The investigators will study the effect of pressure and temperature-dependent density and viscosity on the development of convective instabilities in dense lower-crustal mafic and ultramafic cumulates using a two-dimensional finite element model. Previous work has shown that with an assumption of constant density and viscosity for the lower crust and mantle, Moho temperatures must be ``hot'' (e.g. >600 C at 1 GPa) in order for dense lower crust to become convectively unstable and sink into the mantle on short timescales (~10 Myr). Variable density and viscosity will allow the investigators to consider (1) viscous entrainment of crustal material during the development of an instability and (2) time-varying crustal geotherms. They propose that (1) could play an important role in initiating subduction at passive margins, where a slowly growing lower crustal root exerts a negative buoyancy force on the plate margin. They further propose that (2) can explain the topographic and magmatic evolution of the Tibetan Plateau, where an initially dense, stable, lower crustal root becomes convectively unstable and sinks into the mantle as radiogenic heating decreases the viscosity of the lower crust and lithosphere.
