Late Mesozoic and Cenozoic palaeomagnetism of Australia–II. Implications for geomagnetism and true polar wander

Summary. New palaeomagnetic results from Australia indicate that throughout the Cenozoic era the continent lay further south than suggested by hot-spot data. Moreover, while hot spots give a uniform rate of drift during most of the Cenozoic, the drift rate obtained from apparent polar wander varies considerably.
The discrepancies between the palaeomagnetic and hot-spot results are analysed by comparing the Australian data with those of Europe and the central Pacific. The analysis suggests that the discrepancies are due to: (1) departures of the Earth's magnetic field from the geocentric axial dipole model, and (2), either true polar wander or a non-axial inclined dipole component. It is found that since the mid-Tertiary the dominant non-dipole component has been a quadrupole, and that during this period both the quadrupolar field and the true polar displacement/non-axial dipole component decreased progressively. During the Quaternary, and also at the earliest Tertiary, the non-dipole components appear to have been moderate or small.
The comparison of data sets demonstrates that considerable errors may be incurred when Cenozoic, and presumably earlier, poles from one geographic region are used to derive those of another, widely separated, region. The results also imply that absolute plate velocities estimated from palaeomagnetic data can contain substantial errors, and that hot-spot data may need significant adjustments for true polar wander to yield correct palaeolatitudes.
Finally, the new early Tertiary pole for Australia is used in conjunction with updated early Tertiary poles from other lithospheric plates to reapply the McKenzie test for true polar wander. The results indicate a small true polar displacement since the beginning of the Tertiary. The amount and direction of the displacement, however, differ from those generally obtained from hot-spot data.