Mapping the pattern of convection beneath the plates is central to our understanding of the mode of heat transfer within the earth, the chemical differentiation and evolution of the mantle, the viscosity structure of the mantle, and the forces that drive plate tectonics. Much progress has recently been made toward directly mapping the pattern through the effects of convection on seismic velocities. Two types of velocity variations are expected; lateral heterogeneities related to composition or temperature, and anisotropy caused by fabric or crystal alignment associated with deformation during flow. The purpose of the research proposed here is to map velocity structures in the upper mantle related to convection beneath the plates, with particular emphases on testing hypotheses of the existence of anisotropy in specific regions and on mapping small- scale (wavelengths 500 to 3000 km) features using phases that sample the earth in ways that provide high lateral resolution in otherwise inaccessible areas.
