A magnetotelluric and magnetovariational study of the Gregory Rift Valley, Kenya

Magnetotelluric measurements made at 10 locations in and around the Kenyan Rift Valley in 1973–74 have been processed to yield single station induction vectors and magnetotelluric principal impedances. Following a brief description of the field work and the data processing procedure, the results are presented and subjected to both qualitative and quantitative interpretations. The induction vector results indicate a concentration of current below the rift valley. Since the maximum impedances are along the rift for sites within the valley but at right angles to the rift outside it, the apparent resistivity results are qualitatively consistent with the existence of a good conductor there. Within the rift maximum apparent resistivities lie between 2 and 20 Ωm. The persistence of small induction responses and low apparent resistivities down to the shortest periods at which measurements were made suggests that the conductor lies at very shallow depths below the rift. The quantitative interpretation of the magnetotelluric results is given in terms of one‐and two‐dimensional conductivity models. It is shown that the data require the presence of conductive material at a depth less than 8 km below the rift floor but that the required thickness of this upper crustal conductor (> 5 km) is too great to be explained in terms of a conductive infill of the rift. The favoured explanation is in terms of high temperatures and water saturation of the crust under the rift. To satisfy the long‐period data, conductive material is also required at depths of greater than 30 km, corresponding unambiguously to the upper mantle below the rift. Because of the obscuring effect of the crustal conductor, the depth to the top of and the thickness of the upper mantle conductor cannot be resolved.