Star image fits retain a 1.15-day component at GD 394 in three TESS visits

We test the source of the known optical signal near 1.15 days at the white dwarf GD 394. We fit calibrated star images to public TESS pixel data from sectors 15, 76, and 83. All catalog stars in the fitted field can vary at the tested period and at the nearby star's 0.98-day period. With the frequency and reporting phase adopted from an earlier SCI study, the signed GD 394 components are 0.728 ± 0.221, 0.378 ± 0.116, and 0.453 ± 0.070 electron/s. The errors use the largest scatter from four residual-block lengths. All tested trend, catalog, position, and region choices retain a positive component. Tested image errors retain 95.0 to 100.3 percent of an injected GD 394 signal. Transfer from the bright nearby star is below 0.4 percent per unit signal in those tests. Local position fits have best offsets within 0.3 pixel of GD 394. Some noise samples reach the search boundary, so the position tests do not give complete sky confidence limits. The results support a signal near GD 394 under the fitted image model. They do not identify its physical cause or establish a planet. The data overlap the earlier study, and the errors are conditional on the adopted frequency and model. The package preserves data, code, figures, source records, and steps to repeat the work.

GD 394 · TESS · white dwarf · stellar variability · pixel response function · source position · crowded photometry · reproducible astronomy