Comparison of the Melt Distribution in Natural Analogues to Experimentally Produced Microstructures

The generation of melt in the Earth's shallow mantle, and the segregation and eruption of this melt at the Earth's surface as basalt is the main process by which the Earth continues to evolve. During melting low melting point components are separated from more refractory components, which remain behind in the mantle. The low melting point components transported to the surface with the basaltic melt include volatiles such as water and CO2. Together with the process of subduction, this exchange of the interior with the surface has shaped the composition of the atmosphere over the life of the Earth. This process also constitutes the deepest part of the plumbing systems of volcanoes in a range of tectonic settings. Understanding deep melt migration is therefore an integral part of understanding the dynamics of volcanoes.

Processes in the mantle such as the generation of melt can not be observed directly, but are interrogated experimentally as well as by studying natural occurrences of mantle rocks at the Earth's surface. The natural rocks that will be examined for this project are relatively rare in terms of their microstructure and composition. Preliminary observations indicate that a basaltic melt froze in place in rocks of mantle origin, but with an interstitial melt geometry normally observed only in experiments. Thus these rocks form an important link between experiments, direct observations in the field and indirect methods of observation such as seismic imaging of partially molten regions. This project is intended to help validate the application of experimental results on small samples at upper mantle conditions to the different scales in space and time of natural systems, leading to a better understanding the process of melt generation and transport. To this end we will examine thin sections of olivine-rich troctolites by optical and scanning electron microscopy, and collect electron-backscatter diffraction maps. Together these maps and images will be processed in the same way that images from experimentally produced samples are processed to quantify the melt distribution on the grain scale. Additional analysis by microprobe and laser-ablation mass spectrometry will help illuminate the processes that formed the microstructures.