Occurrence rates based on probabilistic catalogs

[christinahedges]

Near the Kepler detection limit, where small planets in long-period orbits are found, detection reliability drops significantly and the confident identification of planet candidates becomes problematic. Distinguishing both instrumental and astrophysical false positives from true planets becomes very difficult: at low Signal-To-Noise (SNR) many techniques that identify astrophysical false positives struggle, and instrumental systematics can closely mimic small-planet transit signals. Methods that exclude such false positives often also tend to exclude planets.

Research described in #2, #7, and #10 will alleviate some, but not all, of these issues. The development of vetting techniques that assign planet candidate probabilities to all detections when the SNR is low is a promising, though challenging, approach that would extend Morton et al. (2016) to all detections. Population-based inference techniques using such a probabilistic catalog, extending, for example, Farr et al. (2015), would allow inferences even when all detections have a low probability of being a planet. This approach would enable a higher-confidence estimate of the occurrence of terrestrial planets in the habitable zone than is currently available.