Consequences of Energy Dissipation by Creep in the Mantle

Abstract

By relating the energy release in creep processes to the observed geothermal flux, we impose a restraint on the supposed strength of a convecting mantle, which is more easily accommodated if convection is confined to the upper mantle. However, independently of any yield point which could reasonably be assumed, stress wave attenuation is not significantly influenced by the superposition of convective creep. The most hopeful seismic test for reasonableness of the convection hypothesis is therefore not attenuation but rather anisotropy of seismic velocity. The energy of the Chandler wobble is not related to the stresses which it causes in the same way as energy and stress are related in seismic waves. This invalidates the direct comparison of Q factors of the wobble and of shorter period stress waves. Thermal diffusion being slow, large bodies of rock behave adiabatically and heat generated by creep in a weak zone makes it selectively weaker. This leads to the expectation that zones of high heat flow of order 10 km across will result from local concentrations of convective creep.