Using surface waves to image melt migration pathways and storage beneath the northern East African Rift

Continental rifting is thought to develop from a combination of mechanical stretching and magma assisted rifting. The northern East African Rift is a unique location where we can observe subaerially the initial stages of rifting through to incipient seafloor spreading, as well as the spatial extent of extensional processes away from the rift valley. Multiple models have been proposed to understand the evolution of lithospheric stretching and magmatism in the northern East African Rift, however previous seismic studies are not directly comparable for areas on and off rift due to variations in method, resolution, and scale. It is vital for our understanding of magma assisted rifting processes to have one model that allows comparisons laterally and in depth, a goal that can be achieved in this environment. Here, I invert surface waves from ambient noise and tele seismic Rayleigh waves extracted from 269 seismic stations present between1999 and 2017 to image the Earth’s velocity structure beneath the northern East African Rift System from 5 – 210 km depth. I then use Love waves from ambient noise data to investigate radial anisotropy at crustal depths to determine crustal layering and the depth and shape of magma storage. Shear velocities are everywhere slow in the mantle, with velocities in the rift slow enough (<4.10 ± 0.04 km/s) to require pervasive partial melt. At asthenospheric depths slow velocity anomalies (<4.15 km/s ± 0.04 km/s, 80 – 130 km depth) are not directly beneath melt-rich crustal regions, suggesting mantle melt is ephemeral and/or melt migrates laterally during ascent. Furthermore, the anomalies are segmented along rift, existing in areas that have not undergone significant crustal thinning (segments ∼110 x 80 km wide, ∼60 – 120 km deep),suggesting segmented melt supply starts prior to significant plate deformation. Off rift a fast lid is present at depths of 60 - 80 km but is obscured within the rift suggesting melt is infiltrating the lithosphere. At crustal depths velocities are laterally heterogenous and some of the changes can be attributed to variations in crustal thickness. However, velocities beneath the Main Ethiopian Rift (MER) and the off rift Ethiopian Plateau are slow enough to require melt which I interpret as ongoing magmatic emplacement both on and off rift. The MER is significantly slower than Afar suggesting crustal thickness may be a factor in melt residence time. Anisotropy is required from 5 – 30 km depth suggesting the crust is inherently layered. Effective medium theory suggests thin compositional layering of felsic and mafic intrusions can account for anisotropy up to 4%,however to reconcile the highest observed anisotropy (7%) and lowest velocities we require 2 -4% partial melt oriented in sills. Along rift, horizontally aligned radial anisotropy (VSH > VSV )gets progressively weaker northwards until VSV > VSH, suggesting melt reorients from sills to dykes as rifting progresses. This thesis indicates there can be significant melt accumulation in the crust both on and off rift. Furthermore, melt supply starts early in the breakup process which in turn informs our understanding of how magma assists rifting