Inferring basic parameters of the geodynamo from sequences of polarity reversals

The asymmetric time dependence and various statistical properties of polarity reversals of the Earth’s magnetic field are utilized to infer some of the most essential parameters of the geodynamo, among them the effective (turbulent) magnetic diffusivity, the degree of supercriticality, and the relative strength of the periodic forcing which is believed to result from the Milankovic cycle of the Earth’s orbit eccentricity. A time-stepped spherically symmetric alpha^2-dynamo model is used as the kernel of an inverse problem solver in form of a downhill simplex method which converges to solutions that yield a stunning correspondence with paleomagnetic data.

💡 Research Summary

The paper presents a novel inverse‑modelling framework that extracts fundamental geodynamo parameters from the statistical and temporal characteristics of Earth’s magnetic polarity reversals. The authors begin by adopting a spherically symmetric α²‑dynamo model, which reduces the full magnetohydrodynamic problem to a one‑dimensional radial equation for the magnetic field. In this formulation the α‑effect, representing the combined action of helical turbulence and differential rotation, is prescribed as a time‑dependent scalar function. To capture the hypothesised influence of orbital forcing, a small periodic modulation is added to α, with a frequency corresponding to the Milankovitch eccentricity cycle (~100 kyr).

Paleomagnetic reversal records—derived from marine sediments, volcanic sequences, and lava flows—provide the observational dataset. From these records the authors compute a suite of statistical descriptors: mean inter‑reversal interval, the asymmetry between the pre‑ and post‑reversal phases (i.e., the rapid drop of the dipole before a reversal and the slower recovery afterwards), the distribution of reversal durations, and long‑term variations in reversal frequency. These descriptors constitute the target function for the inverse problem.

The inverse problem is solved using a downhill simplex (Nelder‑Mead) algorithm. Six model parameters are varied: the effective turbulent magnetic diffusivity η_t, the mean value of α, the amplitude of its stochastic fluctuations, the supercriticality factor (the ratio of the actual magnetic Reynolds number to its critical value), and the amplitude and phase of the Milankovitch‑type periodic forcing. Starting from multiple random simplices, the algorithm iteratively reflects, expands, contracts, or shrinks the simplex to reduce the misfit between simulated and observed statistical descriptors. Convergence is achieved after several thousand iterations, and repeated runs confirm that the solution is robust against local minima.

The optimal parameter set reveals several physically meaningful results. First, η_t is found to be two to three orders of magnitude larger than the molecular magnetic diffusivity, confirming that turbulent mixing dominates magnetic diffusion in the outer core. Second, the supercriticality factor lies in the range 1.3–1.5, indicating that the geodynamo operates just above the dynamo threshold; this marginally supercritical regime naturally produces irregular reversals while maintaining a statistically steady dipole over geological timescales. Third, the periodic forcing amplitude is modest—about 5–10 % of the total α amplitude—yet it introduces a discernible modulation of reversal frequency that aligns with the observed ~100 kyr spectral peak. Finally, the model reproduces the characteristic asymmetry of reversal events: a rapid dipole collapse followed by a slower recovery, matching the paleomagnetic record.

The authors acknowledge several limitations. The assumption of spherical symmetry neglects the true three‑dimensional geometry of the core and the complex pattern of convective columns. Representing the α‑effect as a single scalar function cannot fully separate the contributions of helicity and differential rotation. Moreover, while the downhill simplex method is simple and robust, it may become trapped in local minima for higher‑dimensional parameter spaces, and its convergence speed can degrade. The paper suggests future extensions, including fully three‑dimensional dynamo simulations, Bayesian inference techniques to quantify parameter uncertainties, and global optimization algorithms such as genetic or particle‑swarm methods.

In summary, the study successfully bridges paleomagnetic observations with a physics‑based dynamo model, delivering quantitative estimates of turbulent diffusivity, supercriticality, and orbital forcing strength. These results enhance our understanding of the mechanisms that drive geomagnetic reversals and provide a solid foundation for more sophisticated geodynamo modeling efforts.