How does the Earth`s crust grow at divergent plate boundaries? A unique opportunity in Afar, Ethiopia

Abstract

In three field campaigns, we have acquired usable, generally high quality, magnetotelluric (MT) data at 40 sites, most with transient electromagnetic (TEM) data for static shift correction, over two magmatic segments in Afar, Ethiopia, a challenging fieldwork environment. Two profiles are across the rift axes and roughly orthogonal to them, one across a currently active segment, the other across an inactive segment. Activity on the former began in September 2005 with a volcanic eruption, the injection of a mega-dyke along the full 60 km long segment, 163 earthquakes with magnitude greater than 3.9, and over 8m of crustal movement. The third is an oblique profile towards the Dabbahu volcano that erupted in September 2005. In addition, we occupied a few sites to the north of the currently inactive segment, the final one as close as accessible to an area of rapid subsidence, thought to be deflation of a deep magma chamber filling a shallower chamber along the active segment to the north. After robust processing, dimensionality and geoelectrical strike direction assessment, static shift correction and rotation into geoelectrical strike co-ordinates, the MT data along the three profiles have been inverted for 2D models of resistivity beneath the profiles. The result for the first profile, across the active Dabbahu segment, was presented in the report on loan 855, along with background information on current activity on the segment. It can usefully be read in conjunction with this report, which therefore concentrates on the results from the other two profiles. In contrast to the Dabbahu segment model, the profile across the currently inactive Hararo segment is characterised by broadly resistive material in the sub-surface, suggesting an absence or only small quantities of melt in the sub-surface there. The model along the oblique profile indicates a very sharp contrast from relatively resistive material in the crust away from the volcano to the south-west to a very conductive sub-surface in the immediate vicinity of the volcano, to depths of approximately 35km. We interpret this to be a magma chamber beneath the volcano. The conductor extends to the south-west at sub-Moho depths, with slightly lower conductivities presumably indicating lower melt concentrations. The model agrees well with that for the Dabbahu segment at their point of intersection. We will be attempting 3D inversion of the full data set, though the distribution of sites is far from ideal for this. We have not yet undertaken a full sensitivity analysis of the models across the second and third profiles. The data from all three profiles are being interpreted in conjunction with other geophysical, geological and remote sensing data to understand the processes associated with continental rupture.