Upper mantle shear wave velocity structure beneath the East African plateau: evidence for a deep, plateauwide low velocity anomaly

The shear wave velocity structure of the upper mantle beneath the East African plateau has been investigated using teleseismic surface waves recorded on new broadband seismic stations deployed in Uganda and Tanzania, as well as on previously deployed stations in Tanzania and Kenya. Rayleigh wave phase velocities at periods between 20 and 182 s, measured with a two-plane wave method, have been used to create phase velocity maps, and dispersion curves extracted from the maps have been inverted to obtain a quasi-3-D shear wave velocity model of the upper mantle. We find that phase velocities beneath the Tanzania Craton and areas directly north and west of the craton are faster, at all periods, than those beneath the Western and Eastern branches of the East African Rift System. At periods <50 s, the western branch is slower than the Eastern Branch, but at periods greater than 50 s, this relationship is reversed. Anisotropy is found at all periods, with a generally northsouth fast polarization direction. The shear wave velocity model shows a seismically fast lithosphere (lid) beneath the Tanzania Craton to depths between 150 and 200 km. The fast velocities in this depth range extend to the north beneath the Uganda Basement Complex and to the east beneath the northern Tanzania divergence zone, indicating that these regions together form a rigid block around which rifting has occurred within weaker mobile belt lithosphere. The Eastern and Western branches are slower than the craton at lithospheric mantle depths, and both branches show variable structure in the upper 200 km of the mantle, with the lowest velocities found beneath areas of Cenozoic volcanism. At depths greater than similar to 225 km, a low velocity anomaly is present beneath the entire East African plateau that may extend into the mantle transition zone. Velocities in the low velocity region are reduced by =10 per cent relative to lid velocities, and if attributed only to temperature variations, would represent an unrealistic thermal perturbation of >400 K. Consequently, it is likely that the velocity reduction reflects a combination of thermal and compositional changes, and also possibly the presence of partial melt. The width and thickness of the low velocity anomaly is greater than typically expected for a plume head and is more easily attributed to an upward continuation of the lower mantle African superplume structure into the upper mantle.