The Effects of Chemical Density Differences on Convective Mixing in the Earth ' S Mantle

Subducted lithosphere and crust are chem1cally d1stinct from amb1ent mantle, and some have hypoth.,sized that the result1ng anomdlous buoyancy may affect m1x1ng and, in particular, may CciUSe chem1cal heterogeneity topers ist. To explore th1s effect, two-dimensional convective mix1ng calculations have been carried out. Boundary cond1tions m1micking features of plate kinematics have been imposed and negatively buoyant tracers placed beneath converging plate margins to simulate subduct1on. The principal parameter of interest is the density anomaly ratio: l!p/(p 0 a l!T), the rat1o of the chemical to the thermal buoyancy. For a density anomaly ratio appropriate to complete separation of the oceanic crust from the rest of the 1 i thosphere, it is marginally possible for the tracers to segregate into the warm region near the bot tom and neutrall ze the thermal buoyancy there, for internally heated flows, and to segregate to the base of the uprising limbs of convection, for bottom heated flows. However, for the more likely scenario of oceanic crust and lithosphere maintaining their layering after subduction, there is neither segregation nor neutralization of ther!l)al buoyancy. The average residence t1me is the average time tracers remain in the box from subduction to sampling beneath r1dges and could be increased by 4% from the passive case at the very most if the crust and lithosphere remain layered. Even in the more extreme case of separation of crust from lithosphere, the residence times are only increased by 15% from the passive case.