TESS Archive reanalysis narrows habitable-zone planet list

Astrobiology Web reported that researcher Juan F. Pagés Navarrete submitted a paper to arXiv on Nov. 2, 2026 reanalyzing 187 TESS Objects of Interest with only one or two observed transits, using all public TESS data through Sector 107.

When a planet crosses the face of its star, it blocks a small fraction of starlight in an event called a transit. TESS looks for these repeating dips to identify candidate planets, and each detection is logged as a TESS Object of Interest. With only one or two transits, the orbital period cannot be measured directly, so the catalogue leaves it undefined or sets it to the spacing between the two events.

This is the central validation problem described in the Astrobiology Web report. A catalogue-period criterion uses the spacing between the two observed transits as the orbital period, the rule that generated the original habitable-zone flags for these pairings. Under that rule, 28 duo-transits appear to sit in the conservative habitable zone. But once real periods are considered, only about two are expected to remain there, and both are giants.

Among the planet-sized members of that same set of 28, none are expected to actually occupy the conservative habitable zone. Resolving their periods shows they are aliases of planets that orbit closer to their stars. An alias is a different orbital period that fits the same sparse pattern of observed transits, so the catalogue’s period placeholder can point to the wrong distance from the star.

The new analysis targets all 187 TOIs catalogued with P = 0 or P >= 200 d. The study uses all public TESS observations through Sector 107 to revisit them.

The flux-based method weights each alias by per-epoch transit likelihood, transit duration and a geometric prior. It then examines coherence across the observations to see whether a candidate period lines up with the data. The method also checks data edges, possible contamination by other planets and instrumental artefacts before accepting any period solution.

Solving a period means choosing the true orbit from among aliases. Fixing by exclusion removes the rival periods until one remains. These steps apply to candidates whose periods were previously undefined or set to the transit gap. The reanalysis resolves 28 periods and fixes eight more by exclusion, giving those candidates a settled orbital period where the catalogue previously offered only a gap or a blank.

The method recovers 12 published periods blindly. Blind recovery means the method finds periods that have already been published without being given them beforehand.

Injection-recovery tests add a new transit to pin down the period and then verify whether the method returns it. In these tests, 71 per cent of such periods came back correctly and none came back wrong. When an additional transit constrains the orbit, the method usually finds the right period and does not return an incorrect one.

Some genuinely habitable-zone planets were previously missed or misclassified because sparse transits left their period undefined or wrong. The study brings back three previously published cases around M-dwarf stars. It also validates four planets with new periods, establishing orbital timing that was previously uncertain or incorrect.

TOI-2298 b is one of the four newly validated planets. A 146.58-day orbit around a K-dwarf star puts it in the conservative habitable zone. A radius of 3.0 Earth radii makes it a sub-Neptune.

TOI-6075 b has a measured period of 416.5 days, though an 833-day alias remains possible. Its radius is 1.7 Earth radii. It is only the second validated transiting planet below 2 Earth radii with an orbital period longer than one year. The possible 833-day alias means the longer orbit cannot yet be ruled out entirely.

The study releases alias-marginalised habitable-zone probabilities for the whole sample of 187 TOIs. The released probabilities marginalise over the possible orbital periods for every candidate. A habitable-zone likelihood is assigned without depending on choosing a single alias, across the full set of 187 TOIs.

The paper runs 17 pages with 12 figures and 6 tables. It is by Juan F. Pagés Navarrete and was submitted to MNRAS as MN-26-3478-P. The manuscript was posted as arXiv:2611.03019, and Table 2 was released as an ancillary file.

The arXiv identifier for the submitted paper is arXiv:2611.03019v1. The preprint is assigned the DOI 10.48550/arXiv.2611.03019. The arXiv submission file is 619 KB in size. The manuscript was submitted to arXiv on Nov. 2, 2026 at 12:17:59 UTC.

The submitter is listed as Juan Francisco Pagés Navarrete. The paper is tagged with Earth and Planetary Astrophysics (astro-ph.EP) and Instrumentation and Methods for Astrophysics (astro-ph.IM).