The first disappearing planet slipped out of view while TESS watched

A giant planet that NASA’s TESS space telescope had watched cross the face of its star, pass after pass, has stopped crossing it. Each transit grew shorter and shallower as the planet’s path slid toward the edge of the star, until the telltale dip in starlight failed to appear at all. The planet was not destroyed or swallowed. Its orbit is tipping, and astronomers caught the moment it tipped out of line with Earth.

That makes it the first confirmed planet around a single star whose transits were seen to disappear, according to a UCLA-led team writing in The Astrophysical Journal Letters. The likely culprit is a second giant planet that nobody has seen directly. Its tilted orbit drags on its neighbor and makes both orbits wobble like a spinning top. Astronomers usually infer changes like this over centuries; here one played out within a single space mission.

The vanished planet is TOI-2535 b, a hot Jupiter: a gas giant about 0.84 times the width of Jupiter that hugs an old orange dwarf star roughly 1,000 light-years away. It completes an orbit in about 8.2 days, around ten times faster than Mercury circles the Sun. Hundreds of hot Jupiters are known. This one stood out because its transits kept arriving up to six hours early or late, a sign that something else in the system was pulling on it.

Recommended ReadingA planet orbiting backwards around a red dwarf has no giant neighbor to blame

“This is the most extreme example we’ve seen because the two planets are interacting so strongly,” said Samuel Yee, a 51 Pegasi b Postdoctoral Fellow at UCLA and the paper’s corresponding author. “The planet’s orbit is tilting so quickly that its transit disappeared while we were watching.”

How TESS caught a planet slipping out of view

A transit happens only when a planet’s orbit lines up almost exactly with our line of sight, so that the planet passes between its star and Earth and blocks a sliver of light. TESS hunts for those slivers across millions of stars. It returned to TOI-2535 in five separate observing stretches spread over seven years, and ground-based telescopes filled in transits between them, including small observatories in Croatia, Chile, Switzerland and Australia.

The team measured two things for every transit. One was timing, pinned down to within two or three minutes and compared with the schedule a perfectly regular orbit would keep. The other was shape. A path across the middle of the star gives a long, deep dip, while a path near the rim gives a short, shallow, grazing one. Astronomers call that crossing line the impact parameter, and for TOI-2535 b it climbed steadily from one observing stretch to the next.

Reconstructed, the planet’s path seems to have swept across more than half of the star’s face, crossing the stellar equator on its way to the edge. In the latest TESS data there was no transit anywhere in the window where the planet should have shown up, even allowing for its six-hour timing swings. To test that blank result, the team planted fake transits in the data: a planet barely grazing the rim would still have been recovered about half the time.

Measurements of the star’s back-and-forth wobble with the NEID spectrograph on the WIYN 3.5-meter telescope at Kitt Peak, in Arizona, point to a planet of roughly Jupiter’s mass on the transiting orbit. They also carry extra scatter that a single planet cannot explain.

A neighbor nobody has seen

The authors modeled the timing swings and the shrinking transits together. Both fit a nearby companion planet whose orbit is tilted by at least five degrees relative to TOI-2535 b’s. When two orbits are inclined like that, their planes slowly swing around, the way the axis of a top traces a circle, and the transiting planet’s path drifts across the face of its star.

The hidden companion is probably at least as massive as Saturn. It also appears to sit in or near a resonance, meaning the two orbital periods form a simple ratio. The best-fitting solutions cluster around ratios of 2 to 1 and 3 to 2, though 3 to 1 and 4 to 1 also remain possible. That would make the system an oddity twice over: hot Jupiters rarely have close neighbors at all, and when they do, the neighbor is almost always a small planet, not a second giant.

Co-author Erik Petigura compares the effect to eclipses. “If the moon’s orbit around the Earth was perfectly aligned with the Earth’s orbit around the sun, we would see an eclipse every month,” he said. “But because the moon’s orbit is tilted, eclipses are less common.”

Faster than any planetary tilt seen before

Other planets have been caught changing their transits, but more slowly. TOI-216 b and K2-146 c went from grazing to full transits over 4.5 and three years, and a planet in the TOI-4504 system began transiting three years after astronomers first observed it. The orbital planes in TOI-4504 and TOI-216 swing around on cycles of about 250 and 150 years. The models that work for TOI-2535 point to a cycle of decades.

Its six-hour timing swing, about 3% of the orbital period, is among the largest ever seen for a short-period giant planet. Such changes are rare: one search of 561 systems from NASA’s Kepler mission found only 15 with significant long-term changes in transit duration. Compact multi-planet systems also tend to be flat, with orbits aligned within 1 or 2 degrees.

There was one earlier hint of a vanishing act. A Kepler candidate called KOI-120.01 faded and disappeared during that mission, but its system contains at least two stars, and the signal may come from a pair of eclipsing stars rather than a planet. TOI-2535 b is the cleaner case: a planet that is certainly real and still there, simply no longer lined up with us.

What the disappearance does not settle

The neighbor’s identity is the biggest open question. No single solution fits the data uniquely, so its mass, its orbital period and whether it orbits inside or outside the hot Jupiter remain unknown. Some configurations with very steep tilts, up to about 60 degrees, matched the data over short runs but tended to fall apart in long simulations, which favors gentler tilts.

The five-degree minimum tilt also depends on that unseen planet. If confirmed, TOI-2535 would join a short list of compact systems with orbits this misaligned. The host star brings its own puzzle: it is an old orange dwarf about 0.86 times the Sun’s mass, moving fast enough that it very likely belongs to the Milky Way’s thick disk, an older stellar population, though pinning down its age needs measurements of its chemistry.

Then there is formation. A hot Jupiter with a giant partner this close could not have been flung inward on a stretched-out orbit, the route thought to produce most hot Jupiters. A slow inward drift through the gas disk around the young star explains the resonance, but that disk should also have flattened the orbits. Something later must have tilted them, and the paper lists candidate mechanisms without choosing one.

“Because TOI-2535 b has a close neighbor, it probably migrated by slowly spiraling in instead of getting flung in,” Yee said. “We may be seeing a formation channel that doesn’t happen very often but is very interesting.”

Questions readers ask about the disappearing planet

What is a disappearing planet?

It is a planet whose transits, the small dips in starlight it causes when it crosses its star, stop happening. TOI-2535 b is the first confirmed case around a single star. The planet still exists; its orbit has tilted so that it no longer passes between its star and Earth.

Why did TOI-2535 b stop transiting?

The gravity of a hidden neighboring planet on a tilted orbit makes both orbits precess, or swing around like a wobbling top. Over seven years of TESS observations, the planet’s path slid from across the star’s face to its edge and then off it.

Will TOI-2535 b transit again?

Yes, but not soon. The UCLA team expects the planet to cross its star again sometime in 30 to 50 years. Its hidden neighbor may come into view sooner, if its own tilting orbit lines up with Earth first.

What comes next

TESS observed the system between April 2019 and March 2026, and the transits were missing in that final March stretch. The study was published on November 9, 2026. With only the first four years of TESS data, the team notes, the shrinking transits would not have been securely detected, and TESS’s growing baseline should turn up more systems like this one. Identifying the neighbor will take more densely sampled measurements of the star’s wobble. The planet itself will not cross its star again for 30 to 50 years, so its hidden partner may be the next thing anyone sees.

Reference: Yee et al., “A TESS Disappearing Act: The Vanishing Transits of TOI-2535 b,” The Astrophysical Journal Letters, 2026. DOI: 10.3847/2041-8213/aeb4a4

Tags: astronomy, nasa, exoplanets, hot Jupiter, TESS, TOI-2535 b

More Like This

Stellar Siblings Host ‘Teenage’ Exoplanets Dicovered by NASA’s TESS

A lava planet that should be bare rock reflects 74% of its starlight, Webb finds

James Webb found methane on a giant planet that is neither scorching nor frozen

A planet nursery was caught spinning for the first time, and parts ignore the rules

James Webb found a planet where rock clouds form each morning and clear by night

TRAPPIST-1’s two closest planets are bare rocks — a 400°C temperature gap proves it