Exploring Warm Jupiter Migration Pathways with Eccentricities II. Correlations with Host Star Properties and Orbital Separation

TL;DR

This study uses a uniform RV-based refitting of 200 warm Jupiters to infer the population-wide eccentricity distribution and its dependence on host metallicity, stellar mass, and orbital separation via hierarchical Bayesian modeling. The inferred distribution is well described by a Beta form with $\alpha=1.00^{+0.09}_{-0.08}$ and $\beta=2.79^{+0.28}_{-0.26}$, with about $27^{+3}_{-4}\%$ of planets on nearly circular orbits and $73^{+3}_{-3}\%$ dynamically hot. Metal-rich host stars harbor warmer giants with higher eccentricities, while stellar mass and orbital separation within $\approx$0.5–2.0 $M_{\odot}$ and 0.1–1 AU show no significant effect on the underlying $e$ distribution; thick-disk hosts appear systematically less dynamically excited than thin-disk hosts. Collectively, the findings support planet-planet scattering as a major contributor to warm Jupiter architectures and highlight metallicity as a key factor shaping eccentricity distributions in this population.

Abstract

Warm Jupiters with orbital periods of $\approx$10-365 d represent a population of giant planets located well within the water ice line but beyond the region of tidal influence of their host star relevant for high-eccentricity tidal migration. Orbital eccentricities offer important clues about the formation and dynamical history of warm Jupiters because in situ formation and disk migration should imprint near-circular orbits whereas planet scattering should excite eccentricities. Based on uniform Keplerian fits of 18,587 RVs targeting 200 warm Jupiters, we use hierarchical Bayesian modeling to evaluate the impact of host star metallicity, stellar mass, and orbital separation on the reconstructed population-level eccentricity distributions. Warm Jupiters take on a broad range of eccentricities, and their population-level eccentricities are well modeled using a Beta distribution with $α$ = 1.00$^{+0.09}_{-0.08}$ and $β$ = 2.79$^{+0.28}_{-0.26}$. We find that 27$^{+3}_{-4}\%$ of warm Jupiters have eccentricities consistent with near-circular orbits ($e$ $<$ 0.1), suggesting that most warm Jupiters (73$^{+3}_{-3}\%$) detected are dynamically hot. Warm Jupiters orbiting metal-rich stars are more eccentric than those orbiting metal-poor stars -- in agreement with earlier findings -- but no differences are observed as a function of stellar host mass or orbital separation, at least within the characteristic ranges probed by our sample ($\approx$0.5--2.0 $M_{\odot}$ and 0.1--1 AU, respectively). In this sense, metallicity plays a larger role in shaping the underlying eccentricity distribution of warm Jupiters than stellar mass or final orbital distance. These results are broadly consistent with planet scattering playing a major role in shaping the orbital architectures of close-in giant planets.

Figures (7)

• Figure 1: Eccentricity plotted as a function of semi-major axis for all 200 warm Jupiters with new, consistently fit orbits in this analysis. Circles correspond to warm Jupiters in single planet systems and stars represent warm Jupiters in multi-planet systems. Symbols are colored blue if the warm Jupiter orbits a metal-poor host ([Fe/H] $<$ 0) and orange if they orbit a metal-rich host ([Fe/H] $>$ 0). The grey shaded area shows tracks of constant angular momentum which planets follow during tidal circularization. Here we have adopted a mass of 1 $M_\mathrm{Jup}$ and a radius of 1 $R_\mathrm{Jup}$ around a Sun-like star. The dashed black line displays where $f$-mode tidal dissipation could speed up tidal migration (Wu2018; Dong2021).
• Figure 2: Left: Stellar host metallicity measurements compiled from the literature and used in this analysis compared with metallicities derived from Gaia DR3 RVS spectra (GaiaCollaboration2022). Individual warm Jupiter host stars are colored by Gaia log $g$ to highlight main-sequence stars (log $g$$>$ 4.0 dex) and post main-sequence stars (log $g$$<$ 4.0 dex). The dashed line represents the 1:1 relation, while the solid line shows the best-fit linear relation. Middle: Stellar masses used in this analysis compared with uniformly inferred FLAME masses from Gaia. Systems that lie well above the 1:1 line have Gaia surface gravities suggesting they are massive post-main-sequence stars. Right: Host stars plotted on a Gaia $M_{G}$ vs $G_{BP}$ -- $G_{RP}$ color-magnitude diagram. Iso-mass tracks spanning 1.0--3.0 $M_\odot$ are shown with solar metallicities ([Fe/H] = 0 dex). The most evolved stars tend to have the highest stellar masses.
• Figure 3: Plotted in orange is the metallicity distribution of $\approx$20,000 field stars in the solar neighborhood within $\sim$40 pc (Nordstrom2004). The blue histogram shows the metallicity distribution of warm Jupiter host stars in this study. These hosts are, on average, more metal-rich than the field star population.
• Figure 4: Toomre diagrams illustrating $\mathit{U},\mathit{V},\mathit{W}$ Galactic kinematic space velocities of the warm Jupiter host stars in this analysis using their Gaia DR3 measurements. Contours separate stars that are kinematically most similar to the thin disk, thick disk, and halo (see Bensby2003). We only plot systems with kinematic uncertainties less than 20$\%$. Left: Here, symbol colors relate to stellar metallicity, showing a trend in which the most metal-poor host stars in our sample belong preferentially to the galactic thick disk. Right: Here, symbol colors scale with the eccentricity of the warm Jupiter. All warm Jupiters orbiting thick disk stars have near-circular orbits.
• Figure 5: Top: Host star metallicity as a function of planet eccentricity for all warm Jupiters in single planet systems (blue) and multi-planet systems (orange). Middle: Stellar host mass as a function of eccentricity for all warm Jupiters. Bottom: Semi-major axis as a function of eccentricity for all warm Jupiters. For all panels, the size of the circles scale with minimum planet mass, $m_p \sin i$.