Re-analysing the dynamical stability of the HD 47366 planetary system

Abstract

Multi-planet systems around evolved stars are of interest to trace the evolution of planetary systems into the post-main sequence phase. HD 47366, an evolved intermediate mass star, hosts two giant planets on moderately eccentric orbits. Previous analysis of the planetary system has revealed that it is dynamically unstable on timescales much shorter than the stellar age unless the planets are trapped in mutual 2:1 mean motion resonance, inconsistent with the orbital solution presented in \cite{2016Sato} (hereafter: S16), or are moving on mutually retrograde orbits. Here we examine the orbital stability of the system presented in S16 using the $n$-body code {\sc Mercury} over a broad range of $a$--$e$ parameter space consistent with the observed radial velocities, assuming they are on co-planar orbits. Our analysis confirms that the system as proposed in S16 is not dynamically stable. We therefore undertake a thorough re-analysis of the available observational data for the HD 47366 system, through the Levenberg-Marquardt technique and confirmed by MCMC Bayesian methodology. Our re-analysis reveals an alternative, lower eccentricity fit that is vastly preferred over the highly eccentric orbital solution obtained from the nominal best-fit presented in S16. The new, improved dynamical simulation solution reveals the reduced eccentricity of the planetary orbits, shifting the HD 47366 system into the edge of a broad stability region, increasing our confidence that the planets are all that they seem to be. Our rigorous examination of the dynamical stability of HD 47366 stands as a cautionary tale in finding the global best-fit model.