The Gaia-Kepler Stellar Properties Catalog. II. Planet Radius Demographics as a Function of Stellar Mass and Age

Abstract

Studies of exoplanet demographics require large samples and precise constraints on exoplanet host stars. Using the homogeneous Kepler stellar properties derived using Gaia Data Release 2 by Berger et al. (2020), we re-compute Kepler planet radii and incident fluxes and investigate their distributions with stellar mass and age. We measure the stellar mass dependence of the planet radius valley to be $d \log R_{\mathrm{p}}$/$d \log M_\star = 0.26^{+0.21}_{-0.16}$, consistent with the slope predicted by a planet mass dependence on stellar mass ($0.24-0.35$) and core-powered mass-loss (0.33). We also find first evidence of a stellar age dependence of the planet populations straddling the radius valley. Specifically, we determine that the fraction of super-Earths ($1-1.8 \mathrm{R_\oplus}$) to sub-Neptunes ($1.8-3.5 \mathrm{R_\oplus}$) increases from $0.61 \pm 0.09$ at young ages (< 1 Gyr) to $1.00 \pm 0.10$ at old ages (> 1 Gyr), consistent with the prediction by core-powered mass-loss that the mechanism shaping the radius valley operates over Gyr timescales. Additionally, we find a tentative decrease in the radii of relatively cool ($F_{\mathrm{p}} < 150 \mathrm{F_\oplus}$) sub-Neptunes over Gyr timescales, which suggests that these planets may possess H/He envelopes instead of higher mean molecular weight atmospheres. We confirm the existence of planets within the hot sub-Neptunian "desert" ($2.2 < R_{\mathrm{p}} < 3.8 \mathrm{R_\oplus}$, $F_{\mathrm{p}} > 650 \mathrm{F_\oplus}$) and show that these planets are preferentially orbiting more evolved stars compared to other planets at similar incident fluxes. In addition, we identify candidates for cool ($F_{\mathrm{p}} < 20 \mathrm{F_\oplus}$) inflated Jupiters, present a revised list of habitable zone candidates, and find that the ages of single- and multiple-transiting planet systems are statistically indistinguishable.