Measuring Transit Signal Recovery in the Kepler Pipeline IV: Completeness of the DR25 Planet Candidate catalog

TL;DR

This paper empirically evaluates the detection efficiency of the Kepler pipeline for planet candidate identification, revealing high overall detection rates and dependencies on stellar and observational parameters, crucial for accurate exoplanet occurrence rates.

Contribution

It provides the first comprehensive empirical measurement of Kepler pipeline detection efficiency across various parameters, improving the accuracy of exoplanet occurrence rate calculations.

Findings

Detection efficiency averages 90-95% for strong signals.

Detection efficiency weakly depends on the number of transits and orbital period.

Detection efficiency is more strongly influenced by stellar temperature and noise properties.

Abstract

In this work we empirically measure the detection efficiency of Kepler pipeline used to create the final Kepler Threshold Crossing Event (TCE; Twicken et al. 2016) and planet candidate catalogs (Thompson et al. 2018), a necessary ingredient for occurrence rate calculations using these lists. By injecting simulated signals into the calibrated pixel data and processing those pixels through the pipeline as normal, we quantify the detection probability of signals as a function of their signal strength and orbital period. In addition we investigate the dependence of the detection efficiency on parameters of the target stars and their location in the Kepler field of view. We find that the end-of-mission version of the Kepler pipeline returns to a high overall detection efficiency, averaging a 90-95% rate of detection for strong signals across a wide variety of parameter space. We find a weak dependence of the detection efficiency on the number of transits contributing to the signal and the orbital period of the signal, and a stronger dependence on the stellar effective temperature and correlated noise properties. We also find a weak dependence of the detection efficiency on the position within the field of view. By restricting the Kepler stellar sample to stars with well-behaved correlated noise properties, we can define a set of stars with high detection efficiency for future occurrence rate calculations.