Six TESS PDC light curves give HAT-P-2 upper RMS limits below 19 ppm at harmonics 79 and 91 under the primary Gaussian model at four fixed coherence times from 0.1 to 3 days

We test random optical signals at orbital harmonics 79 and 91 of HAT-P-2. Earlier infrared data showed signals near these orders. A prior SCI study tested stable signals in six TESS visits. We use the same six public 120-second light curves and test loss of coherence within a visit. We fit two random signals with an exponential cosine covariance. The fixed coherence times are 0.1, 0.3, 1, and 3 days. Each signal has the same stationary variance in all visits. The model has independent sector blocks and fitted Gaussian white and red noise. The data do not require either signal. Profile upper ranges at the eight-test family reference are 18.19, 15.76, 13.92, and 13.07 ppm RMS for harmonic 79. The ranges for harmonic 91 are 17.79, 15.73, 12.92, and 10.84 ppm RMS. These values describe the corrected delivered optical flux. An aperture-flux control gives a weaker upper range of 27.25 ppm at the shortest coherence time. They are model-based likelihood ranges, with no exact confidence guarantee. Residual tails depart from a Gaussian model. We report background, aperture-flux, and event-mask controls, numerical checks, and 32 Gaussian simulations. The result tests a defined class of stationary random signals. It does not test rare short events or prove that the infrared result is false. The package contains source data, code, full fits, figures, provenance, and reproduction instructions.

HAT-P-2 · TESS · stellar oscillations · finite coherence · Gaussian process · public data · profile likelihood