A Paleointensity-Based Test of the Geocentric Axial Dipole (GAD) Hypothesis

The GAD model is central to many aspects of geophysics, including plate tectonics and paleoclimate. However, significant departures from a GAD field over geologic time have not been ruled out, particularly for the Precambrian. Here, we investigate a test of the GAD model using published paleointensity data. Our goals are to determine if paleointensities can shed light on the validity of the GAD model, and hence to see if they provide constraints on the evolution of the geodynamo throughout earth history. Using numerical dynamo models, we show that intensity distributions can be fairly well characterized by the first three zonal Gauss coefficients (dipole, quadrupole and octupole), although time-averaging tends to broaden the range of intensities. The dynamo models indicate that the ancient core, prior to nucleation of the inner core, may have had a significant (up to 10%) contribution of the zonal octupole. We then investigate the connection between the measured paleointensities assembled in the PINT database and the GAD model by means of predicted theoretical frequency distributions for various simple models (GAD, GAD ± small zonal quadrupole or octupole components). Hitherto, paleointensities have often been analysed in terms of corresponding virtual dipole moments (VDMs). But this rather begs the question because a GAD model is assumed in order to derive a VDM. By using raw field values reported from each sampling site we eliminate dependence on the GAD hypothesis. We find that models consisting of one or two different GADs cannot explain the data, but 3- or 4-GAD models can fit the data surprisingly well, and adding a ±5% octupole significantly improves the fit.