OrbitalHub

Astronomers working with the James Webb Space Telescope have produced the first three-dimensional map of an exoplanet’s atmosphere, and the planet they picked turned out to be a great choice. WASP-17b, a hot Jupiter about 1,300 light-years from Earth, has an atmosphere that is roughly circular but stretched out on one side, like an egg pulled gently at the pole. The team, led by Lili Alderson of Cornell, reports the result in Nature Astronomy on 14 October 2026 (Alderson et al., Nature Astronomy, 14 October 2026). The paper is more than a portrait of a single weird world. It demonstrates a new technique, called transmission strings, that turns Webb’s eclipse observations into a latitude-by-latitude picture of a planet’s limb. With that picture in hand, the team can finally check, observation by observation, what hot Jupiter circulation models have been predicting for two decades.

Hot Jupiters are easy to find and brutally hard to characterize. They are gas giants the mass of Jupiter or a few times over, parked in orbits of just a few days around their host stars, with equilibrium temperatures above 1,500 K. Because they transit, you can split their light into a spectrum and read the chemistry of the atmosphere from the depths of absorption lines. That part is mature. What has been hard is geometry: a transmission spectrum averages over the entire thin ring of atmosphere that the planet shows us at transit, and the average smears together the cool morning limb and the hot evening limb, the equator and the poles, and any place on the planet where the chemistry is different from the rest. For most hot Jupiters, you get one spectrum and the knowledge that it hides a lot of weather underneath. Alderson and her colleagues, working with Bristol astrophysicists Hannah Wakeford and David Grant, who developed the technique, figured out how to extract more.

The trick is in how Webb watches a planet disappear behind its star. As the planet slides into eclipse, different parts of its atmosphere are hidden at different times, and the light curve that Webb records during that slide is sensitive to where on the limb the atmosphere is hot, cold, cloudy, or clear. Alderson’s team had previously used that approach to measure global temperature maps, but the transmission-strings method goes further: instead of treating the limb as a single ring, it breaks the ring into angular slices, runs the eclipse light curve through a model that knows how a transit signal changes as the star progressively covers different pieces of the planet, and asks which slice is bright in which wavelength. The output is a map of the limb in temperature and chemistry, around the full circumference of the planet, with separate measurements for the equator, the poles, and everything in between. The phrase “transmission strings” comes from the way those angular slices line up along the limb, like beads on a string.

WASP-17b is a long way from a typical hot Jupiter target. Discovered in 2009, it was the first exoplanet found with a clearly retrograde orbit, meaning it swings around its star in the opposite direction from the star’s spin. It is also one of the puffiest planets on the books, with roughly half Jupiter’s mass stuffed into nearly twice Jupiter’s radius, putting it among the most extreme low-density worlds confirmed to date. That puffiness makes its atmosphere large and easy to probe, and its F-type host star gives Webb a bright, clean signal. It has also been a workhorse for Webb’s exoplanet team: in March 2023, the telescope watched WASP-17b for several eclipses using three instruments at once, NIRISS, NIRSpec, and MIRI, covering wavelengths from 0.6 to 12 microns, under the DREAMS survey run by Webb’s Telescope Scientist Team. That same dataset, used by Grant and colleagues to find the first evidence of quartz clouds in an exoplanet atmosphere in 2023 (ESA/Webb, WASP-17 b transmission spectrum, October 2023) and by Gressier and colleagues to measure a supersolar metallicity in the dayside in 2024, became the input for the new limb map.

The result, when the team unfolded the data, was the egg shape that the press releases kept reaching for. The atmosphere is circular at the equator, where a banded jet stream carries material from the planet’s blistering dayside around to its permanently frozen nightside. As the limb is traced up toward the poles, the atmosphere gets lopsided. One side, the evening side, where gas has just finished its rapid eastward journey from the dayside, is hotter; the other side, the morning side, where gas has crossed the long dark face of the planet and cooled, is colder. The team puts the temperature difference at 650 K between morning and evening limb at the equator, consistent with the predictions of circulation models (Alderson et al., 2026, Nature Astronomy). That confirmation matters because, until now, hot Jupiter circulation models were a chain of inferences: people used eclipse and phase-curve data to infer the dayside temperature, used that to constrain the nightside, used the difference to infer the wind speed. Transmission strings lets you measure the morning-to-evening contrast directly, which is the comparison the models need to be tested against.

Alderson was careful to note that the measurement is exactly what theory predicted. The morning-evening contrast, she told reporters, lands almost exactly where hot Jupiter circulation models say it should, bang-on the trend (Cornell Chronicle, October 2026). That kind of confirmation is rare and welcome in a field where surprises are more common than agreements. The paper also reports detections of water and carbon dioxide in the limb spectrum at high signal-to-noise, and a transition in atmospheric structure between equator and pole, which suggests that the equatorial jet does not just transport heat but also reshapes where the gas rises and falls. As Alderson put it in the same press cycle, the work demonstrates that planets are three-dimensional objects whose atmospheric behavior, as they orbit their stars, matches what general circulation models predict (Cornell Chronicle, October 2026).

WASP-17b has, in some sense, been preparing for this paper for years. The 2023 quartz-cloud finding showed the planet’s atmosphere was chemically interesting, not just hot. The 2024 supersolar-metallicity work pinned down the bulk composition. The 2026 paper, by Alderson and her colleagues, is the moment where the meteorology of WASP-17b becomes something you can put on a wall. Among the more than 6,000 confirmed exoplanets, WASP-17b is now the benchmark for three-dimensional atmospheric characterization, the place future transmission-strings results will be compared against. The technique does not require any new hardware: any transiting exoplanet with a good Webb eclipse light curve is a candidate, and there are dozens already in the archive.

There are also genuine limits to read into the result. The map is one snapshot of one planet at one phase of its orbit, taken in March 2023, and 650 K is a large number but it is also a number that depends on which latitudes the model chose to slice at. The team is already applying the method to other hot Jupiters in the DREAMS survey and beyond, and the next round of papers will start to reveal whether the morning-evening contrast varies from planet to planet in the way general circulation models expect, or whether some worlds break the pattern in instructive ways. What the Webb team has, for the first time, is a tool that lets them ask those questions in three dimensions rather than averaging them away.

Image: Planck map of cosmic microwave background temperature fluctuations across the full sky. Credit: ESA and the Planck Collaboration, CC BY 4.0, via Wikimedia Commons.

The universe is full of missing mass and nobody has the faintest idea what it is made of. Roughly five-sixths of all matter is dark, and the only handle cosmologists have on it is its gravity, which is an awkward position to be in when one candidate answer is something as ordinary as a black hole. It is a stranger position still when the candidate in question is not ordinary at all. A scattering of black holes with masses in the trillions of Suns would be invisible to any telescope and would sit far outside any galaxy, and until recently nobody had found a way to rule them out. A paper accepted by Physical Review Letters finds one, and it does the job using the oldest light in the universe as the instrument.

These hypothetical objects have a name: stupendously large black holes, or SLABs, for a category of primordial black holes heavier than roughly 100 billion Suns. The name was coined by Bernard Carr, Florian Kuhnel and Luca Visinelli, whose 2021 review of the observational constraints on such objects noted how thin the coverage was. Between 10^12 and 10^18 Suns there were, they wrote, no published constraints on a SLAB population at all. That gap mattered, because the range from about 10^14 to 10^19 Suns sits in an awkward sweet spot. Such an object is too heavy to sit comfortably inside a galaxy, which rules out nothing, while being light enough that simple arguments about disrupting dwarf galaxies or heating surrounding gas do not settle the question. Carr and colleagues judged that a population of primordial black holes in that band could still hold a meaningful fraction of the dark matter while remaining hidden. No such object has ever been found. According to Space.com, the heaviest black hole anyone has confirmed so far may weigh only about 36 billion Suns, which leaves a gap of several orders of magnitude between what has been seen and what the arguments had to cover.

The new result comes from Christopher Gerlach, Yann Gouttenoire, Antonio Iovino and Nicholas Leister, working at the Mainz Institute for Theoretical Physics and at New York University Abu Dhabi. Their paper, Closing the Mass Window for Stupendously Large Black Holes, was accepted by the journal on 9 September 2026 and is scheduled to appear in print on 26 October. The argument is indirect and rather elegant. Rather than trying to find the black holes, the team asks what the universe would have looked like if they had been there.

The trick is that a population of discrete objects does not behave like a smooth fluid of particles. If dark matter consisted of particles spread evenly at the moment the universe became transparent, the primordial density field would be smooth everywhere, with ripples that grew in a well-behaved way as the universe expanded. A population of individual black holes is lumpy by construction. The gaps between them depend on where they happen to fall, and that statistical unevenness writes itself into the earliest perturbations in the universe. Cosmologists call this contribution isocurvature, and the microwave background constrains it hard. Gerlach put the intuition plainly when he spoke to Space.com about the result. If dark matter is packed into a handful of gigantic objects, “the universe becomes grainy on very large scales,” and the grain is something we can look for.

Formally, what the team computed is the isocurvature power spectrum that a monochromatic, non-spinning population of primordial black holes would have produced, and then how far that spectrum overshoots the tightest available bounds. The bounds came from a general analysis of isocurvature modes in the CMB and the baryon acoustic oscillation signal, computed with the CLASS and MontePython codes by Buckley, Fernandez and their collaborators. Planck’s own three-point limits on how much of the primordial fluctuation can be isocurvature rather than the usual adiabatic kind were used as an independent check, and they reach the same conclusion. At the largest of the three wavelengths probed, the isocurvature fraction of the primordial fluctuation is capped at about 2.5 percent. A population of black holes each weighing 100 billion Suns or more overshoots that cap across the entire band of interest, and it does so under the more conservative of the two truncation assumptions the authors tested, so the result does not hinge on a technicality.

One boundary deserves flagging. The exclusion runs up to about 10^19 Suns and then stops abruptly, and the reason is not that the constraint weakens but that the objects stop being possible. A primordial black hole that heavy would have had to form before recombination, and anything heavier could not have been made in the early universe at all. The paper also notes that the same range of masses was closed independently and shortly before it was submitted, by an analysis of Lyman-alpha absorption data. A different line of work, published in the Journal of Cosmology and Astroparticle Physics this year, attacks astrophysical SLABs from an unrelated direction: their shadows would be wide enough to appear as dark patches against the microwave background itself, which rules out anything above 10^17 Suns within the last scattering surface and caps the dark matter fraction they could supply at roughly one part in a hundred thousand. Two separate techniques, the same conclusion.

Read carefully, the result is a statement about proportion rather than existence. Gerlach’s group is not claiming that stupendously large black holes cannot exist, only that they cannot make up a significant share of the dark matter. That is still a meaningful narrowing, because the class was a live candidate, and it arrives with consequences in two other directions. The same calculation strengthens the case that the nanohertz gravitational wave background reported by pulsar timing arrays cannot be sourced by primordial black hole mergers. It also challenges scenarios that attribute the surprisingly massive early galaxies seen by the James Webb Space Telescope to primordial black holes seeding density fluctuations.

The forward-looking part is more interesting than the demolition. The paper points at next-generation microwave background instruments, LiteBIRD and the Simons Observatory, as the experiments that will tighten these bounds further, and it flags two ideas for the heaviest masses where the microwave background runs out of leverage: ringdown gravitational wave signals from the formation of very massive primordial black holes, and strong gravitational lensing of the microwave background itself. The team is also turning its attention to the opposite end of the mass range, asking whether the tiny asteroid-mass black holes that remain an allowed candidate can be shut out as a major constituent of dark matter. For now one candidate is gone. The list of things that could be filling the universe’s missing mass is a little shorter, and the sky’s oldest light did the removing.

Figure adapted from Ondratschek et al. 2026 (arXiv 2609.20933), reused under CC BY-SA 4.0.

Most of the Sun is uneventful, and that is exactly the problem. Sunspots get the headlines, flares get the coverage, and the rest of the surface, the part astronomers call the quiet Sun, gets treated as background scenery. Yet a magnetic field still threads every one of those quiet patches, and it has to carry energy upward into a corona that reaches temperatures above a million kelvin. A balloon-borne telescope that spent six and a half days above the Arctic Circle in July 2024 has now delivered the sharpest look yet at what that field is doing in the places nobody was looking. Three papers published this summer describe a magnetic canopy over a quiet region that is nothing like the simple expanding arch that textbook models have assumed for decades. The canopy turns out to be full of narrow threads, some of them running against the grain of the surrounding field, and a companion simulation says those threads are twisted ropes of magnetic flux, the same kind of structure that powers flares.

Why the quiet Sun earns this much attention comes down to area and plumbing. Quiet-Sun regions cover most of the solar disk, so whatever they do, they do almost everywhere. Their magnetic fields are weak, an order of magnitude below the values in active regions, which is why ground-based telescopes struggle to measure them: atmospheric turbulence blurs exactly the fine structure that matters. The chromosphere, the thin atmospheric layer sitting on the visible surface, is also where the Sun’s energy problem lives. Something has to move energy from a surface too cool to power the corona into a corona with temperatures above a million kelvin, and the candidate answers all involve magnetic fields that are far more tangled than a smooth model would suggest. Spicules, the narrow jets that shoot up from the surface, have been promoted in recent years as the main carriers of mass and energy into the upper atmosphere, yet their magnetic field strength has never been pinned down with confidence. As Universe Today notes in its coverage of the papers, the true strength inside those jets had resisted measurement. If the quiet Sun turns out to be full of twisted flux ropes, then the same structures that set off flares in active regions are threaded through places the Sun barely seems to be doing anything at all.

The instrument that cracked this open was built by people who wanted a telescope with a mirror no space agency was likely to fund. Sunrise III is a one-metre solar telescope hung under a helium balloon that floated at roughly 35 kilometres, above nearly all the air that distorts solar observations from the ground. It lifted off from the Esrange Space Center near Kiruna in Sweden on 10 July 2024 at 6:24 in the morning CEST, after several weather delays, and over the next six and a half days it tracked westward along the Arctic Circle and across the Atlantic while the Sun never set (Max Planck Society launch release). The observatory weighed about three and a half tonnes, the balloon expanded to around 130 metres across, and the instruments aboard returned roughly 200 terabytes of data, enough to resolve structures as small as 50 kilometres on the solar surface (NAOJ on the first images). Four space probes and ten ground telescopes watched the Sun at the same time as part of a coordinated campaign. The payload was an international suite, including SCIP, the Sunrise Chromospheric Infrared spectroPolarimeter, a near-infrared instrument built and led by Japan’s National Astronomical Observatory of Japan, which measures the full polarization state of several spectral lines at once and can follow the magnetic field continuously from the photosphere up through the chromosphere (NAOJ release on the results).

Three papers from that flight appeared in the Astrophysical Journal Letters over the summer, released together by NAOJ and the National Institute for Natural Sciences, with Masahito Kubo of the NAOJ SOLAR-C project, Yoshihiro Naito of the Graduate University for Advanced Studies, and Yukio Katsukawa of the NAOJ solar science observatory named as the contact authors. Kubo and colleagues report on a quiet region near disk centre, mapped in a single uninterrupted scan of 107 minutes with a field of view 58 arcseconds on a side and a polarimetric precision of three to four hundredths of a percent of the continuum (Kubo et al. 2026). Their maps show elongated thread-like magnetic structures above the internetwork, the darker lanes between the bright network lanes, and no obvious counterpart directly below them in the photosphere. The threads are narrower than an arcsecond, which on the Sun is a distance of roughly 700 kilometres, and their line-of-sight field strength under the weak-field approximation comes out 10 to 20 gauss weaker than the canopy around them. In the clearest cases the polarity of a thread is the opposite of the canopy it sits inside.

The other two papers deal with the rest of the puzzle. Ondratschek and colleagues, working with a chromospheric extension of the MURaM magnetohydrodynamic model built at the Max Planck Institute for Solar System Research, fed the observed magnetic map into a three-dimensional radiative simulation and reproduced the fine structure in the synthetic polarization signal (Ondratschek et al. 2026). In the model, the opposite-polarity features appear about a megametre above the surface, running for two to seven megametres with widths near one megametre, and they belong to twisted flux ropes that persist through most of a 21-minute sequence. Naito and colleagues looked instead at the solar limb, where off-limb spicules can be seen against the sky, and measured their fields directly (Naito et al. 2026). Their line-of-sight values reach 20 gauss and stay roughly constant over the first three arcseconds above the limb, then spread out to about 40 gauss between three and six arcseconds, and the way the line profiles change shape with height says the wider number is really a statement about how deeply the line-forming layer can see into a forest of overlapping spicules rather than a measure of any one jet.

The measurement behind all of this is old physics applied with new precision. A magnetic field splits the wavelengths of a spectral line into components, and the polarization of those components, described by the four Stokes parameters, encodes the field’s strength and direction. Because Ca II 854.2 nanometre light forms in the chromosphere and Fe I 846.8 nanometre light forms in the photosphere, observing at both wavelengths shows how the field changes with height. What a balloon buys is the freedom to do that without the atmosphere smearing the finest structures, and the numbers reflect it: a ten-second dwell at each slit position and a polarimetric noise floor of a few hundredths of a percent is enough stability to invert a single line for a magnetic field and trust the result. The MURaM-ChE model then handles interpretation, taking a three-dimensional field configuration, computing what an observer would actually measure, and matching that synthetic signal against the real one.

What the three papers together say is that the Sun’s least active regions are not simple. The magnetic field above them is organized into tight, twisted substructures that can carry the opposite polarity to their surroundings, embedded in the apparently ordered canopy that springs from the network. That has a direct consequence for solar physics beyond the quiet Sun, because twisted flux ropes are the structures that reconnect and release energy, and they are usually associated with flares in active regions. Finding them in the quiet Sun suggests the inventory of twisted field on the disk is much larger than anyone assumed, and that small-scale reconnection may be an ordinary part of the quiet atmosphere rather than an exceptional event. The instruments also have limits worth naming: six and a half days is a short window, the balloon could observe only a narrow band of wavelengths compared with a space telescope, and one field of view cannot map the whole Sun. Still, the data volume will keep feeding the community for years, and the next step is a dedicated space observatory built for the same job, SOLAR-C, which NAOJ already runs as a project and which can watch the same tangled field for months instead of days.

A white dwarf does not need to do anything dramatic to set the record books of stellar astrophysics. It only needs to wait. For V445 Puppis, the wait ended this week when a team led by John Mills at the University of Warwick finally traced the system back to its ingredients: a compact stellar corpse pulling material from a rare stripped helium star. The RAS press release for the Royal Astronomical Society’s National Astronomy Meeting in Birmingham described the result, and it pinned down what astronomers have wondered about since November 2000, when a point of light in the constellation Puppis brightened and then vanished behind its own debris.

The eruption produced a bipolar outflow stretching more than 1.5 trillion kilometers, with clumps of gas clocked at roughly 8,900 kilometers per second, about 3% of the speed of light, racing through the ejecta. The press release calls them “bullets” and notes that nothing of their kind has ever been seen in any other nova.

Almost every nova we know about is hydrogen-rich. A white dwarf siphons gas from a Sun-like companion, the stolen layer gets compressed and hot, and a thermonuclear runaway on the white dwarf’s surface flashes the system into view. The whole cycle takes a few weeks to a few years, then the dwarf goes back to feeding. Helium novae are different in a way that sounds like physics trivia but is genuinely awkward for the textbooks: the accreting layer contains almost no hydrogen at all. The fuel has been swapped out, and only a handful of candidates in the Milky Way match that description. V445 Puppis, the only one currently confirmed, is essentially the only laboratory on hand for studying what such an explosion actually looks like before, during, and after.

That matters because helium novae are one of the proposed pathways to Type Ia supernovae. White dwarfs that keep accreting material approach the Chandrasekhar mass limit, where the star can no longer support itself against electron degeneracy, and the result is one of the most important kinds of explosion in observational cosmology. Type Ia supernovae are the standard candles behind the 1998 discovery of the accelerating expansion of the universe, the work honored by the 2011 Nobel Prize in Physics. Understanding their progenitors is one of the standing problems in stellar astrophysics, and helium novae sit on one of the more plausible candidate routes.

The same logic explains why a 25-year wait is not the embarrassment it sounds like. The eruption’s own dust disk made any direct view of the central binary impossible through optical and ultraviolet light, and the only way to see through was to wait for the obscuring veil to thin enough for a new generation of instruments. That window has now opened.

V445 Puppis erupted in late November 2000 and was classified almost immediately as a nova, although the spectrum looked strange from the start. There was helium, carbon, and a long list of heavier elements, but essentially no hydrogen, the element that defines nearly every other nova on record. For the next 25 years the system sat inside a thick dust torus, bright in the infrared from the re-radiated light of the eruption but stubbornly opaque at the wavelengths where stellar spectra carry information about the two stars at its center.

The new result came from a deliberate effort to combine four observatories. Mills pulled infrared images from the European Southern Observatory’s Very Large Telescope, optical images from the Hubble Space Telescope’s Wide Field Camera 3, long-term spectroscopy from the Southern African Large Telescope, and photometric monitoring from NASA’s TESS satellite. The four cover the infrared through visible bands, plus time-resolved brightness measurements that make it possible to extract the orbital period. John Mills’ research page at the University of Warwick describes the same approach in his group’s broader work on hot and cool binaries.

With the dust clear enough, the spectrum showed strong ionized helium lines consistent with material actively falling back onto the white dwarf. The brightness curve gave an orbital period of 3.7 days, about twice the value the most widely cited prior estimate had produced. A 2023 paper in The Astrophysical Journal Letters, “V445 Puppis: Dustier than a Thousand Novae,” had estimated the eruption threw off between 0.01 and 0.1 solar masses of gas, which would actually mean the white dwarf is losing mass over repeated eruptions rather than approaching a Type Ia explosion. The 3.7-day orbit and the active accretion add two more data points to that open debate.

Embedded inside the bipolar outflow, the Hubble imaging also resolved distinct compact knots of gas, possibly oxygen-rich, moving at speeds comparable to the fastest ejecta ever measured in a classical nova. The RAS announcement explicitly notes that no analogous structures have been observed in any other nova, which is why Mills flagged the origin of the bullets as the unresolved problem in the talk.

The work was reported at NAM 2026 in Birmingham on Wednesday 22 July 2026, in a talk titled “Long-term evolution of the helium nova V445 Puppis and the emergence of the underlying binary.” As of this writing the results have not yet gone through formal peer review, a point Universe Magazine’s coverage makes explicit, but the data sources themselves are public and the orbital measurement should be straightforward to confirm.

A classical nova does not destroy the white dwarf; the explosion only blows off the freshly accreted surface layer. The energy budget is well understood: hydrogen-rich material piles up under degenerate conditions, the base reaches roughly 10 million K, and a CNO-cycle hot layer drives the runaway. The peak luminosity is comparable to a small star, which is why distant novae can briefly outshine the combined light of their host galaxies.

Helium novae follow the same broad pattern with hydrogen-poor fuel, and the mechanics are fuzzier because so few examples exist. The white dwarf in V445 Puppis cannot be accreting hydrogen-rich gas, which is what motivates the “stripped helium star” companion model, a star that has already lost its hydrogen envelope through prior interaction. The models get the orbital period, the spectral lines, and the mass loss to line up at least approximately, although the geometry of how the white dwarf pulls material off its donor’s outer layers is still being worked out.

The bullet problem is the genuinely new piece. A 2009 infrared study had identified clumps moving at 6,700 km/s in a narrow hourglass-shaped outflow, with knots at 8,450 km/s. Mills’ combined analysis finds similar speeds and identifies discrete oxygen-rich clumps at roughly 8,900 km/s, fast enough to cross the Earth-Moon distance in under five minutes. The knots are not fast enough to be a separate explosion, but they are too fast to be passive material carried by a slow wind, which is why the team assumes some kind of post-eruption engine. So far none of the obvious candidates, including a re-ignition on the white dwarf surface and gravitational rebound, cleanly fits both the timing and the composition.

The imaging pipeline is the more straightforward part. Hubble’s WFC3 narrow-band filters F502N, F680N, and FQ727N isolate the emission from ionized oxygen, nitrogen, and sulfur separately, which is what made the oxygen identification possible. Combining frames from two different epochs (2013 and 2015) let the team measure proper motion of each knot directly.

After more than two decades behind a self-built dust wall, V445 Puppis is now visible as what it has been all along: a white dwarf feeding helium-rich gas from a rare stripped companion. The 3.7-day orbital period replaces the older 1.87-day estimate, the dust mass loss points toward a white dwarf that is being drained rather than grown by each eruption, and a catalog of fast oxygen-rich knots has no clean explanation yet.

The next obvious step is a longer TESS baseline plus higher-cadence UV spectroscopy with HST. If the binary truly is no longer accreting efficiently, the 2023 dust-mass argument gains weight and the path to a Type Ia supernova through this channel looks harder than it did five years ago. If the bullets can be linked to a specific instability, the same mechanism probably applies to a few helium-rich extragalactic novae that have otherwise looked like clean hydrogen-poor events. Either way, V445 Puppis has stopped being a curiosity and started being a test.

Image credit: NASA.

The James Webb Space Telescope, the NASA/ESA/CSA observatory at the Sun-Earth L2 point, has caught an exoplanet doing something a planet that close to its star really should not be able to do. HD 3167 b is a rocky world roughly one and a half times the size of Earth, and it zips around a K-type star in the constellation Pisces once every 0.96 Earth days. At that distance, the dayside should be glowing red-hot bare rock, the kind of surface where silicates are molten and any gas should have been stripped away by stellar wind and high-energy photons long ago. Instead, the dayside is noticeably cooler than the theoretical maximum, and a new analysis of a JWST secondary eclipse says the most likely reason is an atmosphere. It is the coldest lava world with evidence of an atmosphere on record, and the first result from a survey of ten such worlds led by astronomer Megan Weiner Mansfield. The team’s writeup, as covered by Space.com, lays out what makes the detection hard to explain any other way.

That word cool does not mean comfortable. The planet’s equilibrium temperature sits at about 1786 K, somewhere between a slow oven and molten glass. Lava worlds previously caught with atmospheres have all been hotter than roughly 2200 K. HD 3167 b is hot enough that the rock on its star-facing side is almost certainly molten, and cool enough that something other than bare rock is controlling its dayside temperature.

The result, published in The Astrophysical Journal Letters, comes from a single JWST observation using the Mid-Infrared Instrument’s Low Resolution Spectrometer, the team that has been doing most of the heavy lifting on small rocky exoplanets for the last two years. The team watched HD 3167 b slip behind its host star, a maneuver called a secondary eclipse, and measured how much mid-infrared light disappeared. They measured the white-light eclipse depth at 38 plus or minus 11 parts per million. That number is more than five standard deviations shallower than a dark bare rock with no heat redistribution would produce at this distance from the star. The only good explanations are an atmosphere that reflects some incoming starlight back into space, an atmosphere that moves heat from the day side to the night side, or some combination of the two.

The lead author is Brandon Park Coy, a graduate student at the University of Chicago working in planetary scientist Edwin Kite’s group. Kite is the second author and a co-author on a long line of papers trying to figure out which rocky exoplanets hold onto their air. The HD 3167 b paper, described in a University of Chicago news release, is the first published result of a survey that Megan Weiner Mansfield, who did her PhD at Chicago and is now at the University of Maryland, is running. The program is looking at ten lava worlds with JWST and asking a single question. Below what temperature do rocky planets start losing their atmospheres, and above what temperature do they keep them? The published sample so far suggested a gap. Ultra-hot lava worlds such as 55 Cancri e kept their air. Cooler rocky worlds, the ones too cold to be lava worlds at all, often did not. The boundary lived somewhere in the middle, and nobody had measured it.

HD 3167 b sits right at that boundary. Its dayside is too cool for a bare rock, so it has an atmosphere. Its equilibrium temperature is far below the ultra-hot threshold, so it is the first cool end of the population in question. The boundary should sharpen from a guess into a measured curve as the program keeps adding worlds to the sample.

The detection technique itself is worth slowing down on. A secondary eclipse is the small dip in light that happens when a planet passes behind its star as seen from Earth. The dip is small because the planet is much cooler than the star and emits much less light at the wavelengths being observed, but it is measurable, and the depth of the dip tells you how much of the observed light is the planet’s own thermal emission. Doing this in mid-infrared light, around 15 micrometers for this observation, is what lets JWST see the heat of a planet whose dayside is below 2000 K. The same measurement in visible light would be much less informative, because at those wavelengths the planet contributes essentially nothing.

Why does the dayside being cooler than the maximum matter? On a bare rock world with no atmosphere, the side facing the star absorbs starlight, heats up, and re-radiates that heat as infrared light. The temperature that produces is set by the rock’s reflectivity and the distance to the star, and it is the highest possible dayside temperature for that orbit. Add an atmosphere, and two things change. First, the atmosphere can carry heat from the day side to the night side, evening out the temperature. Second, clouds or hazes can reflect some starlight back into space before it ever reaches the surface, lowering the dayside temperature below the bare-rock maximum. Venus is the textbook example of the first effect, with almost no temperature difference between its equator and poles or between its day and night sides. Earth is the textbook example of the second effect. HD 3167 b is showing one or both on a world whose dayside is liquid rock.

What the atmosphere is actually made of, the team is careful to say, is not known. The single JWST spectrum is not precise enough to pick out individual molecules. The paper notes that the dayside emission spectrum is compatible with a silicate-rich composition consistent with vaporized rock, but also with heavier gases like carbon dioxide, carbon monoxide, or water. Picking among those will take follow-up spectroscopy, ideally with JWST’s Near-Infrared Spectrograph, which covers a different wavelength range. The team is also hoping that the rest of the ten-planet program will start to fill in a statistical picture, where several planets measured the same way start to show consistent or inconsistent composition patterns.

The planet itself is also interesting in detail. Its host star, HD 3167, is a K-type main-sequence star slightly smaller and cooler than the Sun, and the planet was first spotted in 2016 in data from the Kepler space telescope’s extended K2 mission, then confirmed by Vanderburg and collaborators in a dedicated follow-up. Like most ultra-short-period rocky worlds, it is tidally locked, with one side permanently facing the star and the other side in permanent night. The new paper refines the planet’s radius to roughly 1.6 Earth radii and confirms the bulk density is slightly lower than a pure Earth-composition planet would suggest, consistent with the planet having a substantial gas layer on top of the rock.

For the broader exoplanet field, the result is a small but real data point in a much larger argument. Most rocky exoplanets found so far, the ones that orbit small cool stars and sit at Earth-like distances, do not have atmospheres. JWST has now measured enough of them to know that, and that has driven a lot of the conversation about whether Earth-like planets are rare or common. Lava worlds are the opposite case. They are easy to observe because they are hot, and the evidence so far is that most of them have something on top of the rock. HD 3167 b extends that trend. It does not change the picture for habitable-zone rocky worlds, but it does sharpen the boundary case where atmospheres start to fail.

There is one quote in the paper and its accompanying press release that captures why a team would spend JWST time on a planet that cannot possibly host life. Lava worlds are useful because they let you watch the processes that matter for all rocky worlds at the extreme end of their temperature range. A planet that is too hot for life can still tell you something about how its rock interacts with its air, how heat moves around, and how clouds form. None of that is unique to lava worlds. All of it shows up in the physics of habitable worlds too. The difference is that on a lava world you can measure it directly, and on a habitable world you mostly have to guess.

A planet heavier than Neptune but the size of Earth should not, by the rules of how planets form, be a rock. Anything above roughly ten Earth masses is supposed to grab gas from the protoplanetary disk and become a sub-Neptune or larger. Yet a small catalog of worlds has long carried the label “Mega-Earth,” rocky objects sitting above that ten-Earth-mass line, and that label itself colors what people assume about planet formation at the upper end of rocky worlds. A paper published today by Casey Brinkman and Nicolas Cowan at McGill’s Trottier Space Institute argues that the entire category may be a measurement artifact (Brinkman & Cowan 2026, arXiv:2610.12600, reported on by Astrobiology.com the same day). Every candidate Mega-Earth in the literature, the authors show, orbits a magnetically active star. Active starspots and plages mimic and inflate the radial-velocity signal of an orbiting planet, and the chance that all of these outliers would land on active hosts by coincidence is below one percent. The implication is uncomfortable but specific. The Mega-Earths were never there. The detection technique was just seeing starlight flicker, and the label stuck because the underlying objects were just hard enough to disprove.

The category was supposed to mark the upper edge of rocky planet formation. When Xavier Dumusque and colleagues at the Harvard-Smithsonian Center for Astrophysics used the HARPS-N spectrograph on La Palma in 2014 to weigh Kepler-10c, the result landed at seventeen Earth masses packed into a radius only 2.35 times Earth’s (Harvard Gazette, June 2014). That density, on the order of seven grams per cubic centimeter, was consistent with a compressed rocky interior under extreme pressure. The team coined the term “Mega-Earth” (Wikipedia: Mega-Earth) on stage at the Boston meeting of the American Astronomical Society, and Dimitar Sasselov called Kepler-10c the “Godzilla of Earths.” The discovery paper was published in The Astrophysical Journal in June 2014. Subsequent surveys added more candidates with apparent masses between ten and forty Earth masses and radii sitting inside the rocky planet regime, and at various points the literature has listed six to eight such worlds, depending on which mass-radius cuts are applied. Each one pressed against a theoretical limit set by collisional mantle stripping and post-formation erosion that no primordial rocky planet should be able to cross.

What jumps out of the resulting scatter plot is that every white-circle Mega-Earth candidate lies on the active side, while the gray-circle population of ordinary rocky and sub-Neptune-mass planets spans inactive through quiet. The four Mega-Earths the authors identify, WASP-84 c, TOI-1347 b, TOI-2322 c, and HD 63433 b, all orbit stars with log R prime HK above negative four point seven five. Three additional planets sit close to the Mega-Earth density regime: GJ 523 b, TOI-815 c, and HD 18599 b. All seven orbit active stars. The probability of that alignment by chance, given how many inactive and quiet-host rocky systems exist in the same sample, comes out below one percent. The authors also note that a previously Mega-Earth-classified world, HD 119130 b, was already re-measured by another group and found to have an inflated mass caused by constructive interference between the planet’s orbital signal and the star’s rotational modulation.

The mechanism behind the contamination is well known but tricky to remove. Starspots and the brighter magnetic regions called plages move across the disk of a rotating star and produce Doppler shifts in the integrated starlight that are indistinguishable, in many pipelines, from the reflex motion of a planet. Starspots suppress local convection and produce false radial-velocity signals at the star’s rotation period. Plages do the same at somewhat different amplitudes. When a planet’s orbital period happens to land near an integer or half-integer multiple of the star’s rotation period, the two signals can constructively interfere and inflate the inferred planet mass by a factor of two or more. That is the situation for several of the candidate Mega-Earth hosts, whose TESS light curves show clean rotational modulation at periods from roughly four to fourteen days. The full radial-velocity datasets that yielded the published masses came from public archives including the NASA Exoplanet Archive, which compiled discovery spectra from HARPS, PFS, and TESS follow-up campaigns. Gaussian-process regression, the tool most often used to handle stellar activity in radial-velocity fits, is only as good as its kernel choices and its priors. When the kernel is too smooth, the process absorbs genuine planetary motion as activity noise. When it is too wiggly, activity noise leaks into the planetary model and inflates the mass.

What the paper does not do is rule out rocky planets above ten Earth masses on theoretical grounds. The runaway gas accretion argument sets ten Earth masses as the threshold above which a core cannot avoid grabbing gas during formation, but it is a threshold, not a wall. Collisional mantle stripping after disk dispersal or photoevaporation of a once-larger planet’s envelope can, in principle, leave a bare rocky remnant. The authors are explicit that those pathways remain possible, just unproven. The Mega-Earth label, in their framing, has been applied to planets whose mass measurements cannot yet distinguish between a real ultra-dense rocky world and a noisy radial-velocity fit around an active star. Re-observing these systems with ESPRESSO on the Very Large Telescope, with HARPS on the TESS-Keck Survey infrastructure, or with NIRPS at La Silla, all of which have the spectral resolution to see activity-induced line-profile variations directly, would settle the question. Several of the candidate Mega-Earths have already been on those queues. The paper, which appeared on the arXiv preprint server on 8 October 2026 and is now in the resubmission cycle for The Astrophysical Journal Letters, lands in a catalog of corrections that have repeatedly eaten specific claimed populations of extreme exoplanets. Stellar contamination has eaten mini-Neptunes hiding inside hot-Jupiter mass measurements, cold Jupiters past ice lines that were artifacts of magnetic cycles, and small-radius planets that turned out to be diluted binaries. Each time, the underlying detection method is sound for inactive host stars and quietly wrong for active ones. Adding Mega-Earths to that list is, in a way, an expected move. The catalog of rocky worlds stays grounded. The Ultra-Massive Rocky Planet threshold, if it exists at all, moves somewhere more difficult to measure from the ground, perhaps requiring JWST transmission spectroscopy of these same systems to constrain atmospheric mass, or years of radial-velocity monitoring as activity cycles rotate through minimum phases. For now the simple chart is enough: a horizontal line at ten Earth masses, with seven suspected outliers sitting on the active side and zero on the quiet side. Pattern recognition cut the question in half.