New Insights into Global Plume Buoyancy and Heat Fluxes from Numerical Models of Plume-Lithosphere Interaction

Earth's dynamic evolution is controlled by the interplay between mantle convection and plate tectonics. While subducted plates stir the mantle, upwelling plumes can lubricate, push, and break up plates. As the surface expression of upwelling plume dynamics, the plume buoyancy flux is traditionally estimated as the cross-sectional area of the hotspot swell multiplied by plate velocity (for intraplate hotspots) or multiplied by the full-spreading rate (for ridge-centred hotspots).This classical approach implies two big assumptions: that the swell is fully isostatically compensated by the hot ponding plume material at the base of the lithosphere; and that this plume material spreads at exactly the same speed as the overriding plate moves. However, geophysical observations and numerical models demonstrate that those assumptions are wrong. Hotspot swells are largely dynamically instead of fully isostatically compensated; to some extent, swells are further compensated by sublithospheric erosion [1]. Moreover, at least some plumes spread faster than plate motion [2]. For example, evidence in the North Atlantic from prominent V-shaped ridges, ephemeral landscapes, and off-axis uplift of oceanic gateways suggests that along-axis asthenospheric velocities can be an order of magnitude faster than the full plate-spreading rate near Iceland [3]. Thus, classical estimates for the buoyancy fluxes of deep-seated mantle upwellings may be strongly biased by surface-plate velocities [4]. Alternative estimates of plume buoyancy flux assume a constant swell decay timescale [4] but without any physical underpinning. As detailed estimates of dynamic seafloor topography are now available [5], it is time to revisit the buoyancy fluxes and, thereby, the mass and heat fluxes carried by mantle plumes.Here, we explore high-resolution regional-scale geodynamic models with a free surface to study plume-ridge interaction and swell compensation. We consider composite diffusion-dislocation creep in our models. We investigate the effects of plume temperature/radius, plate velocity (or spreading rate for ridge-centred hotspots), and mantle rheological parameters on plume-lithosphere interaction and swell support. Preliminary results demonstrate that plume spreading is significantly faster than plate motion for intermediate-to-large plumes at realistic rheological conditions. From this result, we update estimates of plume buoyancy fluxes, showing that the total heat flux carried by plumes across the core-mantle boundary is significantly larger than previously thought.  Reference List1. Cadio et al., 2012; doi:1016/j.epsl.2012.10.0062. Ribe & Christensen, 1999; doi:10.1016/S0012-821X(99)00179-X3. Poore et al., 2011; doi: 10.1038/ngeo11614. Hoggard et al., 2020; doi: 10.1016/j.epsl.2020.1163175. Hoggard et al., 2016; doi: 10.1038/ngeo2709