Posted by What changed for exoplanet astronomy after the James Webb telescope launch was not the pictures. It was the shortlist. Webb spends much of its exoplanet time on a small number of rocky worlds around small stars, measuring whether they have any atmosphere at all, and the results so far have been sobering, useful and considerably less romantic than the press releases suggested.
That is what studying potentially habitable worlds actually looks like from the inside: long exposures, contested error bars, and a steadily growing list of planets that can be crossed off.
What the Telescope Was Actually Given
Webb reached its station near the Sun-Earth L2 point in early 2022 and started science that July. Its advantage over everything before it is infrared sensitivity — a 6.5-metre mirror kept below fifty kelvin behind a sunshield the size of a tennis court, feeding instruments that can hold their calibration through hours of staring at a single star.
Infrared is where the interesting molecules live. Water, carbon dioxide, methane, sulfur dioxide and ammonia all absorb strongly at those wavelengths, so a spectrum taken in the infrared carries chemical information that visible light simply does not. Everything that follows depends on that one design choice.
The Seven Planets Everyone Wanted First
TRAPPIST-1 was always going to be the first serious target. An ultracool dwarf about forty light years away, it hosts seven roughly Earth-sized planets, several of them in the temperature range where liquid water is at least conceivable. No other known system offers so many comparable rocky worlds around one star, which lets astronomers compare planets without also having to account for different stellar environments.
The first results arrived in 2023 and were not what the optimistic version of the story predicted. Using MIRI to measure thermal emission at fifteen microns, one team found the innermost planet, TRAPPIST-1b, radiating essentially like a bare rock — a dayside near 500 kelvin, with no sign of an atmosphere thick enough to move heat toward the night side. TRAPPIST-1c followed, with limits that ruled out a substantial carbon dioxide envelope of the kind Venus wears.
LHS 475b, an Earth-sized planet around a different red dwarf, produced a similar verdict in January 2023: no thick methane-dominated atmosphere, and probably very little air of any kind.
None of that closes the question for the outer TRAPPIST-1 planets, which are cooler and harder to measure and where the observing campaigns are still running. But a pattern is forming, and it points at the star.
It is worth being blunt about the cost of these measurements. A single confident statement about one small planet can consume tens of hours of the most oversubscribed telescope in astronomy, spread across many transits, and the analysis afterwards takes months. Webb’s exoplanet programme is not a survey. It is a handful of extremely expensive case studies, chosen carefully because there is no room to choose badly.
The Problem With Small Stars
Red dwarfs are the most common stars in the galaxy and by far the easiest hosts for this kind of work. They are small, so an Earth-sized planet crossing one blocks a comparatively large fraction of its light. They are dim, so the temperate zone sits close in and transits repeat every few days instead of once a year.
The same proximity is the hazard. Young red dwarfs spend hundreds of millions of years throwing flares and extreme-ultraviolet radiation at whatever orbits them, and a planet close enough to be warm may have had its atmosphere eroded into space long before it settled down. Whether any rocky planet around an M dwarf keeps its air is now one of the defining questions of the field, and Webb was pointed at it deliberately.
A negative answer would matter enormously. Most of the temperate rocky planets we can study in the next twenty years orbit stars like these.
Two Ways to Squeeze Chemistry Out of Starlight
Transmission spectroscopy is the better-known method. During a transit, a thin ring of starlight passes through the planet’s atmosphere and picks up absorption features on the way, so subtracting the ordinary stellar spectrum leaves the atmosphere behind. The effect is often a few hundred parts per million, which is why these observations demand dozens of hours and repeated transits.
Secondary eclipse photometry works differently and has quietly become the workhorse for small planets. Instead of watching the planet pass in front of the star, you watch it pass behind. The tiny drop in total infrared brightness is the planet’s own thermal emission, and how hot the dayside is tells you whether winds are redistributing heat — which requires an atmosphere — or whether the surface is simply baking in place.
Four instruments share the focal plane, and which one gets used depends on the question. NIRSpec and NIRISS handle the near-infrared where water and methane show up most clearly. NIRCam does imaging and some spectroscopy. MIRI reaches out to the mid-infrared, cooled to around seven kelvin by its own cryocooler, and it is the instrument that measures the thermal glow of a rocky planet at fifteen microns — the measurement that decided the TRAPPIST-1b question.
Webb also looks at protoplanetary disks, the dusty rings around young stars where the water and carbon compounds a planet inherits are still being sorted by distance from the star. It is a different timescale for the same question: how ordinary are the ingredients?
K2-18b and the Price of Excitement
The most argued-over result of the mission so far concerns K2-18b, a planet about eight times Earth’s mass orbiting a red dwarf roughly 120 light years away. Webb detected methane and carbon dioxide in its atmosphere in 2023, which was solid work. The same analysis reported a tentative hint of dimethyl sulfide, a gas made on Earth almost entirely by marine microorganisms.
The coverage was enormous and the caveats travelled badly. Independent groups reanalysed the data and disagreed about whether the feature was there at all; later observations sharpened the argument without settling it. Even the planet’s basic nature is unresolved, since a mini-Neptune with a deep hydrogen envelope and a water world with a global ocean can produce uncomfortably similar spectra.
The episode is a fair illustration of where the science stands. Webb can measure the chemistry. Deciding what the chemistry means requires stellar characterisation, atmospheric modelling, independent instruments and a willingness to be wrong in public.
What Habitable Actually Requires
The habitable zone — the orbital band where surface liquid water is possible given a suitable atmosphere — is a filter for building target lists, not a description of a planet. Venus sits near its inner edge and has a surface at 460 degrees Celsius. Mars sits near the outer edge and lost nearly all its air.
Mass, atmospheric pressure, cloud cover, interior heat, magnetic shielding and stellar activity all matter, and most of them are unmeasurable for planets this far away. What Webb contributes is a way to replace one assumption — a planet at the right distance is probably temperate — with an actual observation, even when that observation says the planet has no atmosphere and the assumption was wrong.
Null Results Are Doing the Real Work
A finding that a planet is airless reads as a disappointment and functions as data. It constrains how much atmosphere red dwarf planets retain, it tests models of atmospheric escape, and it stops the community from spending hundreds of hours on a target that cannot deliver.
It also shapes what gets built next. Direct imaging of a genuinely Earth-like planet around a Sun-like star is beyond Webb’s reach; that job belongs to future observatories, and the target lists they will use are being assembled now, one tedious spectrum at a time. The James Webb telescope launch bought the survey work that makes the next instrument worth designing.
If you want to follow that work rather than the headlines about it, SETIworld keeps track of the campaigns, the disputed detections and the quiet non-detections that end up mattering more.
