Geomagnetic secular variation

Abstract We analyse over 175000 magnetic observations from an interval spanning 1695-1980 to produce a sequence of maps of the magnetic field at the core-mantle boundary; we find that even the earlier data enable us to determine reliable maps. We produce these maps at approximately 60-year intervals through the eighteenth and nineteenth centuries, and at 10-year intervals in the twentieth century. This span of maps is long enough to render straightforward the distinction between static and drifting features in the field: we observe that some features show no sign whatsoever of drift over the entire 285-year time interval, although others drift westwards. In particular, we observe that the secular variation is very low beneath the Pacific ocean, but beneath southern Africa and the South Atlantic ocean we observe rapid secular variation. We interpret the morphology of the static field in terms of a simple model of the dynamo, and conjecture that interactions between the core and the mantle are an important element of the process. As part of the static field we identify four main concentrations of flux, two in each hemisphere, at high latitudes: these features largely account for the Earth’s axial dipole moment. We find unequivocal evidence that magnetic flux has not remained frozen over the time span of our models; much of the diffusive behaviour that we identify is associated with the formation of a pair of flux spots (a ‘core spot’) beneath southern Africa, early in this century. Nevertheless, we are able to construct maps that satisfy a set of necessary conditions for frozen-flux, and use these maps to construct maps of the core surface fluid flow, based on the steady flow hypothesis. Although we find no strong evidence against the steady flow hypothesis, we do find some grounds on which to doubt the validity of the flow maps.