The upper few hundred kilometres of the earth is comprised of a jigsaw of tectonic plates that separate, collide, or slide past each other at plate boundaries in a process called plate tectonics. Separation of a plate along a boundary located within a continent can result in the division of the continent into two or more parts and eventual formation and growth of a new ocean basin. This is achieved through the addition of new material to the plate, sourced from partially molten magma deep below the surface, and added to the upper crust as near vertical walls of frozen magma called dykes. Therefore, the process by which dykes are emplaced at separating plate boundaries is fundamental to plate tectonics. However, the majority of such plate boundaries lie at the bottom of the Earth's major oceans, and obscured by several kilometers of overlying water. Due to a lack of observations, scientists do not know accurate details of how, where and when dyking actually occurs during plate separation. A tiny proportion of the boundaries between separating plates where dykes occurs are exposed above sea-level. Examples include Iceland, and Afar (Djibouti and Ethiopia). Previous observations show that dyke intrusion does not occur on a continuous basis, but is rather restricted to week, month, or year long episodes of repeated dyke intrusions that occur every couple thousand years. Dyking is generally accompanied by violent earthquakes, fracturing of the Earth's surface, and volcanic eruptions. However, only two dyke intrusions have been observed over the last century, and none since the advent of modern scientific instrumentation. We are therefore still not sure where the source of the molten rock that feeds dykes is located, and how and when molten rock moves within the earth during the formation of a dyke. These fundamental questions can only be answered by combining accurate observations of earthquakes with high resolution measurements of actual plate movements provided by modern global positioning systems and space borne satellites. In September 2005, the intrusion of a 60 km-long basalt dyke, up to 8 metres thick, occurred in the Afar rift, Ethiopia. The event was accompanied by a volcanic eruption and swarm of violent earthquakes over two weeks. I participated in an urgent response to deploy recording equipment in Afar to accurately measure earthquakes, the expansion and contraction of magma chambers beneath the active volcanoes, and measure the movement of the tectonic plates, in the knowledge that future dykes are likely to be emplaced. The extreme effort to acquire this unique data was rewarded in June 2006 when a second dyke (15 km long, 2 metres wide) was injected beneath the rift. I thus now have access to extraordinary data from a once in a generation dyke intrusion event to accurately probe how, when and where dyking occurs at separating plate boundaries. I will be able to accurately determine where the magma came from, when the magma moved within the earth, and how much magma was frozen into the plate as a vertical dyke. The results of my proposed study are of critical importance in improving our understanding of the general process of plate separation. In collaboration with first class scientists from Ethiopia, Iceland, USA and the U.K., I will learn new skills in processing, analyzing and interpreting scientific data while solving a fundamental question on how the Earth continues to evolve. The results of my proposed research will be invaluable to the Ethiopian government in understanding the volcanic and seismic hazards, and potential sources of geothermal energy associated with the spectacular geological features in Afar.