Giant Planet Occurrence in the Stellar Mass-Metallicity Plane

Abstract

Correlations between stellar properties and the occurrence rate of exoplanets can be used to inform the target selection of future planet search efforts and provide valuable clues about the planet formation process. We analyze a sample of 1194 stars drawn from the California Planet Survey targets to determine the empirical functional form describing the likelihood of a star harboring a giant planet as a function of its mass and metallicity. Our stellar sample ranges from M dwarfs with masses as low as 0.2 Msun to intermediate-mass subgiants with masses as high as 1.9 Msun. In agreement with previous studies, our sample exhibits a planet-metallicity correlation at all stellar masses; the fraction of stars that harbor giant planets scales as f \propto 10^{1.2 [Fe/H]}. We can rule out a flat metallicity relationship among our evolved stars (at 98% confidence), which argues that the high metallicities of stars with planets are not likely due to convective envelope "pollution." Our data also rule out a constant planet occurrence rate for [Fe/H]< 0, indicating that giant planets continue to become rarer at sub-Solar metallicities. We also find that planet occurrence increases with stellar mass (f \propto Mstar), characterized by a rise from 3.5% around M dwarfs (0.5 Msun) to 14% around A stars (2 Msun), at Solar metallicity. We argue that the correlation between stellar properties and giant planet occurrence is strong supporting evidence of the core accretion model of planet formation.