Planetary dynamos driven by semiconvection in stably stratified layers

TL;DR

This study demonstrates that semiconvection within stably stratified planetary layers can sustain self-generated magnetic fields in a rapidly rotating, low magnetic Prandtl number regime. Using direct numerical simulations in a rotating spherical shell, the authors show a layered convection state with an inner convective region under a stable layer can drive a self-consistent dynamo, producing dipolar fields and spectral characteristics similar to Jupiter's. The results provide the first direct numerical evidence that semiconvection can power dynamos in such regions, offering a potential mechanism to explain observed planetary magnetic fields and guiding future exploration with more realistic planetary parameters. This work has broad implications for astrophysical magnetic field generation in gas giants, stellar interiors, and exoplanetary contexts.

Abstract

Stably stratified fluid layers are common in gaseous planets, stellar interiors, and planetary cores, and have long been considered incapable of sustaining dynamo action. Here, we show that semiconvection - driven by a destabilizing thermal gradient within an overall stably stratified medium - can, in fact, give rise to self-sustained magnetic fields. Motivated by recent models suggesting that large portions of Jupiter and Saturn may be semiconvective, we perform direct numerical simulations in spherical shells, operating in the planetary-relevant regime of low magnetic Prandtl numbers. From a primary semiconvection instability, a layered convection state spontaneously develops, consisting of a convective region beneath a stably stratified layer of comparable thickness. Fluid motions in this convective region are strong enough to produce magnetic fields with key features observed in planetary dynamos, including strong dipolarity, realistic field strengths, and spectral characteristics. These results provide the first direct evidence that semiconvection can drive dynamo action in stably stratified regions of gas giants and stellar interiors, with important implications for understanding astrophysical magnetic field generation.

Figures (7)

• Figure 1: Outline of the physical setup of the problem.
• Figure 2: Stability of dynamo action in ${\rm{Ek}}$--${\rm{R}}_m$ space, for a range of simulations, with varying values of ${\rm{Ra}}_C$, ${\rm{Ra}}_T$, ${\rm{Pm}}$ and $Ek$. Dots/crosses represent simulations showing growth/decay in the magnetic energy, coloured by the value of ${\rm{Pm}}$. The black dashed line is an estimate of the critical value ${\rm{R}}_m^C({\rm{Ek}})$. The red circle identifies the position of the simulation presented in detail in Sect. \ref{['sec:nonlinearsim']}. The ranges of parameter values explored are given in Table \ref{['tab:simparams']}.
• Figure 3: Initial condition for the main simulation, showing the axisymmetric part $N^2_\text{axi}$ of the square of the Brunt Väisälä frequency (a) in a meridional slice and (b) along the transect shown in red in (a). The black dashed line shows the background conductive state. (c)--(f) Meridional and equatorial slices of the radial and azimuthal velocity fields $u_r$ and $u_\phi$. Solid black lines on the snapshots show the boundaries of the convective region, where $N^2_{axi}<0$.
• Figure 4: Time-evolution of (a) the kinetic and magnetic energies $E_u^S$ and $E_b^S$, (b) the dipolar energy fraction $f_{dip}$, and (c) the mean radial magnetic field on the outer sphere $\langle b_r\rangle_S$, for a MHD simulation with ${\rm{Ek}}=2\times10^{-6}$, ${\rm{Ra}}_C=2\times10^{10}$, ${\rm{Ra}}_T=1.67\times10^9$, $\Pr=0.3$, ${\rm{Sc}}=3$, ${\rm{Pm}}=0.2$ and $\Delta R=0.5$. Mean values are marked with horizontal dotted lines; The vertical dotted line at $t=0.53t_\eta$ shows the time at which snapshots are shown.
• Figure 5: (a) Azimuthal velocity field shown on both spherical surfaces and meridional cross sections at $\phi=0$ and $3\pi/2$ at time $t=0.53t_\eta$. The flow takes the form of a strong prograde equatorial surface jet, with retrograde motion near the poles and in the interior. (b) Spatial structure of the magnetic field, with $b_r$ shown on the outer boundary, and $|\vec{b}|$ in the interior. The radial field at the surface is strongly dipolar, and the interior field shows very localised activity. The surface field is much weaker than in the interior.