Convection at the melting point: A thermal history of the earth's core

Abstract

Higgins and Kennedy (1971) concluded that the Earth's fluid core has a stable stratification if it is at its melting point. Busse (1972) and Elsasser suggested as an alternative that a hydrostatic-isentropic distribution of particulate solid can produce neutral stability in a partially molten core. Here this suggestion is quantified and a determination is made of the efficiency of the production of fluid motion from the heat flux. This is used to establish that macroscopic convection can exist only if the particulate solid is of sufficiently small size. A thermal history of the core compatible with upper mantle heat flux is advanced in which it is suggested that the inner core is a fairly recent feature. The implication of these results for convection-driven and precession-driven dynamos is that both can function for a small enough suspended particulate, that the convection-dynamo will fail for particles greater than one micron in diameter, and that the precession-driven dynamo probably cannot survive particles greater than ten microns in diameter.