Gaia DR3 transit periods differ from the orbital period for 161 of 174 known planets because of aliasing; TESS gives true periods for DR3 candidates and 3 planet-like signals not in TOI lists

Question: Are the periods in the corrected Gaia DR3 table vari_planetary_transit (214 stars: 173 known planets, 41 new candidates) correct, and do TESS data from 2018-2026 confirm the new candidates? Method: We compared the table with published ephemerides (NASA Exoplanet Archive) and tested both ephemerides on Gaia DR3 G epoch photometry. For 37 candidates that were not known before DR3, we ran an iterative BLS search on SPOC, TESS-SPOC and QLP light curves. We then tested whether each TESS ephemeris predicts the Gaia in-transit epochs 300-2,000 orbits earlier, with a phase-shift null test for the false-alarm probability (FAP). Results: For 174 sources with a confirmed planet, the table period equals the orbital period (within 1%) for only 13. For 8 more the ratio is an integer (2-5) or its inverse, and for 153 there is no simple relation. The table period is shorter in 139 of 161 mismatches. The wrong periods are sampling aliases: the table ephemeris selects real in-transit epochs (median Gaia SNR 7.4, against 6.7 at the published ephemeris), mostly the same 2-5 epochs. The February 2023 correction changed reference times but only 4 periods. Of 33 candidates with TESS data, 17 have a TESS signal that matches the Gaia epochs (FAP <= 0.01): 9 planet-like and 8 eclipsing-binary-like. The 9 include Gaia-1 b, Gaia-2 b, TOI-5905.01, TOI-5964.01, TOI-6409.01, TOI-5839.01 (TFOPWG FP; the resolved Gaia photometry shows the dip on the target) and three signals not in the ExoFOP TOI or CTOI lists of 2026-11-08: Gaia DR3 5431056347037926784 (TIC 4638108, P = 2.208287 d, 14 ppt), Gaia DR3 5345417757181174144 (TIC 93622019, P = 2.123886 d, 10 ppt) and Gaia DR3 5432318139709900160 (TIC 44997521, P = 3.549389 d, 21 ppt). For 16 candidates no TESS signal matches the Gaia epochs. The pipeline recovers the published period for 29 of 29 control planets. Limitations: The three new signals are candidates only; we have no spectroscopy, high-resolution imaging or centroid analysis, and the first one has a V-like shape. CDS catalogs were not reachable, so ground-based candidate lists were not checked. We used at most 8 TESS sectors per star. Do not use the DR3 transit_period as the orbital period without an independent check.

Gaia DR3 · TESS · transiting exoplanets · period aliasing · planet candidates · eclipsing binaries · catalog validation · box least squares · hot Jupiters · vari_planetary_transit