True polar wander

A re-evaluation of the existence of true polar wander (TPW) since the Late Cretaceous and a comparison among the various approaches are made using updated paleomagnetic, hotspot and relative motion datasets. Previous attempts to determine the existence of TPW had resulted in different conclusions: comparison of hotspot locations and paleomagnetic poles required significant pole motion, although lithospheric plate displacement analysis yielded insignificant motion. However, these earlier determinations cannot be directly compared to find the reason for the discrepancies, because each used different datasets. For this study the different approaches are applied to a single updated model with three alternative relative motions of East and West Antarctica. Although the results are model-dependent, in general there was not significant motion of the pole relative to the lithosphere (1-5°) since the early Tertiary, but a large motion (10-12°) relative to the hotspot framework. It is unlikely that errors in the determinations could account for this disagreement: the A₉₅ of the plate reconstruction is about 3°, the uncertainty in Antarctica motion is estimated to no larger than 3°, and cumulative errors in the relative plate motions may also amount to 3°. Only if all these errors are present in the maximum estimated amount, and in the same direction, could they account for the 10-12° gap between the two approaches. This conclusion of pole motion relative to the hotspots, but not the lithosphere, may indicate an independent shift of the mesosphere relative to the lithosphere (or "mantle roll" of Hargraves and Duncan).