Unveiling hidden multiplicity in the TESS archive a systematic vetting of sub arcsecond stellar companions using Gaia DR3

Abstract

We present a high-efficiency astrometric and photometric vetting pipeline to isolateunresolved sub-arcsecond stellar companions contaminating the TESS Objects of Interest(TOI) catalog. Instead of utilizing arbitrary sample thresholds or local processing constraints,we cross-matched the entire active population of the NASA Exoplanet Archive (5966 activecandidates) against the Gaia Data Release 3 (DR3) database. By targeting systems withsignificant Renormalized Unit Weight Error (RUWE > 1.4) and a high Image ParameterDetermination multi-peak fraction (IPDfmp > 5%), followed by an additional astrometricquality-control stage (Section 2.4), we identified a vetted population of 204 sub-arcsecondbinary-star candidates that lack prior duplicity classifications in the literature. We furthercross-matched this catalog against the bulk ExoFOP-TESS imaging-observations log (Section2.5) and find that 78 of the 204 candidates have no high-resolution imaging observationcurrently on record, making them efficient targets for new follow-up. These undetectedcompanions likely induce systematic photometric dilution, leading to underestimated transitdepths and overestimated planetary radii. Confirmation of physical duplicity for any individualsystem requires dedicated high-contrast imaging or speckle interferometry and is notestablished by astrometric indicators or the absence of an ExoFOP record alone.