Testing for Hemispheric Variability in the Earth's Magnetic Field Over Geologic Timescales /
- Author(s): Cromwell, Geoffrey John Deustua
- et al.
The Earth's magnetic field is generally assumed to approximate a geocentric axial dipole (GAD) field when averaged over sufficient time (10⁵-10⁶ yrs). In a GAD field, magnetic declination is zero at every point on the Earth's surface while magnetic inclination and intensity vary predictably with latitude. Departures from GAD are observed in the present field and in paleomagnetic measurements over historical and archaeological timescales. Early compilations of paleodirectional and paleointensity data for the last few million years support the GAD hypothesis with only minor deviations. Most paleomagnetic data are from mid-latitudes in the northern hemisphere, producing a geographic sampling bias that is not accounted for in existing paleomagnetic compilations and geomagnetic field models. Until recently there has not been sufficient global coverage of high-quality paleomagnetic data to adequately evaluate latitudinal field structures across the globe. In this dissertation we add new high-quality paleomagnetic sites from the Arctic and document data quality issues that have affected the development of paleosecular variation (PSV) and time- averaged field (TAF) models. We compare our new Arctic directional results to the existing Antarctic data and a new global compilation of published directional sites. We observe a greater variance of paleodirections in the southern hemisphere, especially at high latitudes, relative to northern hemisphere sites and find that time- averaged geomagnetic field models with a small axial- quadrupole (g⁰/₂) component fit the global data set better than GAD models. Accurate paleointensity estimates of the ancient and modern field are difficult to obtain, but we show that historic Hawaiian volcanic glasses can reliably recover the expected field strength when using appropriate statistical controls. We use this methodology to acquire paleointensity data from Iceland for the last 780 ka, which are consistent with long-term estimates of geomagnetic field strength. Our analysis of high-quality paleodirectional data from the global data set and accurate paleointensity results from the Arctic suggest different geomagnetic field structures between northern and southern hemispheres which are not yet predicted by existing field models