Temperature Gradients and the Convective Velocity in the Earth's Core

Abstract

Investigation of the methods used to calculate the adiabatic temperature gradient in the fluid core of the Earth leads to the conclusion that none of them is valid. This is because they all, in effect, apply to the liquid core theory that is strictly applicable to solids only, without providing any justification for such a step. Thus claims based on such calculations that the core is stable against convection cannot be accepted. Simple dynamical arguments are then used to obtain an estimate for the mean vertical component of the convective velocity, on the assumption that the core is undergoing large-scale convection. The figure obtained is 3 x low4 m s-', about twice previous estimates which were based on the westward drift of the geomagnetic secular variation.