S. P. Lund $^{a}$ and C. Constable $^{b}$
$^{a}$ University of Southern California, Los Angeles, USA. $^{b}$ University of California - San Diego, La Jolla, USA.
Spherical harmonic analyses of historical and paleomagnetic secular variation (PSV) data for the last 3000 years (Constable et al., Phil. Trans. Royal Soc. v.358, p.991, 2000) have been used to reconstruct the positions and movement of high-latitude flux lobes in the Northern Hemisphere. The data indicate that two flux lobes are present persistently, but that as many as four lobes may be present for a few hundred years. The flux lobes move systematically on a millennial time scale but only within limited longitude bands, primarily centered on North America and central Eurasia. However, unlike historical field observations, the prehistoric flux lobes sometimes occur within the tangent cylinder at latitudes greater than 60 degrees N. Flux lobes move consistently in a clockwise manner between 1000 BC - 200 BC and 1000 AD - 1600 AD, but in a counterclockwise manner from 200 AD - 1000 AD. A simple mathematical simulation model using sets of axis-parallel lines of dipoles (analog for main field composed of convection rolls), which move as the flux lobes, can account, to first order, for most of the observed Northern Hemisphere PSV for the last 3000 years. However, this model is unrealistic as an analog for convection rolls in that the lines of dipoles sometimes occur within the tangent cylinder. An alternative mathematical simulation model based on a centered dipole (main field) plus dynamo waves (regional fields; Olson and Hagee, Geophys. J., v.88, p.139, 1987; Hagee and Olson, PEPI, v.56, p.266, 1989) can also explain much of the observed PSV for the last 3000 years. Both models were used to attempt to simulate the dynamics of magnetic field behavior during the Laschamp Excursion (40,000 YBP), a time of anomalous PSV characterized by two millennial-scale cycles of large-amplitude directional looping. Only the dynamo wave model was able to generally reproduce the Laschamp Excursion directional and paleointensity variability. We hypothesize that, if we combine the dynamo wave model (regional fields) with the convection roll model (main field), we can account for most PSV in the Northern Hemisphere for the last 3000 years and keep the convection rolls located outside of the tangent cylinder. Most of the apparent high latitude flux lobe motion would then be due to dynamo wave propagation.