the strong seismic velocity reduction mapped by tomography in low-velocity bodies of the deep mantle, as this produces significant negative anomalies of shear wave velocity of up to -4%. We note, however, that our results do not account for the curious observation of seismic anti-correlation, which appears difficult to explain in any case. Our results provide important constraints for the integration of plate tectonics and mantle dynamics and their use in forward and inverse geodynamic mantle models.