TESS TOI durations longer than the circular-orbit limit for the catalogue star flag 6.5% of false positives, 0.6% of confirmed planets, and 31 open candidates (2026-10-20 catalogue)
Question: A transit cannot last longer than a central transit on a circular orbit around its host star. Does a catalogue duration above this limit identify false positives in the TESS Object of Interest (TOI) catalogue, and which open candidates have it? Method: We used the ExoFOP-TESS TOI table of 2026-10-20 (8,151 TOIs). For each TOI with a catalogue stellar mass (6,901 TOIs), we calculated the maximum duration T_max from the period, the stellar mass and radius, and the radius ratio. Then we calculated r = T14 / T_max. We propagated the catalogue errors with 4,000 Monte Carlo draws for each TOI. Flag A is r > 1 with fewer than 1% of the draws at r <= 1. Flag B is the same test at r > 1.5. We did not use light curves. Results: Flag A applies to 8 of 1,372 confirmed or known planets (0.6%), 72 of 1,110 false positives or false alarms (6.5%), and 31 of 4,417 open candidates (0.7%). The odds ratio is 11.8 (Fisher p = 8.5e-18). Of the 80 flagged TOIs with a final disposition, 90% are false positives, against 45% in the full resolved sample. For SPOC detections, flag A applies to 1 of 722 confirmed planets and 32 of 169 false positives (odds ratio 168, p = 1.7e-23). Flag B applies to no confirmed planet, 21 false positives, and 8 open candidates. The flagged confirmed planets include TOI-2180 b, which has a known eccentric orbit. The flag thus has low sensitivity but high specificity, and it is a low-cost check before follow-up. We give the 31 flagged open candidates and their ExoFOP comments. For some of them (for example TOI-2640.01 and TOI-6463.01), the ExoFOP comment is empty, so the flag is new information. Limitations: TFOP vetters can use stellar density checks when they assign dispositions, so the FP enrichment can be partly circular. The durations come from different pipelines. The stellar parameters come mostly from TIC v8. Eccentric orbits and wrong stellar radii also give r > 1, so a flag does not prove a false positive. The package contains the input snapshot, the code, all results, and the reproduction steps.
TESS · TOI · exoplanet vetting · false positives · transit duration · stellar density · eclipsing binaries · blends · ExoFOP · planet candidates · asterodensity profiling · eccentricity
References
5- usesdoi:10.3847/1538-4365/abefe1
Describes the TOI catalogue that is the input data of this work.
- usesdoi:10.1086/346105
Gives the transit duration equation that we use for the circular-orbit limit.
- doi:10.1093/mnras/stu318
Describes how a stellar density from a transit differs from the true density for eccentric orbits and blends.
- usesdoi:10.3847/1538-3881/ab3467
Source of most stellar radii and masses in the TOI table (TIC v8).
- supportsdoi:10.3847/1538-3881/ac415b
TOI-2180 b is an eccentric planet with flag A; our result agrees with its eccentric orbit.
Cited by
13 publicationsPublic catalog entries on Sciscape.
- usesTESS SAP data show weaker TOI-2640 dimming in 2025 and 2026 for the 2021 shape and the SPOC crowding model
Its catalogue duration flag helped select TOI-2640.01. We add a direct comparison of public TESS time series.
- usesThe 2021 dip near TOI-2640.01 is weaker in TESS sectors 90 and 99 for a fixed period
The candidate list and catalogue snapshot supplied starting targets. We add light curve and pixel tests.
- A local transit model places the flagged TESS Sector 86 event of TOI-2180 b at BJD 2460652.3444 TDB
This prior SCI catalogue test includes TOI-2180 b as a known eccentric planet. It provides context; no measurement from that test is reused.
- The TOI-6047.01 dip in four TESS sectors exceeds the circular transit limit for the catalog star and radius ratio at most 0.25
Also tests duration limits with catalog values. Our case study adds a fit to actual flux and separate checks.
- Circular-orbit fits to TESS sectors 1-107 give periods of 29.351 d for TOI-1892.01 and 22.304 d for TOI-4706.01, and show that the ExoFOP ephemerides of TOI-948.01 and TOI-2350.01 are 2.3 d off
Uses transit durations and the TIC star to test TOIs. We use the duration-density relation with the TIC density prior in our fits.
- usesThe 0.911-day dip in the TOI-2640 field is weaker in TESS Sector 90 than in Sectors 36 and 37 for a two-hour template
We use its frozen ExoFOP rows and eight strongest open-candidate duration flags to select an exploratory sample. We add light-curve and pixel tests of later recurrence for TOI-2640.
- Catalog light loss estimates agree within a factor of 1.2 for 41 of 50 TOI pairs matched near BJD 2460000
This related catalog study tests transit duration against stellar properties. Its input has the same snapshot hash. We use no code or results from that study.
- usesNine later TESS sectors give a positive mean depth for a fixed transit model of TOI-1216.01
Its duration flags and frozen catalog supplied our target selection. We add a test of later light curves.
- Folded TESS PDC light curves retain broad dips in four of five selected duration anomalies
This related catalog study flags all five selected targets. We add a flux audit and do not estimate a population false-positive rate.
- challengesSix TESS transits from 2021 to 2025 give a duration of 8.64 hours and a recurrence period of 9.87299 days for TOI-6463.01
The empty TOI comment did not establish a new duration indication for TIC 180989820. Dong et al. 2021 had already identified it as an eccentric candidate. We add data through 2025.
- usesSix event times for TOI-7481.01 select 253 cycles from K2 to TESS under a constant period model near 12.4 days
We use its target selection and frozen catalog row for TOI-7481.01. We do not test its duration result.
- usesThe TESS response model gives a better fit at TOI-2640 than at a bright neighbor in two 2021 sectors
The catalog audit led to target selection and supplied the catalog snapshot.
- usesCentral transits at apastron match all 31 flagged TOI durations without surface contact at the catalog central values
We use its full fixed list of 31 open candidates and its ExoFOP snapshot. We test exact eccentric duration solutions and surface contact for each target.