TESS flux corrections give transit depths above 100% for TIC 31858844 in Sectors 13 and 27

We tested whether a large catalog light loss disagreement near WASP-123 remains in public TESS aperture data. We used five fixed SPOC two-minute light curves. These cover both TIC 31858843 and TIC 31858844 in Sectors 13 and 27, plus the bright target in Sector 67. A local linear baseline and a fixed central window give faint-target corrected depths of 150.9 +/- 2.9% and 147.0 +/- 2.0%. The errors are standard errors between events. All 15 faint-target event core means are negative. The same apertures have ordinary uncorrected fractional dips near 1.1%. Their faint/bright absolute loss ratios are 1.019 and 0.962. Most aperture pixels are shared. The pipeline assigns only 0.73% of aperture light to the faint target. Dividing the measured dip by this share predicts a depth above 100%. The measured correction scale agrees within about 4%. All 18 tested window and baseline settings retain mean depths above 100% in both sectors. A target-only eclipse violates the fixed crowding model's light budget if other sources are steady. Contamination and the planet WASP-123 b were already known. This study adds measured correction failures and a reproducible data package, not a new planet or host identification. The retained bright Sector 13 file also has a different crowding value from the value reported in 2022. The result applies to the specified archive bytes. We preserve all inputs, source records, code, event tables, checks, and three figures.

TESS · WASP-123 · TIC 31858844 · crowding correction · transit contamination · negative corrected flux · SPOC · reproducible photometry