The Excitation of True Polar Wandering by Extreme Earthquakes over Time

The recent work by Kagan and Jackson (2013) has awakened us to the possibility that there may be five M9 earthquake occurring per century for a long time. The co-seismic inertia perturbation associated to global seismicity from 1977 to 2013 (CMT earthquake catalogue) yields a total displacement of the rotation axis towards the Pacific ocean of about 35 cm, corresponding to almost 1 cm/yr. This significant result, although smaller by one order of magnitude compared to the observed True Polar Wander (TPW) velocity of 10 cm/yr, has motivated us to refine the simulations including the viscoelastic relaxation of the Planet. The viscoelastic relaxation, indeed, is responsible for the rotational instability of the Earth on geological timescales, the hydrostatic bulge being able to readjust to new positions of the rotation axis, and amplifies the earthquake-induced inertia perturbations during the transient phase. In this vein, we develop a time-dependent model of global seismicity on the million years timescale based on (i) plate reconstruction data from, here we have employed G-plates, (ii) the momentum conservation principle relating the seismic moment rate at plate boundaries to the relative velocity between plates and (iii) using the Anderson's theory of faulting. Simulations based on this time-dependent model well agree with the result from the earthquake catalogue in the elastic limit and show that the TPW excursion may reach several degrees and that TPW velocity are only triggered by changes in global seismicity through the million year timescale. Key parameters of our simulations are (a) the viscosity of the lithosphere which controls the timescales at which each earthquake still perturbs the inertia of the planet, (b) the stabilizing rotational effect associated to the non-hydrostatic bulge from mantle density anomalies and (c) the nature of the major density discontinuities in the mantle.