The PI will carry out a numerical study in realistic 3D geometry of the thermomechanical interaction of mantle plumes moving lithospheric plates. The model consists of a rectangular box containing fluid with variable viscosity. he lithosphere is represented by highly viscous fluid at the top of the box that moves with velocity Uo, and the plume is generated by imposing a Gaussian-shaped temperature anomaly on the bottom of the box. Numerical solutions for the flow will be obtained using the hybrid spectral-finite difference method of Christensen and Harder (1991). The solutions will then be used to calculate topography, geoid, and heat flow anomalies, melt production rates, and the chemical and isotopic composition of erupted melts. The models will be constrained by comparing these parameters with observations from various hotspots, with special attention to Hawaii. Questions include the origin of midplate swells (thermal vs. dynamical support), the effectiveness of lithospheric thinning, and the origin of the four-stage evolutionary pattern of Hawaiian volcanism.
