Constraints on the formation history and composition of Kepler planets from their distribution of orbital period ratios

Abstract

The Kepler high-precision planetary sample has revealed a radius valley, separating compact super-Earths from sub-Neptunes with lower density. Super-Earths are generally assumed to be rocky planets that were probably born in-situ, while the composition and origin of sub-Neptunes remains debated. To provide more constraints on the formation history and composition, based on the planetary sample of Kepler multiple planet systems, we derive the distributions of orbital period ratios of sub-Neptune and super-Earth planet pairs and calculate the normalised fraction of near-first-order mean motion resonances. Using synthetic planetary systems generated by the Generation III Bern Model, we also obtain theoretical predictions of period ratio distributions of planet pairs of different compositions and origins. We find that actual Kepler sub-Neptune pairs show a normalised fraction smaller (larger) than the model predictions for water-rich (water-poor) pairs with confidence levels of about two sigma. The derived normalised fraction of actual Kepler Super-Earth pairs is generally consistent with that of water-poor model planet pairs but significantly smaller than that of synthetic water-rich planet pairs. Based on the distributions of orbital period ratios, we conclude that orbital migration has been more important for sub-Neptunes than for super-Earths, suggesting a partial ex situ formation of the former and an origin of the radius valley caused in part by distinct formation pathways. However, the model comparisons also show that sub-Neptunes in actual Kepler multiple systems are not likely to be all water-rich/ex situ planets but a mixture of the two (in situ/ex situ) pathways. Whereas, Kepler super-Earth planets are predominantly composed by of water-poor planets that were born inside the ice line, likely through a series of giant impacts without large scale migration.