On the genesis of the Earth's magnetism.

Few areas of geophysics are today progressing as rapidly as basic geomagnetism, which seeks to understand the origin of the Earth's magnetism. Data about the present geomagnetic field pours in from orbiting satellites, and supplements the ever growing body of information about the field in the remote past, derived from the magnetism of rocks. The first of the three parts of this review summarizes the available geomagnetic data and makes significant inferences about the large scale structure of the geomagnetic field at the surface of the Earth's electrically conducting fluid core, within which the field originates. In it, we recognize the first major obstacle to progress: because of the Earth's mantle, only the broad, slowly varying features of the magnetic field within the core can be directly observed. The second (and main) part of the review commences with the geodynamo hypothesis: the geomagnetic field is induced by core flow as a self-excited dynamo. Its electrodynamics define 'kinematic dynamo theory'. Key processes involving the motion of magnetic field lines, their diffusion through the conducting fluid, and their reconnection are described in detail. Four kinematic models are presented that are basic to a later section on successful dynamo experiments. The fluid dynamics of the core is considered next, the fluid being driven into motion by buoyancy created by the cooling of the Earth from its primordial state. The resulting flow is strongly affected by the rotation of the Earth and by the Lorentz force, which alters fluid motion by the interaction of the electric current and magnetic field. A section on 'magnetohydrodynamic (MHD) dynamo theory' is devoted to this rotating magnetoconvection. Theoretical treatment of the MHD responsible for geomagnetism culminates with numerical solutions of its governing equations. These simulations help overcome the first major obstacle to progress, but quickly meet the second: the dynamics of Earth's core are too complex, and operate across time and length scales too broad to be captured by any single laboratory experiment, or resolved on present-day computers. The geophysical relevance of the experiments and simulations is therefore called into question. Speculation about what may happen when computational power is eventually able to resolve core dynamics is given considerable attention. The final part of the review is a postscript to the earlier sections. It reflects on the problems that geodynamo theory will have to solve in the future, particularly those that core turbulence presents.