Killing Planet Candidates with EVEREST

Abstract

We exploit high-quality photometry from the EVEREST pipeline to evaluate false-positive exoplanet candidates from the K 2 mission. We compare the practical capabilities of EVEREST’s pixel-level decorrelation scheme to the data analysis pipelines widely used at the time of these planet candidates’ discovery. Removing stellar variability from the EVEREST-corrected light curves, we search for potential secondary eclipses. For each object exhibiting a secondary eclipse, we compare the implied brightness temperature of the planet candidate to its calculated equilibrium temperature. We thereby identify objects whose brightness temperature is too high to be consistent with a planet. We identify seven systems previously flagged as planetary candidates in preliminary vetting pipelines, and use EVEREST to instead identify six of them as eclipsing binaries. We also project the importance of optimal photometric vetting for TESS data. We find that the majority of blended eclipsing binaries could be identified using TESS photometry, and a systematic study of that kind could in principle also yield valuable information on the mass ratio distribution in stellar eclipsing binaries.