Estimates of true polar wander since 300Ma.

We investigate true polar wander (TPW) since the Early Permian. We first construct a hotspot reference frame (under the assumption of hotspot fixity) using: (1) a compilation of published hotspot reference frames from 0 to 120Ma, (2) an updated list of active hotspots with improved criteria aimed at detecting their depth origin, (3) a compilation of terrestrial volcanic data suspected to reveal hotspot activity, and (4) a set of plate reconstructions, based initially on paleomagnetism corrected with respect to hotspots. We derive a hotspot Apparent Polar Wander path (APWP) with respect to South Africa which, combined with a global paleomagnetic APWP, allows us to estimate mantle TPW. The computed mantle curve is characterized by slow TPW between the present and 50 Ma, a cusp followed by a polar wander of nearly opposite direction till 120-130Ma. Despite some minor differences, similar trends from several published TPW curves may indicate that former determinations from present back to 120 Ma are real features of TPW. Then a new cusp precedes a period of fast drift marked by a long track between 130 and 180 Ma. Then a change of direction occur at around 200Ma, when a new change in direction occurs in the opposite direction. The APWP curve during the period t=[0-180] Ma grossly aligns along a great circle which pole latitude is located close to the equator and longitude at around 50°E, while the interval t=[180 to 280] Ma, is marked by a different great circle, the pole of which is also located close to the equator with a longitude different by some 50°. In both case, we find that the pole describing the mantle motion is situated close to the minimum Principal Axis of Inertia (PIA), itself positioned close to an axis passing through the Large Low Shear Velocity Provinces (LLSVPs), and close to the maximum degree 2 geoid high under Africa. We then compare our data with a model which compute the temporal evolution of the Principal Inertia Axis (PIA). The observed cusps are also present in the model, while the minimum PIA is also situated close to the equator, with a dispersion in longitude of some 45° not significantly different from that of the observed poles. We discuss the sudden changes of TPW direction (i.e., cusps), in term of mass reorganizations within the mantle linked to changes in subductions.