GP34B-01 INVITED
On the Trail of Bullard's Dynamo
E. C. Bullard proposed the first deterministic models of the geodynamo more than five decades ago, combining low order spherical harmonic degree poloidal flows in the outer core suggested by the geomagnetic field structure on the core-mantle boundary with an axisymmetric zonal toroidal flow implied by the geomagnetic westward drift. Although most of the original Bullard dynamo models were kinematic failures, their modern dynamical counterparts have proven notably successful in accounting for many characteristics of the geomagnetic and paleomagnetic fields, including their dipole dominance, the time average axial state, preference for westward drift, low frequency secular variation, and the existence of polarity reversals and excursions. However, theoretical considerations supported by some recent numerical and laboratory dynamos indicate that small scale convection and turbulence may be more important than large scale flow in the dynamo generation process because the outer core is in a high Rayleigh number, low Ekman number dynamical regime. In this talk the properties of large scale laminar versus small scale turbulent numerical dynamos are compared. Evidence is presented that large scale flows similar to those envisioned by Bullard are an important part the geodynamo in spite of the turbulent state of the core, because of the non-uniform heat flow from the core into the heterogeneous lower mantle. Dynamo models and laboratory experiments show that non-uniform thermal coupling with the lower mantle generates large scale flows in the outer core that affect the symmetry of the geomagnetic field and may cause non-uniform growth of the inner core. Dynamo models also indicate that time dependence of core-mantle thermal coupling influences the frequency of polarity reversals and the timing of superchrons.