A hundred light-years away, two Earth-sized worlds orbit a quiet red dwarf in the zone where life might take hold — and astronomers, wielding humanity's most powerful space telescope, went looking for moons. The James Webb Space Telescope brought those distant planets into sharper focus than ever before, yet the star itself refused to stay silent, its churning surface drowning out the faint whisper any moon would leave in the light. The discovery is less a failure than a reckoning: the data may already hold the answer, waiting for a mathematical mind to separate signal from stellar noise.
JWST's exomoon hunt stalled by stellar noise, but data may hold answers
The evidence could be sitting in their dataset right now, waiting for the right mathematical key.
Mark
So they pointed the most powerful telescope we have at what seemed like the best place to find a moon around another planet, and found nothing?
Mimi
They found nothing about moons, yes. But they learned a lot about the planets themselves—made measurements ten times more precise. The problem is that the star itself is so noisy that any moon signal gets buried.
Luke
How noisy are we talking?
Mimi
The star's surface turbulence creates a noise pattern of about 46 parts per million. A moon would only dim the starlight by 20 parts per million. So the noise is more than twice as loud as the signal they're looking for.
Mark
That seems like a fundamental problem. Can they just look at different stars?
Mimi
They could, but TOI-700 d and e are genuinely the best candidates we have—Earth-sized planets in the habitable zone with the right orbital stability. And the data is already collected. The researchers think if someone develops a better algorithm to filter out the stellar noise, the evidence of a moon might already be in what they've recorded.
Luke
That's a big "if." Has anyone actually built such an algorithm yet?
Mimi
Not that we know of. But the paper suggests it's theoretically possible. The noise has a pattern—it oscillates every 16 minutes. If you can model that pattern precisely enough, you might be able to subtract it out.
Mark
So they're saying the first exomoon might already be discovered, just not yet recognized?
Mimi
In this dataset, possibly. Though they can only rule out moons smaller than Ganymede on short orbits. Smaller moons or those on longer orbits would still be invisible.
Luke
How many other datasets like this exist? Are there other telescopes collecting similar data?
Mimi
JWST is the most powerful tool we have for this work. There may be other observations, but this is the most sensitive data collected so far on these particular planets.
Mark
What happens next?
Mimi
Someone needs to solve the noise problem. When they do, exomoon hunting might finally move from theory to confirmation.
The Pulse
- Researchers from MIT, Harvard, and the University of Chicago targeted TOI-700 d and e — two Earth-sized planets in a habitable zone 100 light-years away — as the most promising candidates in the search for a stabilizing exomoon.
- JWST delivered extraordinary precision on the planets themselves, sharpening orbital and radius measurements dramatically, yet returned silence on the question of moons.
- Stellar granulation — the boiling plasma on the host star's surface — produced noise at 46 parts per million, more than twice the 20 ppm signal a detectable moon would generate, effectively burying any lunar signature.
- The team can only confidently rule out moons larger than Ganymede in tight orbits; everything beyond that threshold remains invisible, hidden in the star's own restless chatter.
- A quiet hope persists: if noise-reduction algorithms can be developed to filter stellar granulation, the evidence for the first confirmed exomoon may already exist inside the data collected.
A hundred light-years away, two Earth-sized worlds orbit a quiet red dwarf in the zone where life might take hold — and astronomers, wielding humanity's most powerful space telescope, went looking for moons. The James Webb Space Telescope brought those distant planets into sharper focus than ever before, yet the star itself refused to stay silent, its churning surface drowning out the faint whisper any moon would leave in the light. The discovery is less a failure than a reckoning: the data may already hold the answer, waiting for a mathematical mind to separate signal from stellar noise.
The moon shaped Earth in ways we are only beginning to fully appreciate — steadying its axial tilt, moderating its climate, and perhaps warming its early oceans enough to kindle life. For decades, astronomers have wondered whether such a partnership exists elsewhere. A new analysis from researchers at MIT, Harvard, and the University of Chicago represents one of the most methodical attempts yet to find out, using the James Webb Space Telescope to search for moons around two Earth-sized planets orbiting the red dwarf star TOI-700, roughly 100 light-years away.
TOI-700 d and TOI-700 e both sit within the habitable zone, where liquid water might persist on a surface, and both carry enough gravitational mass to hold a moon in stable orbit. When the team pointed JWST at these worlds, the telescope performed beautifully in one respect: orbital measurements improved tenfold, and radius estimates became two to three times more precise. But on the question of moons, the data returned nothing.
The obstacle was the star itself. A moon transiting in front of TOI-700 would dim its light by roughly 20 parts per million — a signal JWST is theoretically capable of detecting. But the star's surface churns with convective plasma, producing a rhythmic red noise oscillating every 16 minutes at 46 parts per million. That stellar granulation was more than twice as loud as any lunar signal, swallowing it completely.
The team could only rule out moons smaller than Ganymede in very close orbits; anything farther out remained concealed. Yet the researchers noted a quiet possibility within the disappointment: if an algorithm could be developed to mathematically filter out the stellar noise, the data already in hand might reveal a moon hiding just beneath the surface of the star's own light. No exomoon has ever been confirmed beyond doubt, but the tools are sharpening — and the answer, some astronomers now believe, may already be waiting to be found.
The moon made Earth what it is. It steadies our planet's tilt, softens the wild swings of climate that would otherwise render the surface uninhabitable, and may have warmed the early oceans through tidal friction in ways that sparked the first life. For decades, astronomers have wondered whether such a system exists elsewhere—an Earth-like world with its own stabilizing satellite, orbiting some distant star. The search has yielded nothing definitive. But a new analysis from researchers at MIT, Harvard, and the University of Chicago, posted to the arXiv preprint server, describes a methodical attempt using the James Webb Space Telescope to find exactly that. The effort came close. It also revealed why the hunt remains so difficult.
The target was TOI-700, a small red dwarf star roughly 100 light-years from Earth, orbited by several known planets. Two of them—TOI-700 d and TOI-700 e—sit in the habitable zone, that theoretical band where liquid water might persist on a world's surface. Both are roughly Earth-sized. Both, the researchers reasoned, possessed the gravitational heft and orbital stability needed to hold a moon in place. If any exoplanet system was going to host a detectable moon, this seemed like the one.
When Emily Pass and her colleagues pointed JWST at these worlds, the telescope delivered on its promise in one sense. The observations sharpened the team's understanding of the planets themselves dramatically. Orbital measurements improved by a factor of ten. Radius estimates—1.145 times Earth's size for TOI-700 d, and 0.919 times Earth's size for TOI-700 e—became two to three times more precise. The data was exquisite. But it revealed nothing about moons.
Theoretically, JWST should have been able to spot a moon the size of Earth's own Luna. The calculation was straightforward: a moon passing in front of its host star would dim the starlight by roughly 20 parts per million. JWST's instruments are sensitive enough to catch such a dip. The problem was that the star itself was drowning out the signal. As the researchers analyzed their data, they found a repeating pattern of noise—what astronomers call "red noise"—oscillating every 16 minutes with an amplitude of about 46 parts per million. This noise arose from stellar granulation, the roiling and churning of plasma on the star's surface, the same convective turbulence that creates sunspots. At 46 parts per million, this stellar chatter was more than twice as loud as the 20 parts per million whisper a moon would make. Any lunar signal was simply lost in the din.
The constraint this imposed was sobering. The team could only rule out moons smaller than Ganymede, the largest moon in our solar system, if they orbited closer than two days to their planet. Anything beyond that remained hidden. Yet there was a glimmer of possibility in the disappointment. The researchers realized that if someone developed an algorithm capable of filtering out the red noise—mathematically separating the stellar chatter from genuine planetary signals—the data they had already collected might reveal a moon. The evidence could be sitting in their dataset right now, waiting for the right mathematical key to unlock it.
This is not a new problem in exomoon hunting. Previous searches have been derailed by similar obstacles: sunspots masquerading as moons, statistical flukes that evaporated under scrutiny, instrumental artifacts mistaken for real discoveries. The field has learned to be cautious. Some astronomers have begun focusing their search strategies specifically on finding ways to eliminate stellar noise, treating it as the central challenge rather than a peripheral nuisance. For now, no exomoon has been confirmed beyond doubt. But the tools are improving, and the data is accumulating. When someone finally develops the right noise-reduction technique—and astronomers and software engineers tend to be resourceful—the answer may already be waiting in observations like these, hidden just beneath the surface of the star's own restless light.
Notable Quotes
The researchers determined that if someone developed a noise-reducing algorithm and a moon actually exists in the system, the dataset should reveal it.— Emily Pass and colleagues, MIT/Harvard/University of Chicago
Want the full story?
Read the original at Phys.org ↗
Education