Incidence of unresolved binary impostors in Gaia DR4/DR5 exoplanet catalogs

Determine whether the incidence of unresolved near-twin binary stars that mimic planetary signals in the Gaia Data Release 4 and Data Release 5 astrometric exoplanet candidate catalogs is comparable to their incidence in the Gaia DR3 exoplanet candidate catalog, by quantifying the false-positive contamination from photocentric motion and accounting for selection biases and the candidate selection procedures used to construct each catalog.

Background

Unresolved near-twin binaries can exhibit low-amplitude photocentric motion that partially cancels the stars’ opposite orbital motions, producing astrometric signals similar to those of a single star orbited by a dark planetary-mass companion. Follow-up of Gaia DR3 astrometric exoplanet candidates has shown that more than half of the candidates observed with radial velocities were unresolved nearly-twin binaries, highlighting the seriousness of this false-positive channel.

The DR3 candidate list was produced through a complex selection process and is subject to strong observational biases, which may differ in DR4 and DR5. Consequently, the fraction of candidates that are impostor binaries could change substantially in later releases, making it important to quantify and compare contamination rates across Gaia data releases to guide follow-up strategies and interpret catalog statistics.

References

How much will the DR4 and DR5 exoplanet catalogs be contaminated by false positives? It is unclear whether the incidence of unresolved binaries as planet impostors will be comparable to that of the DR3 catalog, given the strong observational biases as well as the complex process used to select the DR3 exoplanet candidates (see \citealt{GaiaCollaboration2023}).

On the Exoplanet Yield of Gaia Astrometry  (2511.04673 - Lammers et al., 6 Nov 2025) in Section 5, Binary stars as planet impostors