Role of inherited structures and magmatism in North Tanzania from high resolution teleseismic P and S body-wave tomographies.

The North Tanzanian Divergence (NTD) is a zone of rift initiation. Its surface expression results from interactions between deep-mantle (mantle plume), lithospheric (inherited rheology and stratification, melting...) and crustal (dyke propagation, fault activation...) processes. However, the role of each process on the observed surface activity is still debated, because highly difficult to decorrelate.

We recently carried out a study to obtain enhanced P and S-wave tomography, from the surface down to 150-200 km depth. The particularity of our method consists in its initial velocity model. It is composed of a 1D IASP91 regional velocity model in which we inserted an a priori 3D crustal velocity model with a fine grid. This crustal model was deduced from an independent local tomography inversion.

The P and S images obtained, resulting from the teleseismic inversion of this hybrid method, show strong contrasted velocity anomalies: from 10 % of P (Vp) and S velocity (Vs) variation on the craton, to -17 % below the rift axis. The anomalies locations are consistent with the surface geology (rifting basin, border faults, volcanoes). At a regional scale, the strongest velocity contrasts correspond to the lithospheric inherited structure (Tanzanian craton and Proterozoic belts) boundaries, which control the propagation of the rift. In particular, the Masai cratonic block, south of the NTD, is inferred to have a strong influence in the rift evolution. The transition from the North-South axial valley into three diverging rift arms (Eyasi, Natron-Manyara and Pangani) is likely due to the change in rheology and to the presence of magma along inherited sutures between the craton and the mobile belts.

However, interrogations about the role of the thermal changes, the melt/fluid presence and the mantle composition in the NTD on these velocity anomalies still remain. To distinguish which parameters are acting in the rift, we realize a Vp/Vs ratio map. With this new data, and in the light of parallel petrological studies, we interpret the Vp/Vs anomalies in term of gas and/or melt concentration zones.