The geodynamo as a random walker: A view on reversal statistics

Abstract

[1] The sequence of durations of the geomagnetic polarity intervals is often described in terms of a nonhomogenous Poisson process with time-dependent reversal rate, reflecting the nonstationarity of the underlying geodynamo process. This view has recently been challenged, and here we show that the first-passage time statistics of random walks taking place on a flat potential relief yields a much more consistent interpretation of the distribution of geomagnetic polarity intervals. A possible physical explanation suggests that the stability of a polarity chron of the Earth's magnetic field is controlled by a sum of statistically independent, randomly behaving, dynamo processes in the outer core. The random-walk hypothesis naturally includes the observed occurrences of very long superchrons, and it provides a new, considerably longer statistical estimate for the total duration of the present Brunhes chron predicting that the probability for a geomagnetic field reversal within the next 30 Ka is ≈5% and the probability that we live in a chron longer than 2 Ma is about 30%.Citation: Shcherbakov, V., and K. Fabian (2012), The geodynamo as a random walker: A view on reversal statistics, J. Geophys.

Variability of Millennial‐Scale Trends in the Geomagnetic Axial Dipole

“…A remarkably similar prediction was made using a solution of the backward Fokker‐Planck equation (Buffett & Davis, ). This agreement may be surprising at first glance because the solution in has the form of a purely diffusive process with no dependence on the drift term (e.g., Shcherbakov & Fabian, ). While the drift term appears in the backward Fokker‐Planck equation, its contribution to the predicted probability is small when the solution is evolved over a short time interval.…”