Fluid Motion and the Toroidal Magnetic Field Near the Top of Earth's Liquid Outer Core.

This work considers two unresolved problems central to the study of Earth's deep interior: (1) What is the surface flow of the complete three dimensional motion sustaining the geomagnetic field in the fluid outer core? (2) How strong is the toroidal component of that field just beneath the mantle inside the core?. A solution of these problems is necessary to achieve even a basic understanding of magnetic field generation, and core-mantle interactions. Progress in solving (1) is made by extending previous attempts to resolve the core surface flow, and identifying obstacles which lead to distorted solutions. The extension relaxes the steady motions constraint. This permits more realistic solutions which should resemble more closely the real Earth flow. A difficulty with the assumption of steady flow is that if the real motion is unsteady, as it is likely to be, then steady models will suffer from aliasing. Aliased solutions can be highly corrupted. The effects of aliasing incurred through model underparametrization are explored. It is found that flow spectral energy must fall rapidly with increasing degree to escape aliasing's distortion. Damping does not appear to remedy the problem, but in fact obscures it by forcing the solution to converge upon a single, but possibly still aliased estimate. Inversions of a magnetic field model for unsteady motions, indicate steady flows are indeed aliased in time. By comparison, unsteady flows appear free of aliasing and show significant temporal variation, changing by about 30% of their magnitude over 20 years. However, it appears that noise in the high degree secular variation (SV) data used to determine the flow acts as a further impediment to solving (1). Damping is shown to be effective in removing noise, but only once aliasing is no longer a factor and noise is restricted to that part of the SV which makes only a small contribution to the solution. To solve (2) the radial component of Ohm's law is inverted for the toroidal field (B_{T }) near the top of the corp. The flow, obtained as a solution to (1), is treated as a known quantity, as is the poloidal field. Solutions are sought which minimize the difference between observed and predicted poloidal main field at Earth's surface. As in problem (1), aliasing in space and time stand as potential impediments to good resolution of the toroidal field. Steady degree 10 models of B_{T} are obtained which display convergence in space and time without damping. Poloidal field noise, as well as sensitivity to the flow model used in the inversions, limit resolution of toroidal field geometry. Nevertheless, estimates indicate the magnitude of B_{T } does not exceed 8times 10^ {-5}T, or about half that of the poloidal field near the core surface. Such a low value favors weak -field dynamo models but does not necessarily endorse a geostrophic force balance just beneath the mantle because partial_{r}B _{T} may be large enough to violate conditions required by geostrophy.