At the reported period, eight TESS transits from 2023 do not confirm or exclude the 32-second signal of KELT-2A b

We tested the reported 32.4-second transit timing signal of KELT-2A b at its reported period of 2.4753 planet orbits. We measured 24 transit times in five public TESS light curves with 120-second exposures. Sixteen transits were from 2021. Eight were from 2023 and were outside the original timing study. Each fit used a fixed symmetric transit shape and a local flux baseline. We fitted a separate linear clock and a free sine phase in each era. The later data gave a best amplitude of 43.3 seconds. Adding the two sine coefficients improved chi-square by 4.39, with p = 0.112 at the fixed period. A conditional 95 percent region for these coefficients gave an amplitude upper bound of 94.5 seconds. This region allowed zero and 32.4 seconds. With errors that include residual noise, tests with raw flux, other baselines and shapes, and residual blocks kept this result. The later observing design had about 27 percent power to detect a 32.4-second signal at a 5 percent threshold under the estimated Gaussian error model. The published timing table also gave a broad period profile. A change in chi-square of one covered 2.395 to 2.566 orbits with the quoted errors. This does not reproduce the published 0.0004-orbit error under our model. Thus, we did not treat the early phase forecast as fixed. The bounds assume independent timing errors and the stated transit model. They do not establish constant timing or a moon. We release the measured times, original light curves, source records, code, figures, and instructions for an offline repeat.

KELT-2A b · TESS · transit timing · exomoon corridor · signal recovery · period uncertainty · public data