Upper mantle structure beneath the Kenya Rift and Arabian Shield has been investigated to advance our understanding of the origin of the Cenozoic hotspot tectonism found there. A new seismic tomographic model of the upper mantle beneath the Kenya Rift has been obtained by inverting teleseismic P-wave travel time residuals. The model shows a 0.5–1.5% low velocity anomaly below the Kenya Rift extending to about 150 km depth. Below ∼150 km depth, the anomaly broadens to the west toward the Tanzania Craton, suggesting a westward dip to the structure.

The P- and S-wave velocity structure beneath the Arabian Shield has been investigated using travel-time tomography. Models for the seismic velocity structure of the upper mantle between 150 and 400 depths reveal a low velocity region (∼1.5% in the P model and ∼3% in the S model) trending NW-SE along the western side of the Arabian Shield and broadening to the northeast beneath the MMN volcanic line. The models have limited resolution above 150 km depth everywhere under the Shield, and in the middle part of the Shield the resolution is limited at all depths.

Rayleigh wave phase velocity measurements have been inverted to image regions of the upper mantle under the Arabian Shield not well resolved by the body wave tomography. The shear wave velocity model obtained shows upper mantle structure above 200 km depth. A broad low velocity region in the lithospheric mantle (depths of ≤ ∼100 km) across the Shield is observed, and below ∼150 km depth a region of low shear velocity is imaged along the Red Sea coast and MMN volcanic line.

A westward dipping low velocity zone beneath the Kenya Rift is consistent with an interpretation by Nyblade et al. [2000] suggesting that a plume head is located under the eastern margin of the Tanzania Craton, or alternatively a superplume rising from the lower mantle from the west and reaching the surface under Kenya [e.g., Debayle et al., 2001; Grand et al., 1997; Ritsema et al., 1999].

For the Arabian Shield, the models are not consistent with a two plume model [Camp and Roobol, 1992] because there is a continuous low velocity zone at depths ≥ 150 km along the western side of the Shield and not separate anomalies. The NW-SE trending low velocity anomaly beneath the western side of the Shield supports the Ebinger and Sleep [1998] model invoking plume flow channeled by thinner lithosphere along the Red Sea coast.

The NW-SE low velocity structure beneath the western side of the Shield could also be the northern-most extent of the African Superplume. A low velocity anomaly beneath Ethiopia [Benoit et al., 2006a,b] dips to the west and may extend through the mantle transition zone. The observed low velocities in the upper mantle beneath the Arabian Shield could be caused by hot mantle rock rising beneath Ethiopia and flowing to the north under the Arabian Shield.