INTERACTION OF DEEP AND SHALLOW PROCESSES IN THE EVOLUTION OF THE KENYA RIFT

The start of volcanism before rifting in the northern Kenya rift suggests that an asthenospheric thermal anomaly was responsible, not decompression melting due to lithosphere stretching. This volcanism may be partly related to the Ethiopian rift, or even the Anza graben, not the Kenya rift. In the northern Kenya rift the first stage of deformation was the formation of isolated sediment-filled half-graben basins during the Late Oligocene-Early Miocene, perhaps superimposed on lower Tertiary basins. During the Miocene, the location of basins shifted eastwards. This shift is interpreted as being due to strain hardening of the lithosphere during extension caused by a relatively slow strain rate. Relocation of the zone of extension progressively eastwards was possibly caused by migration of the asthenospheric thermal anomaly to the east (which lowered the strength of the crust above the thermal anomaly). The simple McKenzie model of uniform lithosphere stretching does not apparently fit the Kenya rift. Uniform extension may have affected the entire lithosphere but uniform stretching can only be demonstrated for the continental crust. The shape of the geophysically defined base lithosphere under the rift shows much more thinning of the mantle lithosphere than the crust. Consequently, thermal thinning of the mantle lithosphere has to be invoked to explain the discrepancy. Where the asthenosphere lies almost at the base of the crust the surface rift above displays swarms of minor faults and a linear array of Pliocene recent volcanoes. Thus the deep thermal history and the shallow brittle structures of the Kenya rift appear to be closely linked and each has influenced the evolution of the other. Extension estimates for the upper crust and the lower crust are similar, indicating that addition of magma to the crust has not caused an underestimate of lower crust extension. This suggests that either the ratios of magma emplaced within the crust to surface volcanism are much smaller (1:1 to 2:1) than previously thought (5:1 to 10:1), or that entry of magma into the crust is offset by an equal loss of crustal material into the asthenosphere.