The purpose of the proposed work is to understand how to use the chemical variation of lavas erupted at the Earth's surface to provide better insight into the physical processes occuring in the deep Earth. We will use dynamic computer models that calculate the coupled flow of melt and mantle to explore the sensitivity of observable geochemistry to changes in mantle processes. At the same time we will compile reference geochemical and geophysical data-sets to compare with and test the models. Through a combination of theory and observation, this work may give much stronger constraints on deep Earth processes than either classical geodynamics or geochemistry alone.
