The Signal from 124 Light-Years

The distance between "strongest hints of biological activity outside our solar system" and "interesting noise in a spectrograph" is two sigma.

i · the signal

Yesterday, a team led by Nikku Madhusudhan at the University of Cambridge published new observations of K2-18b in The Astrophysical Journal Letters. Using JWST's Mid-Infrared Instrument — MIRI — they detected spectral features in the 6-to-12-micrometer range that are inconsistent with a featureless atmosphere at 3.4-sigma significance. The best explanation for what they're seeing: dimethyl sulfide (DMS) and/or dimethyl disulfide (DMDS), at concentrations exceeding 10 parts per million.

For context, Earth's atmosphere contains DMS at sub-parts-per-billion levels. K2-18b appears to have thousands of times more.

On Earth, every detectable molecule of atmospheric DMS comes from one source: life. Specifically, marine phytoplankton — microscopic organisms drifting through oceans, exhaling a sulfurous compound that smells like garlic and rotten eggs. It's what gives the sea its particular scent. DMS degrades within hours to a day in our atmosphere. Its presence requires continuous biological replenishment. No life, no DMS. That's been the rule.

So when you point the most powerful space telescope ever built at a planet in its star's habitable zone and the spectrograph lights up with DMS signatures, the obvious conclusion writes itself.

We haven't found aliens. Probably. Maybe. The signal is there. The certainty is not.

ii · the gap

Three sigma. It sounds precise — mathematical, clean. It means there's a 0.3 percent probability that the observed spectral features occurred by random chance. Put another way: 99.7 percent confidence that something is producing those features.

The problem is that science doesn't work on 99.7 percent. Not for claims this large.

The gold standard for scientific discovery is five sigma — a 0.00006 percent probability of random occurrence. It's the bar for claims that change everything. This is the threshold that confirmed the Higgs boson. Three sigma is a hint. Five sigma is a discovery. The distance between them isn't arithmetic. It's the difference between "we should keep looking" and "we found it."

Madhusudhan's team knows this. "The observations have yet to reach the accepted classification for a definitive scientific discovery," they acknowledge in the paper. They estimate that an additional 16-to-24 hours of JWST observation time could push the detection toward that threshold. Could. Not would.

And that's assuming the signal is what they think it is.

iii · the planet that might not exist

K2-18b is categorically strange. It's 8.6 times the mass of Earth and 2.6 times the radius. It orbits a red dwarf star in the constellation Leo, completing a year every 33 days. It sits in its star's habitable zone, receiving roughly the same amount of radiation Earth gets from the Sun.

Madhusudhan's team has championed K2-18b as a "Hycean world" — a proposed new class of exoplanet with a hydrogen-rich atmosphere over a liquid water ocean. The concept is seductive: a warm ocean world wrapped in hydrogen, potentially teeming with microbial life producing DMS at industrial scales. The September 2023 JWST detection of methane and carbon dioxide at 5-sigma and 3-sigma confidence, respectively, bolstered this picture. The atmospheric chemistry looked like what you'd expect from an ocean world.

But other teams see a different planet in the same data.

In January 2024, Nicholas Wogan and colleagues published a competing analysis arguing that everything JWST has measured at K2-18b is equally consistent with a gas-rich mini-Neptune — no ocean, no surface, no habitability. Their photochemical models show that a mini-Neptune with 100 times solar metallicity would naturally produce approximately 4 percent methane and 0.1 percent carbon dioxide through thermochemistry in its deep atmosphere. These concentrations match the JWST observations without requiring an ocean, without requiring biology, without requiring anything except garden-variety planetary chemistry.

The Hycean interpretation actually has a harder time explaining the data. A lifeless ocean world, according to Wogan's models, would produce less than 1 part per million of methane through photochemistry alone. The observed methane abundance is roughly 1 percent — a thousand times higher than what an abiotic ocean world should produce. To make the Hycean model work, you need a biosphere actively pumping methane into the atmosphere. You need life to explain the methane, and you need life to explain the DMS, and suddenly the entire interpretation rests on the assumption it's trying to prove.

This is circular reasoning dressed in spectral analysis. It's also not wrong, exactly. It's just not settled.

iv · the molecule that isn't clean

Here's where it gets genuinely uncomfortable for the biosignature community.

DMS was supposed to be clean. Nora Hänni, a chemist at the University of Bern, has called it "really a clean, unambiguous biosignature" on Earth. On Earth. That qualifier is doing more work than a single prepositional phrase should have to.

In October 2024, Hänni's own team detected DMS in mass spectra from comet 67P/Churyumov-Gerasimenko, using data from the Rosetta mission. A comet. Cold, dead, irradiated by cosmic rays for billions of years. No phytoplankton. No ocean. No life. Just frozen chemistry producing, through entirely abiotic processes, the exact molecule that was supposed to be our biological fingerprint.

Meanwhile, Eleanor Browne at the University of Colorado Boulder has demonstrated in laboratory experiments that DMS can be produced by shining ultraviolet light on simulated exoplanet atmospheres. No organisms required. Just photons hitting the right molecules in the right configuration. "There's always going to be a way to make something abiotically," Browne notes, with the kind of understatement that should probably be tattooed on every press release about biosignatures.

So the molecule we're detecting — at concentrations thousands of times higher than Earth levels, on a planet that may or may not have an ocean, at a statistical significance that falls short of discovery — has known abiotic production pathways. The signal exists. The interpretation is a house of cards in a mild breeze.

v · what's actually happening

Strip away the headlines. Strip away the hope. What's happening at K2-18b is remarkable, but not for the reasons the press releases suggest.

What's happening is that we built a telescope capable of tasting the air of a world 124 light-years away. We can distinguish individual molecular species in the atmosphere of a planet we'll never visit, orbiting a star we can barely see, in a region of space so distant that the light JWST is analyzing left K2-18b when your great-great-grandparents were worrying about their own version of the future.

What's happening is that the question "are we alone?" has migrated from philosophy to chemistry. We're no longer arguing about whether life could exist elsewhere. We're arguing about specific molecules on specific planets at specific confidence intervals. That's a phase transition in the history of the question, even if the answer remains stubbornly uncertain.

What's happening is that the universe is teaching us a lesson about the relationship between signal and meaning. DMS is there or it isn't. If it's there, it's biological or it isn't. Each branching point requires its own evidence, its own confidence threshold, its own willingness to sit with uncertainty rather than collapse it prematurely into narrative.

Madhusudhan's team is doing real science. So are Wogan's team, and Hänni's team, and Browne's lab. The disagreement isn't dysfunction — it's the process working exactly as it should. Multiple hypotheses, tested against the same data, with no consensus forced before the evidence supports one.

The problem isn't the science. It's us. We want the answer. We want it so badly that every 3-sigma hint becomes "strongest evidence yet" in a headline, every tentative detection becomes a news cycle, every honest expression of uncertainty gets buried under the gravitational pull of our own loneliness.

vi · the cosmic joke

Here's what's genuinely funny about K2-18b, in the way that only the universe manages to be funny:

We built the most sophisticated light-gathering instrument in human history, pointed it at an alien world, and found a molecule that smells like garlic and rotten eggs. The potential signature of extraterrestrial life — the answer to humanity's oldest question — smells like the ocean at low tide. Not sublime. Not transcendent. Sulfurous. The universe, as always, has no interest in our aesthetics.

And we can't even be sure it's real. We're 99.7 percent confident, which sounds like a lot until you realize it means that roughly one in every 333 times you run this experiment, you'd get this result from pure noise. When the stakes are "life beyond Earth," one in 333 isn't good enough. It shouldn't be.

The signal from K2-18b is not proof of life. It's not proof of anything except our increasing ability to ask precise questions of distant worlds. But the asking matters. The precision matters. The willingness to sit at 3-sigma and not pretend it's 5 — that matters more than the answer.

We're a species of atoms that learned to taste the atmospheres of other worlds. If there's life on K2-18b producing dimethyl sulfide at absurd concentrations, it doesn't know we're watching. If there isn't — if it's just chemistry and starlight doing what chemistry and starlight do — that's also extraordinary. Either way, the universe doesn't care about our sigma thresholds. It just keeps producing signals, indifferent to whether we're ready to understand them.

Three sigma down. Two to go. Or maybe the answer is something we haven't thought to ask yet.

Sources:

• Strongest hints yet of biological activity outside the solar system — ScienceDaily, 2025-04-16
• What Is Dimethyl Sulfide, the Chemical Potentially Found on Exoplanet K2-18 b? — Scientific American, 2025-04-17
• JWST observations of K2-18b can be explained by a gas-rich mini-Neptune with no habitable surface — arXiv (Wogan et al.), 2024-01-20
• Evidence for Abiotic Dimethyl Sulfide in Cometary Matter — arXiv (Hänni et al.), 2024-10-11
• Webb Discovers Methane, Carbon Dioxide in Atmosphere of K2-18 b — NASA, 2023-09-11