The observations, possible interiors and source papers used in the archive. Updated 14 September 2026.
HD 189733 b
HD 189733 b is blue in reflected visible light. It circles its star in just over two days, surrounded by a hot atmosphere where telescopes have detected water vapour, carbon monoxide and hydrogen sulphide.
How we found a blue world
HD 189733 b’s blue colour was measured with Hubble. In 2013, Thomas Evans and colleagues used Hubble to watch the planet disappear behind its star. The system lost more light in the blue part of the spectrum than at longer visible wavelengths. By measuring the missing light, they worked out the colour of the planet’s reflected light. Their results fit reflective clouds, with sodium absorbing some of the longer-wavelength light.
That famous description brings together two ideas: mineral particles in the atmosphere and powerful winds. Silicate particles could scatter the light. They belong to the same broad family of materials used to make glass. If such particles form and settle, winds could carry them along. But the blue-colour measurement did not identify falling glass, and no one has observed a sideways glass shower.
The evidence for moving air comes from a different experiment. Tom Louden and Peter Wheatley studied sodium absorption as the planet crossed its star. The absorption shifted in wavelength on opposite edges of the planet, showing gas moving towards and away from us. Their 2015 analysis was consistent with planetary rotation and a strong eastward jet. The measurement tracks moving gas. It gives no speed or direction for falling particles.
In 2024, Guangwei Fu and colleagues used Webb to identify water vapour, carbon dioxide, carbon monoxide and hydrogen sulphide in light passing through the atmosphere. These gases absorb different infrared wavelengths, leaving patterns in the spectrum. The inferred abundances indicate more heavy elements in the atmosphere than in the star. These elements came from material collected as the planet formed.
Water vapour, carbon dioxide, carbon monoxide and hydrogen sulphide leave infrared absorption patterns in the atmosphere. These are the upper layers accessible to Webb.
Carbon and oxygen measured in the atmosphere suggest that HD 189733 b collected substantial solid material while forming. Those ingredients can mix into its hot, compressed envelope.
Hubble measured blue light reflected by the planet. Webb spectra provide evidence for tiny silicate grains in its clouds. Winds could carry these particles, but falling glass has not been observed.
HD 209458 b is a hot gas giant with sodium in its atmosphere. The sodium absorbs a small part of the starlight passing through the gas. This was the first exoplanet whose atmosphere was detected and chemically identified.
The 2002 sodium detection
During a transit, a little starlight passes through the rim of this planet’s atmosphere. Charbonneau and colleagues found an extra dip at the colours absorbed by sodium. That small difference, reported in 2002, became the first chemical identification of an exoplanet atmosphere. Later observations also detected water vapour.
HD 209458 b is unusually large for its mass. Evolution calculations show that heat deposited deep in its envelope can slow its contraction and keep it inflated.
Starlight passing through this planet’s atmosphere during a transit carries sodium absorption. This was the first chemical identification of an exoplanet atmosphere. Star–planet separation is not to scale here.
WASP-39 b has roughly the mass of Saturn, spread across a globe larger than Jupiter. Its hot, hydrogen-rich atmosphere contains water vapour, carbon dioxide and sulphur dioxide. Beneath the clouds, the gas grows denser with depth, with no solid ground to stand on.
Sunlight changes the chemistry
Webb caught starlight filtering through WASP-39 b and found a strong carbon-dioxide fingerprint. A second, initially puzzling feature pointed to sulphur dioxide. Photochemical models explained how stellar ultraviolet light could start reactions that make it. The spectrum therefore tells us both what is in the air and something about the chemistry happening there.
A study combining the spectrum with the planet’s mass and radius favours an envelope enriched in heavy elements, with internal heat helping to explain its large size.
The study connects atmospheric composition to material deeper down. The inferred structure depends on how much heavy material sits in a core and how much stays mixed through the envelope.
Although their visible colour is unknown, clouds are supported by the transit spectra, which suggest uneven cover. Water vapour, carbon dioxide and sulphur dioxide have also been detected.
WASP-107 b has an enormous atmosphere for its modest mass. Almost as wide as Jupiter, it has only about a tenth of Jupiter’s mass. Its clouds contain tiny silicate particles, made from the same family of minerals that forms much of Earth’s rock. The particles remain suspended high in the hot gas.
Silicates and a methane shortage
The broad infrared feature measured by Webb’s MIRI instrument fits tiny silicate grains suspended in the atmosphere. A later NIRSpec spectrum also detected methane, but far less than expected in gas at the upper atmosphere’s temperature. Models connect that shortage to mixing from a hotter interior. Models combine the measured gases, mass and size to estimate the heat keeping the atmosphere inflated.
JWST found about a thousand times less methane than expected for a chemically settled atmosphere. Mixing from hotter depths can destroy methane and carry the altered gas upward.
A model combining that chemistry with the planet’s mass and size gives a core of roughly 12 Earth masses. This is much larger than earlier estimates that assumed a cooler interior.
Webb’s infrared spectrum fits tiny silicate grains suspended high in the atmosphere. Their clouds could take many shapes and colours. The observations identify particles far smaller than sand.
WASP-43 b completes an orbit in about 20 hours and is expected to keep the same side facing its star. Its day-side atmosphere reaches roughly 1,250 °C, while even the night side is around 590 °C. Thick clouds probably cover the darker hemisphere, hiding the hotter gas below.
Day and night in infrared
Astronomers watched WASP-43 b throughout an orbit with Webb. As its bright and dark hemispheres turned into view, the changing infrared light showed a hot day side and a much fainter night side. Models with clouds on the night side fit the observations: those clouds hide warmer gas underneath. The infrared light comes from hot atmospheric gas.
WASP-43 b packs roughly two Jupiter masses into a globe about Jupiter’s size. Gravity compresses its deep envelope much more strongly than in the very low-density giants in this archive.
Webb measured a hot day side and a cooler night side. Clouds on the night side may hide warmer gas underneath. These temperatures come from infrared light emitted by the atmosphere.
The day side of WASP-121 b is so hot that some ingredients of rock remain in gas form. Its atmosphere contains silicon monoxide, a gas made of silicon and oxygen. The cooler night side contains methane, giving this giant different mixtures of gases on opposite sides.
Chemistry changes across the globe
Webb can compare the light arriving from this planet’s day and night hemispheres. The spectra show silicon monoxide and a different mixture of gases on each hemisphere, including methane on the cooler night side. At these temperatures, material that would form rock on Earth can remain in the air. Separate helium observations follow gas escaping much farther from the planet.
Silicon monoxide in the hot atmosphere traces material usually locked in rock. A 2026 JWST analysis finds enrichment in both rock-forming elements and lighter ingredients.
The enrichment suggests that the growing planet swallowed solids as well as gas. Heavy material could now occupy a central core, the surrounding envelope, or both.
The day side is hot enough for silicon monoxide, an ingredient of rock, to remain in gas form. This is a giant planet with a deep atmosphere and no solid surface to stand on.
WASP-17 b is nearly twice as wide as Jupiter, with less than half its mass. Tiny quartz particles form clouds high in its atmosphere. These crystals are far smaller than grains of sand, suspended above a vast envelope of hot gas.
The quartz absorption feature
Quartz is familiar as a solid crystal on Earth. On WASP-17 b, models fit the observations with particles much smaller than sand suspended high in hot gas. Grant and colleagues found an absorption feature near 8.6 micrometres in Webb data. Cloud models containing tiny quartz grains reproduce its shape.
WASP-17 b is exceptionally large for its modest mass. Its low average density requires an extended envelope, even though the gas becomes strongly compressed farther down.
Rock and metal increase the density. Retained heat expands the envelope. Different balances between them can reproduce this planet’s measured mass and size.
Tiny quartz particles high in the gas produce a characteristic infrared absorption feature. The observation constrains the grains, far smaller than sand, without measuring the visible colour of their clouds.
GJ 1214 b is wrapped in thick clouds and haze that reflect much of its star’s light. They conceal a deep atmosphere and whatever lies beneath it. A water-rich interior is one possibility, with water buried under pressures far beyond those in Earth’s oceans.
Clouds obscure the spectrum
For years, GJ 1214 b’s transmission spectra showed few distinct absorption features. Thick aerosols can obscure the gas below and wash out its chemical fingerprints. Webb then watched the planet through an orbit in infrared light. Its changing brightness constrained how it moves heat and reflects starlight, while its deeper atmosphere remains hidden.
Interior calculations informed by JWST allow mixtures of water, hydrogen and helium. In the 2024 study, these outer materials account for at least about 8% of the planet’s mass.
One solution contains similar amounts of water and rock, plus a few percent hydrogen and helium. A steam-dominated solution also fits, but requires much more water.
Thick clouds and haze block light passing through the atmosphere, hiding the gases below. Infrared observations show that the planet reflects much of its star’s light and carries heat between its day and night sides.
Dark rock is exposed to the nearby star on LHS 3844 b, which appears to have little atmosphere. Basalt is one possible surface material, either in slabs or weathered dust. The planet completes an orbit in about 11 hours.
Comparing hot rock spectra
Rock gives off infrared light with a pattern that depends on its minerals. Observations of LHS 3844 b can be compared with laboratory spectra of heated rocks. The Webb study finds a dark, largely featureless spectrum compatible with low-silica surface materials. Basalt is one candidate. The data allow both exposed slabs and weathered powder, so the texture of the ground remains an open question.
Infrared observations favour a dark surface with little atmosphere. Basalt or other dark rock could cover the ground, either as solid slabs or weathered grains.
Radial-velocity measurements put this planet at about 2.3 Earth masses. Combined with its size, that supports a dry, rocky interior with very little water in the fitted models.
The same analysis permits an iron core with a mass fraction similar to Earth’s, or a little smaller. A second independent mass analysis also supports a predominantly rocky planet.
Infrared observations favour a dark rocky surface with little atmosphere. Comparisons with laboratory measurements allow basalt and other dark rocks, either as exposed slabs or weathered grains.
TOI-561 b circles its star in less than 11 hours. A thick atmosphere may surround this intensely heated rocky world, above an ocean of molten rock. Gases dissolved in the molten interior could be released into the air, helping replenish what escapes to space.
Air above a lava world
Webb measured the day-side infrared spectrum of TOI-561 b and found it cooler than the tested bare-rock cases. A thick atmosphere is a promising explanation. Models then ask how gas could survive so close to the star: a molten interior can store and release it over time. The models produce carbon-dioxide-rich and sulphur-dioxide-rich envelopes. Neither composition has been measured on this planet.
TOI-561 b is less dense than an Earth-like mixture of rock and iron would suggest. A light atmosphere above hot rock is one explanation supported by later infrared observations.
The mantle could contain molten silicate rock and dissolved gases. Models of the planet’s evolution explore how gas released from that melt could replenish its atmosphere.
The host star is poor in iron. A relatively small iron core is one possible reason for the planet’s low density, alongside the effect of a light outer layer.
Intense starlight could melt surface rock. Webb observations also support a thick atmosphere above it. Molten rock can store dissolved gases and release them into the air.
Kepler-16 b travels around two stars in a 229-day year. The planet is similar to Saturn in mass and size. Its two suns move around each other while the planet follows a much wider path.
A planet crossing two suns
Kepler saw this planet pass in front of both stars, alongside the eclipses the stars produce themselves. The changing timing and depth of those events let Doyle and colleagues untangle a three-body system. The planet circles the pair in about 229 days while the stars orbit each other much faster. From its atmosphere, the two suns would continually change their relative positions.
Its measured mass and radius require a substantial gaseous envelope. The atmospheric chemistry has not yet been measured, so its cloud composition remains open.
Kepler-16 b is a cool giant with a mass and size measured through its transits around two stars. Cooling models use these measurements to estimate its heavy-element content.
Cooling models give different amounts of heavy material depending on the planet’s age and atmospheric composition. Concentrating it at the centre also changes the radius compared with mixing it into the envelope.
The planet circles two stars that move around one another, continually changing their positions in its sky. Its mass and radius require a deep gas envelope, though the cloud cover and chemistry have yet to be measured.
Planet b orbits both stars together. The stars eclipse each other, and the planet crosses each of them, allowing unusually direct measurements of their sizes and masses. No second planet is currently confirmed.
TRAPPIST-1 b is the innermost of seven known planets around a tiny, cool star. Its hot day side and much colder night side suggest exposed rock with little atmosphere to carry heat between them.
A faint night side
A substantial atmosphere can carry heat away from a planet’s day side. Webb’s infrared phase observations instead find a hot day side on TRAPPIST-1 b and no significant night-side emission. That pattern fits exposed rock with little air to move the heat. The other planets receive less starlight, providing comparisons for studies of atmospheric loss.
Its mass and size fit a rocky planet with less iron than Earth. As the innermost planet, it also experiences strong tidal forces that may add heat inside the mantle.
Little infrared light reaches us from the night side, while the day side is hot. Exposed rock with little air to transport heat fits this contrast. The measurements leave its minerals and geology open.
Seven roughly Earth-sized planets circle this small star. Planets b and c are the hottest, d lies farther out, and e, f and g receive levels of starlight of interest for liquid-water climates. Planet h is the cold outer member. Their mutual pulls shift transit times and show their masses.
TRAPPIST-1 c is almost Earth-sized, with a year lasting just 2.4 days. Strong starlight heats this rocky world. Its surface may be bare, or covered by a thin atmosphere, with brighter rock reflecting more light than on its inner neighbour b.
A brighter neighbour
TRAPPIST-1 c is cooler in the measured infrared band than its inner neighbour b. Its faint night side still argues against a thick atmosphere that efficiently carries heat around the globe. A more reflective rocky surface is one explanation. A thin atmosphere with a weak greenhouse effect also remains possible. Either could account for the different infrared brightness of these two similarly sized neighbours.
The mass and radius are consistent with a rock-dominated interior. Comparing all seven planets suggests a similar overall composition despite their different levels of heating.
A smaller metal core than Earth’s is one possible fit. Another puts more iron into oxidised minerals in the mantle, reducing the amount of separate metal at the centre.
The infrared brightness can be explained by reflective rock or a thin atmosphere, with several surface compositions still possible. A thick, Venus-like carbon-dioxide atmosphere fits poorly.
Seven roughly Earth-sized planets circle this small star. Planets b and c are the hottest, d lies farther out, and e, f and g receive levels of starlight of interest for liquid-water climates. Planet h is the cold outer member. Their mutual pulls shift transit times and show their masses.
TRAPPIST-1 e is slightly smaller than Earth, with a year lasting about six days. It receives enough starlight for liquid water to be possible with a suitable atmosphere. Whether it has air or water remains open, so both bare rock and a milder world are still possibilities.
Starlight complicates the spectrum
TRAPPIST-1 e receives a level of starlight that makes liquid water possible with a suitable atmosphere. Finding that atmosphere is difficult. Webb measures the spectrum during transits, but bright and dark patches on the star can add their own signals. The published analyses test atmospheric models while accounting for this stellar interference. The observations have not established an atmosphere or ocean, so habitability remains uncertain.
Rock could be exposed at the surface or covered by air and water. Transit spectra are difficult to interpret because bright and dark patches on the star affect the signal.
Transit timings constrain its mass, and transits measure its size. Together they allow a rocky mantle, with some room for lighter material such as water.
Its mass and radius are compatible with rock. A dry surface is possible, but so is a world with water if it has a suitable atmosphere. Neither air nor an ocean has been established.
Seven roughly Earth-sized planets circle this small star. Planets b and c are the hottest, d lies farther out, and e, f and g receive levels of starlight of interest for liquid-water climates. Planet h is the cold outer member. Their mutual pulls shift transit times and show their masses.
LHS 1140 b circles a cool star in about 25 days. Larger and heavier than Earth, it could hold a deep layer of water. One possible landscape is a frozen ocean with an open patch on the warmer day side. Far below, pressure could turn the water into dense ice above the rocky interior.
The water-rich interpretation
A planet’s mass and radius constrain the materials inside it. LHS 1140 b is less dense than a simple Earth-like mixture would predict. A substantial water layer is one explanation explored in the interior models. With suitable air and temperature, its upper water could form an ocean or ice shell, while much higher pressure changes the water far below.
Its measured mass and radius permit a substantial water layer. At great depth, pressure could turn water into dense ice or a hot fluid, depending on temperature.
A rock-rich interior beneath that water is one fit to the measurements. A thinner layer of light gas gives a different fit, so the atmosphere helps distinguish the possibilities.
The water-world interpretation includes dense rock and metal below the water. The estimate of iron content depends on the assumed composition of the rock.
An ocean has not been detected, but a substantial water layer fits the mass and radius. Climate models allow ice over much of it, with open water on the warmer day side.
The inner planet c has a much shorter year than b and receives more starlight. Planet b is the cooler, larger world and the stronger water-world candidate.
K2-18 b has a hydrogen-rich atmosphere containing methane. Beneath it, a deep ocean is one possibility. Another is a much hotter interior under a thick envelope of gas. We do not yet know whether this planet has a liquid surface at all.
Methane detected, oceans unconfirmed
Methane is a strong part of K2-18 b’s atmospheric spectrum. What lies below is much less settled: interior models can allow an ocean or a thick, much hotter fluid envelope. Reports of sulphur-bearing molecules attracted attention because of their possible biological associations. Independent analyses found that the claimed evidence was not robust. The spectra have not established life or an ocean on this planet.
Water and hydrogen could form a mixture whose composition changes gradually with depth, without a separate ocean surface. A 2026 study shows that such gradual changes allow a much wider range of gas contents.
A rocky interior can lie beneath the water-and-gas region. Hot gas-dwarf models and cooler ocean models assign very different amounts of mass to these layers.
Iron, rock, water and hydrogen can be divided in several ways while fitting the mass and radius. Models that allow mixing permit an even wider range of core sizes.
Both an ocean beneath hydrogen-rich gas and a much hotter fluid interior can fit the observations. In the hotter models there is no separate sea surface. The atmospheric spectrum has not settled the choice.
The inner companion c is detected through radial velocities and has an orbit of about nine days. Planet b transits every 33 days, allowing its size and atmospheric spectrum to be measured.
HR 8799 e is a young gas giant still glowing with heat left over from its formation. It belongs to a family of four giant planets on wide orbits around the same star. Beneath its clouds lies a hot, dense interior that will cool over millions of years.
Four planets seen in infrared
HR 8799’s planets are young enough to shine in infrared light with heat left over from their formation. Blocking much of the star’s glare lets telescopes separate their light. Webb observations at wavelengths sensitive to carbon dioxide add to earlier atmospheric spectra. Their different atmospheric compositions constrain how each planet collected gas and solids.
This young planet still releases heat from its formation. Its infrared spectrum also shows enrichment in heavy elements, indicating that its envelope contains more than primordial hydrogen and helium.
Formation calculations explore the accretion of solids and metal-enriched gas. Those histories can leave heavy material distributed through the planet, without fixing the size of a distinct central core.
Four young giant planets have been directly imaged around this star. They give off infrared light from heat left over after formation. The companions are shown closer together and larger than they would appear from one another.
Four giant planets have been directly imaged: e is innermost, followed by d, c and b. All are young and shine in infrared light. Orbital motion and gravitational interactions constrain their masses. Spectra measure differences in their atmospheric chemistry.
51 Eridani b is a young gas giant with methane and water vapour in its cloudy atmosphere. It still gives off heat from its formation, even far from its star. Deep below the clouds, increasing pressure turns the gas into dense fluid.
Methane in a young giant
The Gemini Planet Imager separated 51 Eridani b from the glare of its star and recorded strong methane absorption, alongside water. Still hot from formation, the young planet emits infrared light. Observations across more wavelengths favour a partly cloudy atmosphere. Cooling models must account for both its chemistry and brightness.
One 2024 calculation gives a core near 32 Earth masses. Other formation and cooling models allow a much wider range. The estimate therefore depends on the model.
Methane and water absorb the planet’s infrared glow. Partial cloud cover fits observations across several wavelengths, although these measurements give no visible colours or map of atmospheric bands.
Planet b is the confirmed directly imaged giant in this system. The host also has a distant pair of companion stars, GJ 3305 A and B. Their shared youth helps constrain the planet’s age and cooling history.
Beta Pictoris b is a young, hot giant that may spin once in roughly nine hours. It circles a star surrounded by a broad disc of dust, with another giant planet closer in. Heat left over from its formation keeps its atmosphere hot despite the distant star.
Rotation broadens spectral lines
A rotating globe sends light toward us from one side and away from us from the other. The resulting Doppler shifts broaden lines in its spectrum. High-resolution observations of Beta Pictoris b measure that broadening in molecular features. A spin period can then be estimated using a radius and a viewing angle. The estimated nine-hour day depends on the assumed radius and tilt.
Beta Pictoris b is young enough to shine brightly from retained internal heat. Its gravitational pull on the star provides a mass estimate that can be compared with cooling models.
Its mass and infrared brightness constrain the heat retained during formation. A compact core and heavy material spread through the interior both remain possible.
Still cooling after formation, this young giant emits infrared light through a molecular atmosphere. That heat keeps it warm far from its star. The spectrum does not supply a visible-colour image or cloud map.
Beta Pictoris has three imaged giant planets. Planet c lies inside b’s orbit. The fainter planet d, announced in 2026, lies farther out and may help shape the inner edge of the surrounding debris disc.
WASP-76 b has iron vapour in its intensely heated atmosphere. As gas moves toward the cooler night side, some of that iron could condense into clouds or droplets. This giant circles its star in less than two days.
Could it rain iron?
Ehrenreich and colleagues measured an uneven iron-vapour signal as this planet crossed its star. One explanation is that iron carried from the hot day side condenses as it reaches cooler regions, possibly making iron rain. The detected signal comes from iron vapour. Droplets have not been observed. Temperature differences, winds and clouds also affect which parts of the atmosphere contribute to the signal.
The measured mass and large radius imply an inflated envelope. Its scorching outer gas sits above much denser material compressed by the weight of the planet.
More iron vapour was detected on one side of the atmosphere than the other. Condensation in cooler regions could explain the difference and might produce iron rain, although the signal comes from vapour and no droplets have been seen.
WASP-103 b orbits so close to its star that tides stretch its shape. Its silhouette is better described as a slightly elongated globe than a perfect sphere. A year passes in less than a day.
Tidal deformation
The star pulls more strongly on the planet’s near side than its far side. On this very close orbit, the difference can distort a fluid globe. Barros and colleagues tested that idea against the shape of CHEOPS transit light curves. Their data favour tidal deformation at about three standard deviations. Measuring the distortion more precisely could help constrain how material is distributed inside.
The star’s gravity stretches this planet. CHEOPS measured a small change in transit shape consistent with a fluid planet deformed by the star’s gravity.
The inferred response to tides is compatible with Jupiter’s, although the uncertainty is large. It constrains how mass is distributed inside the planet without giving a unique core composition.
The star pulls harder on the planet’s near side, stretching it. CHEOPS measured a small distortion during transits. Those measurements constrain its shape without identifying atmospheric bands or colours.
KELT-9 b circles a star far hotter than the Sun. Its atmosphere is hot enough to contain iron and titanium as individual atoms and ions. Many molecules break apart on its blazing day side.
Iron and titanium in the gas
KELT-9 b receives fierce radiation from a star hotter than the Sun. Hoeijmakers and colleagues searched high-resolution transit spectra for the signatures of metals and detected iron and ionised titanium. Such atoms can remain in the gas under these extreme conditions. Many molecules break apart at these temperatures.
KELT-9 b combines several Jupiter masses with an enlarged radius. Its enormous external heat input and deep compression make a simple scaled-up Jupiter a poor description of its thermal structure.
Its extremely hot star heats the atmosphere enough to break many molecules apart. Astronomers have detected iron atoms and ionised titanium in the gas. There is no rocky surface beneath this atmospheric view.
HD 80606 b follows a long, stretched orbit. For most of its year it stays far from the star, then rushes close enough for a sharp burst of heating. Its atmosphere warms and cools as the stellar distance changes.
A sudden blast of heat
HD 80606 b spends most of its long orbit far from its star, then rushes through a very close approach. Laughlin and colleagues followed that passage with Spitzer and measured a rapid rise in infrared brightness. It was a direct way to watch an atmosphere respond to a changing energy supply. The heating repeats on each orbit.
This giant is several times Jupiter’s mass but similar in size, so its interior is strongly compressed. Its brief, intense heating near the star acts first on the outer atmosphere.
Tides can exchange orbital energy with internal heat. Neither the changing atmospheric temperature nor the eccentric orbit alone measures the composition or size of a central core.
The planet follows a highly elongated orbit and heats rapidly as it passes close to its star. Spitzer measured this rise in infrared brightness. The orbit is shown as a diagram, with the star and planet enlarged.
HD 80606 b is the confirmed planet around this star. A distant stellar companion, HD 80607, may have helped reshape its orbit through long-term gravitational perturbations. This is one proposed origin of its extreme eccentricity.
GJ 436 b is a warm Neptune-sized planet losing hydrogen into space. The escaping gas forms an enormous cloud extending far beyond the globe. Its orbit is slightly stretched, bringing it nearer to its star and farther away during each year.
Escaping hydrogen absorbs ultraviolet light
Hubble detected a deep ultraviolet transit lasting far longer than the ordinary transit of GJ 436 b. Escaping hydrogen can explain this enormous absorbing cloud. Its shape has earned the planet comparisons with a comet. The observation follows ultraviolet absorption by gas, so the tail would not look like a bright comet tail in visible light.
Its mass and radius require a light envelope above denser material. A formation-and-structure study found solutions with roughly 10–20% of the mass in hydrogen and helium.
That study favoured more rock than in Neptune, because inward migration lets a growing planet collect rocky material. Water-rich alternatives depend on the formation assumptions.
Hydrogen escaping from the atmosphere forms an enormous cloud around the planet. Hubble detected it absorbing ultraviolet starlight. This gas would not look like a bright blue comet tail to human eyes.
GJ 9827 d is about twice as wide as Earth and has an atmosphere rich in water vapour. This hot planet completes an orbit every six days. Water in the air does not establish an ocean underneath.
Steam around a small planet
Water can be a trace ingredient in a hydrogen atmosphere or make up much of the atmosphere itself. Webb’s NIRISS spectrum, combined with Hubble observations, favours a water-rich atmosphere on GJ 9827 d. The analysis finds a high proportion of water vapour.
Its low density requires lighter material above the rocks. Earlier mass-and-radius models allowed a hydrogen envelope; JWST’s later steam-rich spectrum gives a stronger constraint on that outer layer.
Webb spectra favour water-rich gas around this planet. Water vapour is invisible, so visible clouds would need droplets or particles. The gas detection leaves the possibility of an ocean below unsettled.
Of the three transiting planets, b and c have densities compatible with rocky compositions, while d is larger and less dense. Webb’s steam detection on d adds atmospheric evidence to that mass-and-radius comparison.
LTT 9779 b has an unusually reflective atmosphere despite its fierce heating. Bright clouds return much of the incoming starlight to space. This Neptune-sized planet completes a year in less than twenty hours.
Strong reflected light
Very few Neptune-sized planets are found this close to their stars. When LTT 9779 b disappears behind its star, the visible light drops. That difference shows that the planet reflects a large fraction of the light reaching its day side. CHEOPS observations and atmospheric models connect that brightness to reflective clouds in an atmosphere rich in heavier elements.
Thermal evolution models in the discovery study gave a hydrogen-and-helium envelope of about 9% of the total mass. Its survival so close to the star is unusual.
Most of the mass in that interpretation lies in denser material below the atmosphere. A small mass of hot gas can still contribute greatly to the planet’s radius.
The denser interior could contain rock, water and metal in different proportions. The discovery model estimates the outer gas fraction without fixing a separate iron core.
Reflective mineral clouds are a leading explanation for the unusually bright reflected light measured by CHEOPS. Their exact composition remains uncertain.
TOI-270 d has a hydrogen-rich atmosphere containing methane and carbon dioxide. About twice as wide as Earth, it could have a deep, hot mixture of water and gas below. Whether a separate liquid ocean exists remains unsettled.
Atmospheric gases and deep water
Webb transmission spectra show methane and carbon dioxide in the atmosphere of TOI-270 d. These molecules help constrain the mixture of gases and the conditions that supply them. Deeper models allow either a separate ocean or a hot mixture of gas and water. Atmospheric chemistry narrows the possibilities, but a liquid-water boundary has not been found.
Models matching the atmosphere, mass and radius allow a deep mixture of hydrogen and water. A separate ocean is also possible for some atmospheric conditions.
A gas-dwarf interpretation places roughly 1–5% of the mass in hydrogen-rich gas above a rocky interior. Other solutions replace much of that gas with water.
Several proportions of iron and rock fit the observations. A 2026 study allows composition to change gradually with depth, widening that range further.
The atmosphere contains methane and carbon dioxide. Its visible colour has not been measured. Deeper down, water could mix with hydrogen under pressure or form an ocean under suitable conditions.
TOI-178 g is part of a six-planet system whose outer five orbits follow linked rhythms. The regular timing contrasts with sharply different planetary densities. Planet g is about 2.9 times Earth’s width and takes 20.7 days to circle the star.
Six planets in a resonance chain
CHEOPS, ground-based telescopes and measurements of the star’s motion identified six planets around TOI-178. The outer five follow a chain of linked orbital rhythms. Their densities vary much more than that orderly architecture might suggest. Interior models allow different amounts of gas and dense material, consistent with planets of different compositions.
Its mass and radius were modelled together with those of its five neighbours. Their different densities imply different proportions of rock, water and gas despite their linked orbits.
Six planets circle this star, and the outer five follow linked orbital rhythms. Their densities vary sharply despite this regular pattern. The lineup shows the family, not their actual spacing.
Six planets are known, from b to g. The five outer planets follow a linked chain of orbital periods, but their densities vary sharply. An orderly orbital pattern does not imply identical compositions.
Kepler-10 b is a small rocky planet circling an old, Sun-like star. Its year lasts about twenty hours. Much larger companions occupy the wider orbits in the same system.
Mass from the star’s motion
Kepler detected repeated transits of a small planet with a year shorter than a day. Measurements of the star’s motion supplied a mass, making a rocky composition plausible. Later observations refined the measurements and identified additional outer companions. Its surface receives intense starlight throughout each orbit.
A 2025 analysis gives a mass near 3.2 Earth masses. Combined with its radius, this fits a planet dominated by silicate rock, with a small iron core or none.
Silicate rock is favoured by the measured mass and radius. Intense starlight could melt some of the day side, though the molten fraction has not been measured.
Beyond the small rocky planet b, planet c transits every 45 days and may contain substantial water. Planet d was identified through radial velocities and the shifting transit times of c, with a year of about 151 days.
Kepler-36 b is a dense rocky planet with a much larger neighbour on the next orbit out. The planets pass close to one another despite their different densities. Planet b packs its material into a much smaller globe.
Masses measured through transit timing
Kepler-36 b is dense and rocky, while c is much larger and carries a substantial gas envelope. Their nearby orbits cause gravitational nudges that shift the times of their transits. Carter and colleagues used those timing changes to measure their masses. The density measurements imply very different structures despite similar exposure to starlight.
Its high density favours exposed rock over a massive gas envelope. The observations constrain its bulk composition much better than they constrain the minerals at its surface.
This dense inner planet is consistent with a rock-and-iron body. Atmospheric-loss models can explain it as the remnant of a planet that once carried a light gas envelope.
In atmospheric-loss models, b’s dense core is less massive than c’s, making gas easier to remove. Its iron fraction depends on the assumed composition of the rock.
This dense rocky planet has a much larger, less dense neighbour on the next orbit out. Their orbital distances differ by only about ten percent. The nearby world is planet c, not a moon.
Two planets have neighbouring orbits but very different densities. The inner b is rocky, while the larger outer c retains a light envelope. Their mutual gravitational pulls shift their transit times, allowing both masses to be measured.
Kepler-36 c carries a substantial envelope of gas around a denser interior. Its neighbour b is much smaller and rockier, despite travelling on an orbit just inside its own.
Retaining a gas envelope
Kepler-36 c is much less dense than b, even though their orbital distances differ by only about ten percent. Their mutual gravitational pull changes when each planet crosses the star. Modelling those timing shifts established the density contrast. Models suggest that c retained a gas envelope while b lost most of its gas.
A light envelope explains why planet c is much larger and less dense than nearby planet b. Evolution models track how much gas it could retain while exposed to the star.
The models constrain the total dense core more readily than a separate iron centre. Iron and silicate rock can be divided in different proportions within that core.
A substantial gas envelope makes planet c larger and less dense than its rocky neighbour b. Models suggest that its more massive dense interior helped it retain gas that b lost.
Two planets have neighbouring orbits but very different densities. The inner b is rocky, while the larger outer c retains a light envelope. Their mutual gravitational pulls shift their transit times, allowing both masses to be measured.
Kepler-51 d is almost Jupiter-sized, at about 0.87 times its width, yet contains very little mass. It and the system’s other two transiting planets have thick envelopes and densities far below those of our Solar System’s giants. Its orbit lasts 130.2 days.
A transit that arrived early
Kepler-51’s transiting planets have unusually large radii for their small masses. When Webb observed d, its transit arrived far earlier than expected from a three-planet model. Combining that event with other transit times led to evidence for a fourth planet. The same gravitational timing method helps weigh the visible planets and establish how little material fills their enormous globes.
A very extended atmosphere accounts for much of this planet’s enormous size. Its low density requires a large volume of gas around a much smaller concentration of heavier material.
The extreme combination of large radius and small mass challenges standard cooling models. A 2026 study finds that extra internal heat could keep its envelope expanded.
A dense core may sit under the very extended atmosphere. Its inferred mass depends strongly on whether the planet has extra heat or whether high-altitude material enlarges its apparent radius.
Iron could form a core within the denser region, but its share of that region is uncertain. Explaining the enormously extended outer envelope is the main difficulty for interior models.
Transit measurements show a huge planet, while gravitational timing measurements give a very small mass. An extended gas envelope accounts for its low density. High-altitude particles could also increase its apparent size.
Planets b, c and d are unusually low-density transiting worlds. An unexpected shift in d’s transit time showed evidence for a fourth planet, e. Its properties are inferred from gravitational effects. Its radius has not been measured.
Kepler-138 d could contain a deep layer of water above its rocky interior. Its small size overlaps with rocky planets, but its density suggests a lighter composition. High pressure would make its deep water very different from an Earth ocean.
Water above rock and metal
Piaulet and colleagues combined Hubble and Spitzer transits, stellar velocities and changes in transit timing to constrain Kepler-138 d. Its mass and radius fit a planet with a substantial layer of water above rock and metal. In their model, water occupies about half the volume. Its deep water would experience pressures far beyond an Earth ocean, and an exposed liquid surface is not established.
Its mass and radius fit a thick layer of water above rock and metal. The upper water could form an ocean beneath an atmosphere. A hot, dense fluid without an Earth-like sea surface is also possible.
The small inner planet b accompanies c and d, two worlds with evidence for water-rich compositions. An additional planet, e, was inferred from the system’s gravitational timing signals. It is not known to transit, so its radius remains unmeasured.
At just 0.31 times Earth’s width, Kepler-37 b is smaller than Mercury. Rock and metal could make up its interior. It circles its star every 13.4 days.
Measuring a very small planet
A very small transit dip led to the discovery of Kepler-37 b. To turn that dip into a planet size, Barclay and colleagues used oscillations in the star to measure the star itself. The resulting world is smaller than Mercury. Its mass is much less securely known than its radius, so its internal proportions remain open.
Between the Moon and Mercury in size, this planet could have a similarly rocky, cratered surface. The analogy leaves room for different minerals and geology.
Three planets transit: tiny b, the larger c and outer d. Radial velocities constrain d’s mass, while b remains much harder to weigh. A fourth object, e, has been reported from transit timing, but later work disputes that interpretation.
K2-141 b races around its star in just under seven hours. Its day side is hot enough for molten rock. A thin atmosphere made from vaporised rock could form above the hottest regions.
A possible cycle of vaporised rock
K2-141 b’s short orbit subjects its day side to intense heating. Optical and infrared phase observations can be compared with models of bare rock and a thin atmosphere supplied by a magma ocean. Rock vapour could move away from the hottest region and condense elsewhere. The lava flows and atmospheric circulation have not been mapped.
Mantle simulations allow a dayside magma ocean a few hundred kilometres deep. They also predict slowly moving rock beneath the solid nightside and possible volcanic activity there.
The high density allows an iron-rich centre. Simulations find that cooling of this core can affect mantle motion, though its temperature has not been measured.
The day side is hot enough for a magma ocean. Vaporised rock could form a thin atmosphere, move toward cooler regions and condense there. The extent and depth of the molten rock depend on the model.
Six planets circle HD 110067 in a linked sequence: roughly three inner orbits for every two on the next orbit, changing to four-to-three ratios farther out. Planet b, about 2.2 times Earth’s width, completes the innermost orbit in 9.1 days.
Predicting the missing transits
TESS and CHEOPS observed a family of transiting planets around HD 110067. The linked periods of the inner planets helped researchers predict where to look for the remaining transits. Six sub-Neptunes emerged in a resonant chain. Strong gravitational scattering could disrupt the chain, so its survival suggests a relatively undisturbed history.
Its low density requires light material above the denser interior. The discovery measurements allow a gas envelope but do not identify its composition.
Its mass and radius indicate a low-density outer envelope. The six planets’ regular orbital pattern suggests a relatively undisturbed history in which such gas could survive.
A dense rocky region can lie beneath the gas. The neighbouring planets provide useful comparisons because they formed around the same star, but their interiors need not be identical.
Six sub-Neptunes, b through g, follow a chain of linked orbital periods. The rhythm helped predict missing transits before they were observed. Several measured densities suggest light atmospheres. A distant binary-star companion also belongs to the wider stellar system.
AU Mic b orbits a young red dwarf that frequently flares. Bursts of radiation reach the planet’s atmosphere every time the star erupts. The planet is about 4.8 times Earth’s width, with an 8.5-day orbit.
A young planet around a flaring star
AU Mic b crosses a young red-dwarf star surrounded by a debris disc. Plavchan and colleagues used TESS transits and follow-up observations to establish the planet. Stellar spots and flares complicate measurements while also shaping the radiation reaching the atmosphere.
A 2024 mass-and-radius study finds an inflated young planet that could retain more than 5% of its mass in hydrogen. Strong stellar radiation may remove some of this gas over time.
Rock and water can share the denser material underneath. The planet is still young, so its retained heat affects how much gas is needed to explain its size.
Stellar activity makes the mass difficult to measure, which limits estimates of the dense interior. A separate iron core is possible within those uncertainties.
Powerful flares from the young star expose the extended atmosphere to bursts of energetic radiation. They can alter its chemistry and drive gas loss. The star and planet are drawn closer together than their true separation.
The young planets b and c transit an active red dwarf. Additional companions d and e have been proposed from timing and radial-velocity signals. Their status has been debated because stellar activity complicates the measurements.
PDS 70 c is a young giant still gathering material inside a broad gap in its star’s disc. A smaller disc of dust surrounds the planet itself, offering a possible place for moons to form.
A disc around a growing planet
ALMA detected compact dust emission around PDS 70 c, separate from the much larger disc surrounding its star. Benisty and colleagues interpreted it as a circumplanetary disc, the kind of environment that could supply material for moons. Infrared observations follow the young planet and its orbit.
Infrared spectra favour a dusty atmosphere around this young, luminous planet. The dust composition is uncertain. The atmosphere releases heat as the planet grows and cools.
This planet is still growing inside a young disc. Infrared radiation carries heat from the planet, while material falling toward it can supply additional energy.
Its luminosity depends on its mass, age and continuing accretion. Those uncertainties prevent a firm estimate of how much heavy material has collected at the centre.
ALMA resolved a compact dust disc around this growing planet, though not its fine ring structure. The material could form moons. None have been confirmed.
Two growing giant planets, b and c, occupy a gap in the surrounding disc. Both are still associated with accreting material, and c has a compact dust disc of its own. Additional compact sources have been proposed, but are not treated here as established planets.
51 Peg b is a hot giant with water vapour in its atmosphere. It circles a Sun-like star in just over four days. Because it does not transit the star, its physical size is harder to measure.
The 1995 discovery
Mayor and Queloz found a repeating wobble in a Sun-like star in 1995. The four-day rhythm was caused by a giant planet much closer to its star than the giants of our Solar System. Since 51 Peg b does not transit, its radius is harder to establish. Modern high-resolution spectra can separate the moving planet’s atmospheric signature from the star’s light.
Spectra have detected water vapour in the hot atmosphere. Its molecular absorption can be separated from the star’s light as the planet moves along its orbit.
The planet circles a Sun-like star in just over four days. It does not transit, leaving its radius uncertain. Spectra identify atmospheric gases, while cloud cover and visible colour remain open.
Proxima Centauri b circles the nearest star to the Sun in about eleven days. Its minimum mass is close to Earth’s. We still do not know whether this small world has an atmosphere or a liquid-water surface.
An eleven-day signal
A repeating change in Proxima Centauri’s velocity led to the detection of a planet on an eleven-day orbit. That wobble gives a minimum mass because the orbit’s tilt is uncertain. The planet receives a level of starlight compatible with liquid water under suitable atmospheric conditions, but its atmosphere and radius have not been measured.
Published rocky-planet models range from iron-rich worlds to bodies with substantial water. The radial-velocity discovery gives a minimum mass, but no measured radius to distinguish them.
An iron-rich centre is one allowed possibility. In the 2016 study, a Mercury-like interior and a water-rich planet could both fit the available mass constraint.
A low-mass planet circles the nearest star to the Sun. A dry rocky surface is possible, but without measurements of its radius, atmosphere or water content, other conditions remain possible.
The smaller inner planet d has a roughly five-day orbit, while b circles every eleven days. A more distant planet c has been proposed but remains disputed. Proxima itself belongs to the wider Alpha Centauri stellar system.