Reproducing the stellar-mass dependence of the giant planet occurrence rate with pebble accretion models

Abstract

The stellar mass dependence of the unbiased giant planet occurrence rate may be the best statistical tool to constrain the formation of such planets. This rate rises and falls as a function of stellar mass, peaking around stars of $\sim 1.7{-}2 \Ms$. In this work, we carry out a population synthesis study, using pebble-driven core accretion model of planet formation, to investigate the planet formation conditions that may be responsible for this stellar-mass dependence. We use the inferred giant planet occurrence rated of three combined homogenised radial velocity surveys (EXPRESS, PPPS, and Lick giant star survey) to constrain the models. We find that we can produce a synthetic giant planet population with closely aligned occurrence and properties when we base our model on observationally-supported assumptions that accretion rates are higher and disk lifetimes are shorter around more massive stars, we can produce a synthetic giant planet population with closely aligned properties to the observed distribution. We also find that in this scenario, the runaway gas accretion occurs at a larger orbital distance and earlier times as the stellar mass increases.

Figures (6)

• Figure 1: Stellar metallicity versus mass for the 482 stars of the homogenised combined RV sample Wolthoff2022PreciseSurveys.
• Figure 2: Mass accretion rates from observational studies Wichittanakom2020TheStarsManara2018WhyPopulation, adapted from the Appendix of Iglesias2023X-ShooterEvolution. Linear fits are shown for reference; dashed red line for M$17$, dotted lime for J24, and solid orange line for W$20$. The stars are colour coded by stellar age. There are no individual ages for stars in the sample by Manara2017X-ShooterStars, so we adopt the age of the Chamaeleon I star-forming region to be 1.5 Myr from Galli2021ChamaeleonData. The white highlighted region is the stellar mass range that we are interested in exploring.
• Figure 3: Birth location against birth embryo mass for time 0 for key stellar masses 1 (orange), 1.7 (pink), and 2.4 $\ \rm M_\odot$ (purple). Within the first few au of the host stars, the birth embryo masses are essentially identical.
• Figure 4: Planet occurrence rate (colour coded) as a function of metallicity and stellar mass in the Lick stars in our sample Reffert2006PRECISECOMPANIONWolthoff2022PreciseSurveys (a) and in our synthetic models: M17 (b), W20 (c), and J24 (d). Data has been adapted from Reffert2015PreciseMetallicity and heavily smoothed in (a) so that general trend can be examined. There is a clear maximum in the observed planet occurrence rate (a) for metallicities of about 0.2 and masses of about $2 \ \rm M_\odot$. For our models (b, c, d), the peak of planet formation shifts depending on the initial conditions and is more concentrated as the data does not need to be smoothed.
• Figure 5: Histogram of our mean model giant planet occurrence rates as a function of stellar mass (left) and metallicity (right). Our best mean models are coloured in orange (W20) and lime (J24), respectively. The solid black histograms shows the completeness-corrected observed giant planet occurrence rate in each bin Wolthoff2022PreciseSurveys.