Bayesian Evolution Models for Jupiter with Helium Rain and Double-diffusive Convection

TL;DR

This paper models Jupiter's thermal evolution considering helium rain and double-diffusive convection, reconciling observational data with phase diagrams and exploring parameter spaces using MCMC.

Contribution

It introduces a comprehensive evolutionary model of Jupiter incorporating helium immiscibility, double-diffusive convection, and Bayesian analysis to constrain interior properties.

Findings

Models can match Jupiter's flux, helium content, and radius with appropriate phase diagrams.

Inefficient ODDC leads to a thermal boundary layer and rapid cooling of the molecular envelope.

Models favor a colder-than-SCvH-I EOS, with modest superadiabaticity.

Abstract

Hydrogen and helium demix when sufficiently cool, and this bears on the evolution of all giant planets at large separations at or below roughly a Jupiter mass. We model the thermal evolution of Jupiter, including its evolving helium distribution following results of ab initio simulations for helium immiscibility in metallic hydrogen. After 4 Gyr of homogeneous evolution, differentiation establishes a thin helium gradient below 1 Mbar that dynamically stabilizes the fluid to convection. The region undergoes overstable double-diffusive convection (ODDC), whose weak heat transport maintains a superadiabatic temperature gradient. With a generic parameterization for the ODDC efficiency, the models can reconcile Jupiter's intrinsic flux, atmospheric helium content, and radius at the age of the solar system if the Lorenzen et al. H-He phase diagram is translated to lower temperatures. We cast the evolutionary models in an MCMC framework to explore tens of thousands of evolutionary sequences, retrieving probability distributions for the total heavy element mass, the superadiabaticity of the temperature gradient due to ODDC, and the phase diagram perturbation. The adopted SCvH-I equation of state favors inefficient ODDC such that a thermal boundary layer is formed, allowing the molecular envelope to cool rapidly while the deeper interior actually heats up over time. If the overall cooling time is modulated with an additional free parameter to imitate the effect of a colder or warmer EOS, the models favor those that are colder than SCvH-I. In this case the superadiabaticity is modest and a warming or cooling deep interior are equally likely.