11:05 AM - 11:20 AM
[SIT18-10] Termination of a lunar dynamo inferred from compositionally-driven dynamo models
Keywords:Dynamo evolution, Thermal history, Compositional convection, Lunar magnetic field, Inner core growth
We performed three-dimensional numerical dynamo simulations for an incompressible Boussinesq fluid in a rotating shell. Since the inner core grows in a geological time scale, we change non-dimensional parameters such as the ratio of inner to the outer core radii χ, the Rayleigh number Ra, and the Ekman number E. Simulations are carried out along with evolutional curves of Ra and E as functions of χ, guided by thermal history models of the lunar core [Laneuville et al., 2014 and Sheinberg et al., 2015]. Note that χ varies from 0.1 to 0.7 referring to the seismic observations of the Moon [Weber et al., 2011]. The compositional Prandtl number Pr and the magnetic Prandtl number Pm are fixed at 1 and 5, respectively. The inner and outer boundaries are no-slip and electrically insulating, while the fixed mass flux and zero-flux are set at the inner and outer boundaries, respectively.
Results using several evolutionally curves show that a larger Ra is required to sustain the dynamo in a thinner shell than that in a thicker shell. The magnitude of the curves determines whether a dynamo succeeds or fails at a given χ, while a flow structure in the fluid core after the dynamo shutdown have two common properties independent of this magnitude: Firstly, a strong flow exists only inside the tangent cylinder TC (a virtual cylinder tangent to the inner core at the equator and parallel to the axis of rotation) or exists both inside and outside for successful dynamo, whereas this strong flow inside TC is very weak for failed dynamo, except for χ=0.1. Secondly, the large-scale azimuthal westward flow is particularly strong at the equator. This second feature suggests that the strength of the azimuthal flow could be a marker in terminating the dynamo. Thus, we investigate the ratio of the poloidal to the toroidal component of the kinetic energy in the fluid core KEP/KET for all of the simulation results. We find that a higher value of KEP/KET is required to sustain the dynamos in thinner shells, and the dynamo is not maintained at KEP/KET<0.2, with some exception. The exceptional cases are found near the critical Rayleigh number at χ=0.1 and 0.5. In these cases, the magnetic field is non-dipolar at the outer boundary and the Elsasser number Λ is less than 1. In contrast, in most of the cases, the magnetic field is dipolar-dominant and 1<Λ<10. Therefore, there are two possible scenarios for termination of the lunar dynamo. (1) Relatively strong (1<Λ<10) dipolar field is generated and then stops due to the enhanced azimuthal flow. (2) Weak(Λ<1) non-dipolar field is maintained and then ceases due to the enhanced azimuthal flow. It is suggested that the magnetic fields in (1) and (2) are sustained by different dynamo mechanisms.