PREDICTING PLANETS IN KEPLER MULTI-PLANET SYSTEMS

Abstract

We investigate whether any multi-planet systems among Kepler candidates (2011 February release) can harbor additional terrestrial-mass planets or smaller bodies. We apply the packed planetary systems hypothesis that suggests all planetary systems are filled to capacity, and use a Hill stability criterion to identify eight two-planet systems with significant gaps between the innermost and outermost planets. For each of these systems, we perform long-term numerical integrations of 10{sup 7} years to investigate the stability of 4000-8000 test particles injected into the gaps. We map out stability regions in orbital parameter space, and therefore quantify the ranges of semimajor axes and eccentricities of stable particles. Strong mean-motion resonances can add additional regions of stability in otherwise unstable parameter space. We derive simple expressions for the extent of the stability regions, which is related to quantities such as the dynamical spacing {Delta}, the separation between two planets in units of their mutual Hill radii. Our results suggest that planets with separation {Delta} < 10 are unlikely to host extensive stability regions, and that about 95 out of a total of 115 two-planet systems in the Kepler sample may have sizeable stability regions. We predict that Kepler candidate systems including KOI 433, KOI 72/Kepler-10, KOI 555, KOI 1596, KOI 904, KOI 223, KOI 1590, and KOI 139 can harbor additional planets or low-mass bodies between the inner and outer detected planets. These predicted planets may be detected by future observations.