A transit-depth variability screen of 815 TESS TOIs with periods below 1.6 days finds no disintegrating-planet candidate and shows that wrong catalogue periods cause false depth changes

Question: Do public TESS 2-minute light curves show new disintegrating planets among known short-period TOIs, from transit depths that change from orbit to orbit? Approach: We fitted the depth of each transit of 875 TOIs with periods below 1.6 days (4,724 SPOC light curves, sectors 1-107) with a local baseline. We applied the same fit at off-transit phases as noise controls. We measured the fractional transit-to-transit scatter f and used a pre-set rule with trials correction, odd-even, split-half and same-star control tests. 2,531 injections into real light curves measured the sensitivity. The known disintegrating planet BD+05 4868 Ab was a positive control. Results: 815 TOIs had enough data. The control BD+05 4868 Ab gives f = 0.27 (0.18-0.39) and passes. No TOI is a disintegrating-planet candidate after checks. Two TOIs pass the rule, but crowding (TOI-3070.01) and spot modulation at almost the TOI period (TOI-4710.01) explain them; three others are binaries or have small scatter. 464 of 575 planet-like TOIs have f < 0.5 (95 percent). The rule recovers 86 percent of CV = 0.5 signals at per-transit SNR 4 and passes 1 of 640 constant injections. Wrong catalogue periods made at least four of six rule passes in a first pass. Per-transit times confirm period errors of 0.0002-0.004 days for 17 TOIs, far above the quoted errors. Limitations: Most small planets have per-transit SNR below 2, so their limits are weak; Kepler-1520 b is too faint in TESS. Only TOIs with 2-minute data were tested. No new observations were used.

astronomy · TESS · disintegrating planets · transit depth variability · ultra-short-period planets · TOI · null result · injection-recovery · ephemeris errors · BD+05 4868 Ab · Kepler-1520 b · false positives