A stationary mantle plume heats the overlying lithosphere from below, generating melts and causing lithospheric erosion. When the lithosphere moves away from the plume, it can be re-established by melt residues that have been since attached, partly due to cooling, at the base of the lithosphere. The Etendeka flood basalt region in northwest Namibia used to overlie the Tristan da Cunha mantle plume in the Early Cretaceous and thus provides an ideal place to examine plume-lithosphere interaction. Here we determine the upper mantle shear wave velocity structure of southern Africa down to 400 km depth by waveform inversion of multi-mode Rayleigh waves in order to find the imprints left by the Tristan da Cunha mantle plume. Thick lithosphere with high shear wave velocities is observed beneath the Congo and Kalahari Cratons, extending down to 200 km depth with the largest thickness beneath the Limpopo Belt. Along the landfall of the Walvis Ridge and Damara Belt, our model reveals a thick lithosphere down to 100–200 km depths. The thick lithosphere seems to constitute the Congo Craton, but the thick Congo cratonic lithosphere is extended farther south than observed in most of the previous models. We propose that the lithosphere along the landward extension of the Walvis Ridge was affected by magmatic processes related to the plume, but has been partially reconstructed since then by melt depletion and successive cooling. A high velocity anomaly beneath the northwest coast of Namibia down to a depth of ∼80 km coincides with distribution of the Etendeka basalt at the surface and the seaward-dipping reflectors at the continental margin. At sub-lithospheric depths, a low-velocity anomaly (at 100–200 km depth) and a high-velocity anomaly (at 250–350 km depth) can be detected. The origin of these anomalies could be related to an ongoing edge-driven convection process.
Read full abstract