Influence of turbulence on geomagnetic morphology and time variations

Principal Investigator: Prof  D Gubbins

Co-Investigator: Prof C Jones

Sponsor: Leverhulme

Value: £112252

Dates: 1st January 2007 to 31st December 2010

Summary

Molecular diffusivities are completely unreachable in numerical simulations, and are almost certainly irrelevant because of the effects of turbulence that are not modelled directly. Simple models of turbulence based largely on shear flows may be inappropriate for buoyancy-driven convection in the presence of a self-generating magnetic field, but in any event they leave a numerical problem that will remain intractable for some time to come. Furthermore, in some respects moving towards what appear at first sight to be more realistic values of the parameters produces dynamos that bear no resemblance to that of the Earth and could not operate within the core.

In some respects simple regimes produce more realistic magnetic fields: low Rayleigh number, low inertia, and high Roberts number all produce realistic fields. Lowering the Ekman number may well change this, but clearly we do not understand the subgrid-scale processes properly. The way forward is to study and understand the effects of small-scale turbulence and the effects of different models of turbulence on the large scale dynamo solutions. We are much encouraged by the good agreement, in places, with observation within the simple regime: this is a tractable problem now.