Examining the Biases in Kepler Data due to Long-Term Dynamical Evolution

The Kepler Space Telescope has provided data on thousands of exoplanets and remains the primary dataset for studying exoplanetary demographics. To ensure a representative sample of the true population of exoplanets, many observational biases within this dataset have been measured, but more elusive biases remain unstudied. For example, in the case of multi-transiting systems, the Kepler Telescope was biased towards observing systems with low mutual inclinations, but these inclinations can change significantly over secular timescales due to dynamical interactions. This work has examined biases related to the long-term dynamical evolution of close-in planetary systems. To overcome observational limits inherent to exoplanetary detection methods—which prevent the characterization of such parameters as mass and true inclination—we have used a synthetic population of exoplanets from SysSim (Paper III, He et al. 2020) which was tuned to fit many planetary and system-wide properties from the Kepler data. These systems were integrated 1 Myr using REBOUND (Rein and Liu 2012) with single and multi-transit probabilities measured at each timestep of the simulation using CORBITS (Brakensiek & Ragozzine 2016). Overall, the time-average of the number of transiting planets was found to vary by 0.01% - 3%, depending on the system, on timescales much longer than the four year Kepler mission. The multiplicity distribution (how many systems with different numbers of transiting planets) varied by a few percent over time. Finally, the difference between the mean and median mutual inclinations over time is a proxy for the bias towards small mutual inclinations in the Kepler data; we find this difference to be less than 0.03 degrees on average. These initial results suggest that biases in the Kepler dataset due to the long term secular evolution of exoplanets are small.