Earth's surface is divided into seven major tectonic plates, and a larger number of smaller plates.  The plates are cold and brittle, and move across the planet as the underlying hot, soft, mantle rocks convect like a pot of thick soup as Earth slowly loses heat.  Most people are aware of the plate boundaries where earthquakes and volcanoes are found as the plates slide past each other (large fault systems such as the San Andreas in California), over the top of each other (ocean trenches such as Aleutian Trench), or away from each other (mostly mid-ocean ridges).  However, the largest plate boundary on Earth is the boundary between the brittle plates and the convecting mantle, which geologists call the Lithosphere-Asthenosphere Boundary, or LAB.  Because this boundary is up to 100 km deep and doesn't produce earthquakes, we know very little about it.  The goal of this study is to learn more about the LAB by deploying seismic and electromagnetic instruments in the central Atlantic, from the mid-ocean ridge where the South American and African plates are being formed today, to a distance of 1,000 km from the ridge where the plates are 40 million years old.  When rocks get hot or molten, seismic waves get slower and electrical conductivity gets larger, so we can study how the LAB changes depth and character as the plates get older and colder.

In March 2016 we deployed 39 pairs of electromagnetic (EM) recorders and seismic recorders on the seafloor in the central Atlantic.  The seismic experiment was operated out of the University of Southampton, and through NSF funding EM instruments from Scripps Institution of Oceanography were placed at the same locations as the seismic instruments, thereby taking no additional ship time (which is expensive!).  The seismic instruments needed to record far away earthquakes for a year, so although the EM experiment did not need that much data (the instruments recorded for 70 to 100 days before the batteries failed), they too were left out for a year.  We recovered 38 of the 39 EM instruments in early 2017, all with good data. 

The electric and magnetic fields recorded on the seafloor were processed to create magnetotelluric (MT) data, essentially the ratio of the electric to magnetic fields, and the phase difference between them, as a function of frequency.  The MT data can then be modeled to obtain electrical conductivity estimates as a function of depth and position.  The first thing we discovered as we processed the MT data is that although we were in the middle of the ocean, the African coastline was distorting the MT response.  This is an unusual type of distortion, and although a couple of other studies had observed it in one or two MT stations, nobody had explained it.  We have carried out extensive model studies, and we now understand that the distortion comes from electric currents leaking from the ocean into the electrically conductive part of the mantle.  We have submitted a journal manuscript describing this effect, which is important because details of the distortion are sensitive to the LAB.

The three-dimensional nature of the coastlines makes the interpretation of the MT data difficult, but with the understanding of the distortion we have modified our two-dimensional modeling code to take a version of the MT data that is less sensitive to 3D distortion.  Our models show that there is a layer of partially molten rock at the depths associated with the LAB, that this layer is thin in places and not continuous, and is in fairly good agreement with low seismic velocities and contrasts in velocity.  The depth of the layer increases as one moves away from the ridge, and agrees well with the predicted depth of the 1,200 degree C isotherm, suggesting that melt is being trapped beneath a freezing horizon.  We see magma underneath the ridge, but displaced to one side, perhaps as a result of asymmetric plate motion. 

This work is ongoing and will continue long after the life of this award, but the outcome at this time is that the data collection was successful, we have gained a new understanding of how coastlines affect marine MT data, and that the LAB in the Atlantic is associated with melting at the frozen base of the lithosphere.  Melting at the LAB will act as a lubricant to plate motion, and help decouple the overlying plate from the mantle.

Last Modified: 11/28/2018
Modified by: Steven C Constable