Ancient subducted oceans controlling the positioning of deep mantle plumes

Seismic imaging of the Earths interior reveals plumes originating from relatively hot regions of the lower mantle, surrounded by cooler material thought to be remnants of ancient subducted oceans. Based largely on geophysical data, two opposing hypotheses dominate the discussion on dynamics at the base of mantle: the large hot anomalies are thermo-chemical in nature; or, alternatively, they are purely thermal plume clusters. In previous modelling studies, deep chemical heterogenities have been argued to be essential in developing appropriate present-day plume positions. Here, we quantify how the chemical composition of large, hot regions in the deep mantle influences the location of rising mantle plumes using numerical 3-D global mantle convection models constrained by 410 million years of palaeo-ocean evolution. For the first time, we show that purely thermal convection can reproduce the observed positions of present-day hotspots. By demonstrating that a lower mantle without large-scale chemical heterogeneities can generate appropriate global dynamics, we illustrate the power of sinking ocean plates to stir mantle flow and control the thermal evolution of the mantle. Because our models with a thermo-chemical anomaly reproduce the observed hotspot positions equally well, we posit that the deep hot anomalies in the mantle are purely passive in global dynamics - regardless of their (thermal or chemical) origin.