A solar telescope in New Mexico is being torn down, and the reason is a pool of liquid mercury hidden inside it. The mercury was the clever trick that made the telescope work. It is now the thing that dooms it.
The telescope’s optical system weighs 250 tons, and it had to turn smoothly all day. You can’t spin something that heavy on ordinary wheels. So instead the whole thing floated on a sealed pool of liquid mercury. Mercury is dense and flows easily, so the load rode on the fluid and turned with almost no friction. That is how the 250-ton optical system of the Richard B. Dunn Solar Telescope was designed to move.
However, on January 5, crews found some of that mercury where it wasn’t supposed to be.
We are not engineers, chemists, or occupational-safety specialists, and nothing here is guidance for handling mercury or any other hazardous material. This is our reading of a public engineering story and the agency documents around it, not advice.
Why a telescope floated on liquid metal
The Dunn sits at about 9,200 feet on Sacramento Peak, inside the Lincoln National Forest. The Air Force built it, and it opened as the Vacuum Tower Telescope on October 15, 1969. From the start it was one of the best high-resolution instruments in the world for studying the Sun.
As sunlight comes through the telescope, the Sun’s image slowly rotates over the course of a day. To keep that image lined up with the instruments studying it, you have to turn the whole optical assembly to match. Not just the tube, but the observing room, the lab, and the floor people stand on. That is where the 250 tons comes from. The mercury float let crews rotate the image of the Sun by turning the entire instrument.
The amount of mercury involved surprises people. The system floated on roughly 120 gallons of the liquid metal, sealed in the bearing. Mercury is so dense that this much weighs around 18,000 pounds on its own. And the bearing carrying the full load hung, according to the observatory, on just three bolts, each three inches across.
A puddle of mercury is a serious problem because of what the metal does. It is a neurotoxin, it gives off vapor at room temperature, and once it escapes a sealed system it doesn’t stay put. The real question became whether the bearing could be trusted to hold 18,000 pounds of a poison for years to come. The National Science Foundation brought in the engineering firm Thornton Tomasetti to assess that, and the answer shaped everything that followed.
Why removing the mercury ends the telescope
The mercury isn’t a part you can swap out. It is the thing the entire optical system stands on. Take it away and the telescope can’t rotate, which means it can’t do the one job it was built for. Removing the hazard and keeping a working telescope turned out to be the same decision pointing in opposite directions.
On February 24, 2026, the NSF announced it would remove the mercury, leaving the telescope permanently unusable, then demolish it. In its own words, “After assessing the engineering challenges, anticipated repair costs and safety and environmental risks, NSF determined that removing and safely disposing of the telescope’s mercury is the most prudent course of action.” That is the agency defending its own call, and worth reading as such. Preservation advocates in New Mexico dispute that demolition was the only option, and residents and lawmakers have pushed to save the wider observatory campus.
The Dunn was already near the end of its working life. The Air Force transferred it to the NSF in 1976, and in its final years it ran only in a limited mode since 2019. Much of what it once did best has moved to the NSF’s newer, far larger Inouye Solar Telescope in Hawaii.
What comes down with it
Demolition is scheduled to begin as early as September 2026, and the NSF has said it will contribute up to $100,000 toward an exhibit about the observatory. What ends is more than a building: more than half a century of solar science from a machine that, for a long stretch, was the best in the world at what it did.
On the morning of 2 September 1859, telegraph operators along the Boston-to-Portland line unplugged their batteries, watched the acid-slick terminals go dark, and kept sending messages anyway. The wires were carrying a current the sky was pushing through them. For roughly two hours, the storm itself ran the telegraph.
The event that made this possible had begun the previous day, when English amateur astronomer Richard Carrington, sketching sunspots through a filtered telescope at his private observatory in Redhill, saw two beads of intense white light flare above a complex sunspot group and fade within five minutes. Roughly seventeen and a half hours later, a coronal mass ejection slammed into Earth’s magnetic field. According to the National Environmental Satellite, Data, and Information Service, that night skies from the Rocky Mountains to Panama glowed brightly enough for people to read newsprint by aurora alone.
Auroras were reported over Cuba, over Jamaica, over Hawaii, over Colombia. Gold miners in the Rockies got up and cooked breakfast, thinking dawn had come early.
What the operators actually did
The telegraph network of 1859 was the most sophisticated communications system humans had ever built — tens of thousands of kilometres of copper strung on wooden poles across North America and Europe, powered by wet-cell batteries at every station. It was also, unintentionally, the world’s first planetary-scale antenna for geomagnetic current.
When the coronal mass ejection compressed Earth’s magnetic field, the rapid change induced voltages along any long conductor lying roughly east-west. The longer the wire, the bigger the induced current. Telegraph lines were exactly that: long, conductive, and grounded at both ends.
Contemporary accounts of the storm record what followed. Machines sparked. Paper caught fire. Some keys clacked out garbled text on their own. On the Boston-to-Portland line, the two operators on duty — never named in the surviving transcript, which was printed in the Boston Evening Traveller — realised their batteries were fighting the induced current rather than adding to it. They disconnected the batteries. The line kept working.
Space Daily has previously covered the Boston-Portland exchange in detail, including the transcript-style back-and-forth in which the two men agreed the auroral current was steadier than their own.
More useful is the transit time. Most coronal mass ejections take three to four days to cross the 150 million kilometres between the Sun and Earth. The Carrington ejecta did it in roughly 17.6 hours. That implies a bulk speed above 2,000 kilometres per second, meaning the plasma cloud arrived while its magnetic field was still coherent and aggressive.
A 2024 paleomagnetic reanalysis suggests the storm was even more powerful than the nineteenth-century instruments could capture. IFLScience’s summary of that reanalysis notes that some of the magnetograph traces from 1859 were pinned against their stops — the needles had physically reached the end of their travel and could measure no further.
Why auroras reached Cuba and Hawaii
Auroras form when charged particles funnelled along Earth’s magnetic field lines slam into the upper atmosphere at roughly 100 kilometres altitude, exciting nitrogen and oxygen atoms. The colours — green from oxygen at lower altitudes, red from oxygen higher up, blue and purple from nitrogen — depend on which atoms are being hit and how hard.
Normally the field lines that carry these particles anchor near the magnetic poles, which is why the aurora borealis is a high-latitude phenomenon. During a severe geomagnetic storm, the entire auroral oval expands equatorward. The stronger the storm, the further south (or north, in the southern hemisphere) the glow reaches.
For the Carrington Event, the oval expanded so far that observers in Havana logged red auroras overhead. In Honolulu, the sky reportedly went blood-red. Ship captains in the Caribbean noted the phenomenon in their logs. This is the signature of an extreme storm: not just bright aurora at the usual latitudes, but aurora somewhere it has never been seen before.
The physics behind the wires is the same whether the conductor is a telegraph line in 1859 or a high-voltage transmission cable in 2026. A changing magnetic field induces a voltage in any conductor sitting inside it. Long, grounded conductors accumulate the biggest voltages because the induced electric field integrates over the length of the wire.
Solar storms, as Space Daily has explored elsewhere, do not attack people directly. They attack long metal. In 1859, the only long metal humans had strung across continents was telegraph wire, and there were only a few tens of thousands of kilometres of it. Today’s grid, pipelines, rail signalling systems and undersea cables constitute a target surface millions of times larger.
That is why the same physical event, replayed on modern infrastructure, would look nothing like sparking telegraph paper. It would look like transformers cooking themselves from the inside as DC-like induced currents saturate their iron cores.
What the same storm would do now
The most cited worst-case assessment comes from the US National Research Council’s 2008 report on severe space weather events. That report estimated recovery from a modern Carrington-class event could take four to ten years and cost trillions of dollars, with the damage concentrated in extra-high-voltage transformers that are custom-built, expensive, and made by only a handful of manufacturers worldwide.
Not everyone agrees the picture is that bleak. Grid operators in the United States have been under FERC rules since 2014 requiring geomagnetic disturbance planning. Some transformer designs are more tolerant than others. And the geometry of the storm — which hemisphere gets hit at what local time — matters enormously. A hit at the wrong hour could be catastrophic; the same magnetic pulse a few hours earlier or later, geometrically misaligned with the major grids, would be far less destructive.
The best modern measurement of a Carrington-class event comes not from Earth but from a spacecraft that happened to be in the wrong place at the right time. In July 2012, a coronal mass ejection of comparable magnitude to the 1859 storm erupted from a region of the Sun that had rotated just past the Earth-facing line. The ejecta missed Earth. It hit NASA’s STEREO-A spacecraft squarely.
STEREO-A’s instruments were built to measure exactly this sort of plasma cloud. They survived the encounter and returned detailed data on the shock front, the magnetic field orientation, and the particle energies. Reconstructions suggest that had the eruption occurred a week earlier, with Earth in the firing line, the induced ground currents would have exceeded the 1859 event.
The May 2024 G5 storm — the strongest to hit Earth in more than two decades — was, on Ledvina’s assessment, roughly an order of magnitude weaker than Carrington. Auroras reached Florida and southern Europe. Some GPS-guided tractors in the American Midwest were knocked off-line during planting season. The grid held.
Proposed defences, from prosaic to exotic
Most of the practical mitigation work is unglamorous: adding neutral-current-blocking capacitors to transformer neutrals, hardening spare-transformer stockpiles, refining space-weather forecasts so operators can shed load or reconfigure grids before a storm arrives. The Space Weather Prediction Center at NOAA runs continuous monitoring and provides the alerts that grid operators act on.
Then there are the more ambitious proposals. A 2026 paper in the journal Space Weather, summarised on Hackaday’s coverage of the concept, proposes what its authors call StormWall: releasing a cloud of neutral alkali-metal gas at geosynchronous orbit ahead of an incoming coronal mass ejection. The gas gets ionised by the storm plasma, and in the process absorbs — the paper’s simulations suggest — up to fifty per cent of the storm’s coupling energy before it reaches the magnetosphere.
The mass required is not small. The simulation called for 384,048 kilograms of gas, which the authors estimate as roughly six Starship launches to GEO. Whether that is feasible depends on assumptions about Starship’s eventual payload capacity that remain, in 2026, unproven at scale.
What the 1859 record does and does not tell us
The Carrington Event is the calibration point for every modern space-weather risk model. It is also a single data point. Discover Magazine’s overview notes the estimated recurrence interval is roughly once every 500 years, but that figure comes from ice-core nitrate spikes and tree-ring carbon-14 anomalies with wide uncertainty bars. Some paleomagnetic studies suggest even more powerful storms may have occurred in 774 CE and 993 CE, both detected as radiocarbon spikes rather than direct observations.
What the 1859 record does not tell us is what happens when a storm of that magnitude hits a civilisation that runs on transformers, semiconductors and just-in-time logistics. The Boston-to-Portland operators had the option of unplugging their batteries and carrying on. A modern grid operator watching geomagnetically induced currents climb through a substation’s neutral does not have that option. The transformer either holds or it does not, and if it does not, the replacement is on a barge from South Korea.
The Sun is currently near the peak of that cycle. Activity will decline over the next few years toward solar minimum around 2030, then climb again. A Carrington-class event is not tied to the peak — the 1859 storm actually occurred as Cycle 10 was descending — so the elevated risk window is not neatly bounded.
Somewhere on the visible disc of the Sun tonight, a sunspot group is being tracked by half a dozen space-weather forecasting centres. Most such groups produce nothing more than routine flares. Occasionally one produces a coronal mass ejection that misses Earth by a few degrees of solar longitude. The 1859 event happened because one group, at the right moment, pointed itself squarely at the third planet and let go. The wires were waiting.
On 18 August 1868, a French astronomer named Pierre Janssen stood in Guntur, on the eastern coast of India, and watched the Moon slide across the Sun. Totality ran for roughly six minutes, an exceptionally long window by the standards of eclipse work. He pointed a spectroscope at a solar prominence — one of those enormous loops of glowing gas that arc off the Sun’s edge — and split its light into a rainbow of thin coloured lines. Among them was a bright yellow line at 587.5 nanometres that matched nothing in any laboratory on Earth.
The gas that produced that line would not be isolated on this planet for another 27 years.
When it finally was — coaxed from a lump of uranium ore by a Scottish chemist named William Ramsay in 1895 — it was named for the place where it had first been seen. Helios. The Sun. Helium is the only element on the periodic table first discovered somewhere other than Earth.
What Janssen actually saw
The instrument Janssen carried to Guntur was a spectroscope — a prism, essentially, mounted so that sunlight passing through it fanned out into a spectrum. Nine years earlier, the German physicist Gustav Kirchhoff, working with the chemist Robert Bunsen, had shown that each element, when heated, emits light at its own signature set of wavelengths. Sodium glows a particular yellow. Hydrogen glows a particular red. If you could split the light of a distant object and read the pattern, you could read its chemistry.
Astronomers had been waiting for a total eclipse to try this on the Sun. The corona and the pink prominences at the Sun’s edge are normally drowned out by the glare of the disk. Only when the Moon covers that disk do they become visible, and on most eclipse days that window is two or three minutes at best.
Janssen got six. In the spectrum of a solar prominence he saw a bright yellow line sitting near, but not quite on, the familiar sodium doublet. According to the account preserved by Science Friday, the line was so bright that Janssen suspected it might be observable in ordinary daylight, without waiting for another eclipse.
He was right, and he did not wait long to prove it. By ten o’clock the next morning, 19 August 1868, he was observing prominences against the undarkened Sun, using the spectroscope alone to isolate their light from the glare — a result documented in Françoise Launay’s study of the Guntoor observations in the Journal of Astronomical History and Heritage.
The Englishman who arrived at the same line
Five thousand miles away in England, Joseph Norman Lockyer had been working on the same problem — how to see prominences without an eclipse — and had arrived at the same technique independently. On 20 October 1868, two months after Janssen’s observations in India, Lockyer saw the same yellow line in ordinary daylight.
Both men’s reports reached the French Academy of Sciences on the same day. The Academy struck a commemorative medal carrying both their profiles, with Apollo’s chariot crossing the face of the Sun.
Lockyer went further than Janssen in one respect. After failing, with the chemist Edward Frankland, to reproduce the line in the laboratory, he proposed that it came from an element that did not exist — or had not yet been found — on Earth. He gave it a name: helium, from helios, the Greek word for the Sun.
Naming an element on the basis of a single unexplained line was audacious, and Lockyer knew it. In print he mostly called the line D3, after its position beside the sodium D1 and D2 lines. For the next quarter century, plenty of chemists refused to accept there was an element behind it at all.
Why chemists would not accept it
The problem was not that Lockyer’s spectroscopy was wrong. The problem was that no one could put the gas on a bench.
Nineteenth-century chemistry was built on samples you could weigh, dissolve, heat, and combine. An element you could only see in the light of a body 150 million kilometres away was, to many working chemists, not really an element at all. It was a line. The line might belong to some familiar element behaving strangely at high temperature. It might belong to something in the Earth’s atmosphere that the light had passed through on its way to the spectroscope. It might belong to an instrumental artefact.
The 27-year figure also flattens a middle chapter. As Space Daily has noted in its account of the gap between helium’s detection and its terrestrial isolation, the Italian physicist Luigi Palmieri recorded the same 587.5-nanometre line in Vesuvian lava in 1882. That was the first sighting of helium in earthly material. What nobody had, for another thirteen years, was the gas itself.
Janssen, meanwhile, kept chasing eclipses. In 1870, hoping to observe totality from Algeria while Paris was under siege in the Franco-Prussian War, he escaped the city by balloon. Clouds obscured his view.
The uranium ore that finally settled it
The gas came out of a rock.
In 1895, in London, William Ramsay was hunting for argon in minerals when he turned to cleveite, a variety of uraninite that released an unknown gas when treated with acid. An American chemist, William Francis Hillebrand, had already noticed the gas a few years earlier and taken it for nitrogen. Ramsay ran it through a spectroscope.
There was the yellow line.
Ramsay sent a sample to Lockyer for confirmation. Lockyer wrote of watching the discharge tube light up: “When I received it from him, the glorious yellow effulgence of the capillary, while the current was passing, was a sight to see.” The element named for the Sun had been sitting inside a lump of ore dug out of a Norwegian mine.
What Ramsay was really seeing, though no one at the time understood it in these terms, was radioactive decay. Uranium and thorium atoms in the Earth’s crust break down slowly over hundreds of millions of years, and one product of that decay is an alpha particle — which, once it captures a pair of electrons, is a helium atom. The helium in Ramsay’s flask had been accumulating in the ore since the rock formed.
The sequence that followed rearranged the periodic table, though helium was not the element that cracked it open. Argon came first: Ramsay and Lord Rayleigh announced it in August 1894, and it fit nowhere in Mendeleev’s 1869 chart. Helium, isolated seven months later, told Ramsay that argon was not an anomaly but the second member of a family. Working with Morris Travers, he pulled neon, krypton and xenon out of liquefied air in the summer of 1898, and took the 1904 Nobel Prize in Chemistry for the whole column.
The chemistry of a solitary yellow line in an Indian sky turned out to be the chemistry of an entire family of elements no one had suspected existed.
Why helium was hiding on Earth
Helium is the second most abundant element in the universe — roughly 24% of ordinary matter by mass, behind hydrogen’s 75%. Almost everything else, from the iron in blood to the oxygen in the air, makes up the remaining one to two per cent.
On Earth it is vanishingly rare. Helium makes up about 5 parts per million of the atmosphere, roughly 0.0005%. The reason is gravity. Helium atoms are so light and so fast that in the thin upper atmosphere a steady fraction of them exceed escape velocity and drift off into space. The Earth cannot hold on to them.
What helium exists here is trapped underground, produced by the same radioactive decay that filled Ramsay’s flask and captured in the same geological formations that trap natural gas. Nearly all commercial helium is separated from that gas, a by-product of drilling for methane. The United States, Qatar, Algeria and Russia dominate supply, and the chain is brittle: when Qatar was blockaded in 2017, laboratories and hospitals began running down stocks within weeks.
The strange thing about the element named for the Sun is that most of what modern civilisation uses it for happens near absolute zero. Liquid helium boils at 4.2 kelvin, about −269°C, lower than any other substance. That makes it the only practical coolant for the superconducting magnets inside MRI scanners, NMR spectrometers, and particle accelerators like the Large Hadron Collider. A conventional hospital MRI holds somewhere between 1,500 and 1,700 litres of liquid helium in a cryostat wrapped around its main magnet. Let the magnet warm and it loses superconductivity, and the machine stops working.
Spacecraft use it too. Rocket engines are purged with helium before and after firing to clear residual fuel and oxidiser, and liquid hydrogen tanks are pressurised with it. The Saturn V used helium to pressurise the tanks that carried Apollo to the Moon. SpaceX’s Falcon 9 uses helium for the same job, stored in composite overwrapped pressure vessels; the failure of one such vessel inside the second-stage liquid oxygen tank destroyed a Falcon 9 and the Amos-6 satellite on the pad at Cape Canaveral on 1 September 2016, during propellant loading ahead of a static fire test.
Balloons are the least of it.
The eclipse legacy
Janssen’s 1868 observation began a tradition. Total solar eclipses became laboratories. In 1919, British teams organised by Frank Watson Dyson and Arthur Eddington photographed stars near the eclipsed Sun from two sites, Príncipe off West Africa and Sobral in Brazil, and measured the deflection of their light by the Sun’s gravity. The error bars were wide and the precise confirmation would take decades more, but the result supported Einstein’s general theory of relativity and made him a public figure almost overnight.
The son of Norman Lockyer, W. J. S. Lockyer, was quoted in The New York Times after the 1919 result, dismissing its practical importance: the discoveries, he said, did not personally concern ordinary human beings, and only astronomers were affected. General relativity is now the correction that keeps GPS satellites accurate. Every smartphone navigation fix depends on it.
Eclipses had produced, within a fifty-year span, a new element and a new theory of gravity. And the technique Janssen used — splitting the light of a distant object to read its chemistry — is now the backbone of astronomy. Every exoplanet atmosphere characterised by the James Webb Space Telescope, every distant quasar analysed for its composition, every star sorted by spectral type, is being read by the method he pointed at a solar prominence in Guntur.
The gas he identified is now leaking, atom by atom, out of the Earth’s atmosphere and back into space. Some of it drifted up from balloons at children’s birthdays. Some of it boiled off from MRI machines in hospital basements. Some of it vented from rocket pads at Cape Canaveral. All of it is going the same place — up, and out, and eventually back toward the Sun that gave it its name.
When NASA’s Parker Solar Probe passed through the solar corona during perihelion encounters at the heliospheric current sheet, its instruments recorded energetic protons at energies far above what existing models of particle acceleration at that location could account for.
A team led by Mihir Desai at the Southwest Research Institute, with James Drake, Marc Swisdak, and Zhiyu Yin from the University of Maryland’s Institute for Research in Electronics and Applied Physics has published an analysis of those measurements in The Astrophysical Journal Letters, identifying magnetic reconnection at the heliospheric current sheet as the mechanism responsible. The proton energies detected were, in the study’s framing, approximately a thousand times greater than the available magnetic energy per particle that models of this process had predicted.
The finding matters for two reasons: it identifies a source of energetic particles close to the Sun that was not anticipated, and it raises questions about how far the existing understanding of reconnection-driven particle acceleration extends.
What Parker measured at the current sheet
The heliospheric current sheet is a vast, warped surface in the solar wind where the Sun’s magnetic field reverses polarity. Parker crosses it during perihelion passes, when the spacecraft dips inside the corona. At those distances, the current sheet is narrower and more structured than it appears in the near-Earth solar wind, and the reconnection events occurring there have characteristics that instruments further from the Sun cannot resolve.
Within the exhaust region downstream of a reconnection event, where magnetic field lines have broken and rejoined and the released energy is being distributed into the surrounding plasma, Parker detected trapped protons at energies up to approximately 400 keV. The particles were confined within magnetic islands, structures that form within the reconnection exhaust as reconnected field lines roll up into closed loops. Magnetic island merging, as those loops interact and combine, appears to be the mechanism that accelerated the protons to the observed energies.
The thousand-fold discrepancy between what the models predicted and what Parker measured is the number that requires explanation. Reconnection as a particle accelerator was not a new idea: the process has been theorised to play a role in accelerating charged particles in a range of astrophysical environments. What was not predicted was that the energisation within near-Sun reconnection exhaust could reach this level through the island-merging pathway.
Why this source was not in the models
Current models of solar energetic particles, the high-speed charged particles that propagate through the solar system and can affect spacecraft and astronauts, have generally attributed the most energetic events to large-scale shocks: the driven shock fronts of coronal mass ejections, and the slower-building shocks of co-rotating interaction regions. Reconnection at the current sheet was considered a secondary contributor at best, capable of producing modest particle energies close to the Sun but not a major accelerator of the particles that eventually arrive as energetic storm particle events near Earth.
The Parker measurements suggest that picture needs revision. The magnetic island merging mechanism Desai and colleagues identify can evidently produce proton populations at energies that challenge the boundary between what reconnection does and what shock acceleration does. Whether this source can explain a significant fraction of observed near-Earth energetic particle events, or whether it contributes mainly at energies and scales that remain confined close to the Sun, is a question the paper does not yet resolve.
What this adds to the corona heating problem
A separate thread connects to the longstanding question of why the solar corona is so much hotter than the surface below it. The corona reaches temperatures of several million degrees Celsius while the photosphere sits at around 5,500°C. The energy source sustaining that gradient has not been fully identified. Reconnection, along with wave heating, is one of the leading candidates.
If reconnection at the current sheet is producing particles at the energies Parker measured, it is also depositing energy into the surrounding plasma at rates that may be higher than the models assumed. That does not resolve the corona heating problem, but it shifts the accounting. The energy going into particle acceleration through magnetic island merging is energy that must come from somewhere in the magnetic field, and tracing it may tighten the constraints on how much reconnection contributes to coronal heating overall.
Drake and Swisdak contributed the reconnection theory component of the analysis. The collaboration between the observational data from Parker and the theoretical modelling capacity at the University of Maryland is the structure the paper’s analysis rests on.
What comes next
Parker Solar Probe has now completed more than two dozen perihelion passes, with each close approach yielding data at distances and speeds no prior mission could reach. The heliospheric current sheet crossings that provided the data for this paper are repeatable: Parker will continue to encounter the current sheet during future perihelia, allowing the team to look for the magnetic island signature in additional events and test whether the particle energies observed here hold up as a consistent feature rather than a single anomaly.
Solar Orbiter, the European Space Agency mission operating in coordination with Parker, offers complementary measurements at somewhat greater distances. Comparing what the two spacecraft measure of the same particle populations as they propagate outward from the Sun will help establish which features originate close to the source and which are modified by the intervening solar wind. The energetic particle observations that puzzled researchers in the pre-Parker era now have a candidate mechanism. Whether that mechanism accounts for a large or small fraction of what actually reaches Earth remains the open question.
In 2002 two astronomers at Johns Hopkins University, Karl Glazebrook and Ivan Baldry, averaged the light of more than 200,000 galaxies and first announced that the collective color of the universe was a pale turquoise. Within weeks they took it back. A mistake in their color software had skewed the result, and the corrected answer was a milky off-white that the public soon nicknamed cosmic latte.
The color was never the point of the research. It started as a footnote to a serious survey of the light galaxies give off, and it became, briefly, one of the most talked-about findings in astronomy. Newspapers ran with the idea that the cosmos had a single average shade, and home decorators took an interest in a color drawn from every galaxy at once. The retraction, when it came, was almost as widely reported as the original claim.
How you average the color of the universe
The two were studying the cosmic spectrum, built from the Australian Two-degree Field Galaxy Redshift Survey, which had measured the spectra of more than 200,000 galaxies. By combining many tens of thousands of those spectra, they produced a single volume-averaged, luminosity-weighted result, in effect the average color of light emitted by the nearby universe. Their real aim was to read the history of star formation out of that combined light, since young blue stars and old red stars leave different fingerprints in it.
That blend is why the spectrum landed where it did. It is a mixture of young, hot, blue populations and old, cool, red ones, and the average of the two is neither strongly blue nor strongly red. The paper noted that the resulting color is not the kind a single glowing object would produce, which was the first hint that turning a spectrum into a single perceived shade is less straightforward than it sounds.
On a whim, Glazebrook decided to work out what that spectrum would look like to a human eye, using the standard color-matching functions that describe how people perceive different wavelengths. Integrating the cosmic spectrum through those functions returned a blue-green that the team announced as “a few percent greener than pale turquoise.” Announced at a January 2002 meeting, the idea that the universe had an average color, and that the color was greenish, caught the public imagination at once.
The white-point mistake
Not long after the turquoise claim made news, Mark Fairchild of the Munsell Color Science Laboratory at the Rochester Institute of Technology got in touch. Fairchild had, almost literally, written the book on the subject, the 1998 text “Color Appearance Models.” He and his colleagues found that the free software the astronomers had used had set its “white point” incorrectly. The white point is the light that the eye treats as plain white, and it shifts with surroundings, a little yellow under household bulbs, a little blue on some screens. The wrong setting had tinted the whole result green.
A color scientist at the same laboratory, Francis Imai, put the problem in everyday terms. “Tungsten lights, for example, can make the white point a little bit yellowish,” he explained. “Some monitors have a bluish white point. Your visual system tries to adapt, so that you assume that color is white, but it’s actually yellowish or bluish.” The freeware had effectively chosen the wrong reference for white, and the error pulled the whole result toward green.
When the white point was fixed for someone viewing the light in a dark room, the color came out beige. Adjusted for daylight it turned faintly red, and for indoor lighting it shifted blue. Baldry made the dependence vivid: “if you look at all the light in the universe from a room that has a red neon light, then it may appear turquoise,” he said. “But that’s not a standard perspective.”
Glazebrook did not hide from the mistake. “It’s our fault for not taking the color science seriously enough,” he said. “I’m very embarrassed. I don’t like being wrong, but once I found out I was, I knew I had to get the word out.”
The astronomers posted their spectrum data online and invited anyone to recompute the color. From a flurry of suggested names, cosmic latte stuck, chosen for the milky off-white shade.
What the color actually means
The “color of the universe” is not a property of space the way temperature or density is. It is a perceptual quantity, and the turquoise blunder showed exactly why that matters: the same spectrum reads as beige, red or blue depending only on the lighting you imagine yourself standing in. There is no neutral place to view all the galaxies from, so there is no single true color, only a color for a stated set of assumptions.
It is also an average, luminosity-weighted across one survey’s galaxies at modest distances, not a tally of every photon the cosmos has ever released. The galaxies in the 2dF sample are relatively nearby in cosmic terms, so the figure describes the light of the present-day universe rather than the redder glow of its distant, earlier reaches. And it was, by the authors’ own account, a lighthearted aside to a paper about how fast stars have formed across cosmic time. The beige is real enough, but it describes a calculation and a convention, not a wall you could paint to match the sky.
The naming, too, was informal. Cosmic latte was a popular choice rather than a scientific term, and it appears in galleries of color values more often than in the astronomical literature.
A beige worth keeping
What lingers is less the color than the correction. Two researchers made a vivid, widely repeated claim, learned within weeks that they had got it wrong, and said so in plain language rather than letting the prettier answer stand.
The universe, averaged and viewed in the dark, is roughly the color of milky coffee. The more durable lesson is in how quickly that beige replaced the turquoise once someone checked the color science.
When a very massive star burns through its fuel, the textbook ending is bleak. Nothing is left to push back against gravity, the core caves in on itself, and everything funnels down to a single point of infinite density called a singularity, wrapped in an event horizon that lets nothing escape. A black hole. Two theoretical physicists at Goethe University Frankfurt have now written down a different ending, one in which the same collapse triggers something like a miniature Big Bang inside the star and stops short of a black hole entirely.
In their solution to Einstein’s equations, a tiny region of dark energy switches on at the center of the imploding star and begins to expand, the way our own universe did in its first instant. That expansion pushes outward just as gravity pulls in, and if the two forces settle into balance, the collapse freezes. What remains is a gravastar: an object almost as compact as a black hole, but with no singularity at its heart and no horizon around it.
What a gravastar actually is
The gravastar, short for gravitational vacuum condensate star, is not a new idea. It was first proposed in 2001 by physicists Pawel Mazur and Emil Mottola, who suggested that gravitational collapse might end not in a singularity but in a ball of dark energy held together by a thin shell of ordinary matter.
Dark energy is the same repulsive ingredient that cosmologists invoke to explain why the universe’s expansion is speeding up. Inside a gravastar, it would exert an outward pressure that props up the object’s enormous mass, which would otherwise keep falling inward. The thin shell of matter at the surface, helped by gravity, acts like a belt holding the whole thing in place.
The appeal is that a gravastar sidesteps the two features of black holes that physicists find hardest to accept. There is no singularity, the point where the equations of general relativity stop making sense, and there is no event horizon, the one-way boundary that traps everything behind it. Mazur and Mottola argued their version would also be thermodynamically stable and would dodge the information paradox, the long-running puzzle of what becomes of the information that falls past a horizon. The trade-off is that a gravastar can be made nearly as dense and compact as a black hole, which is also the problem: from the outside, the two would look almost identical.
For about 25 years that idea sat with a hole in it. Researchers could describe what a finished gravastar might look like, and a few had modeled one forming from a collapsing shell of matter, but no one had shown how a solid, collapsing ball of matter would actually turn into one. That is the gap the new work tries to close.
The Big Bang that stops a collapse
The Frankfurt physicists started from one of the oldest models in the field, the Oppenheimer-Snyder description of a uniform sphere of pressureless matter, or dust, collapsing under its own weight. Left alone, that sphere becomes a black hole. The new step was to seed an expanding patch of dark energy, technically a de Sitter region, at the very center and follow what happens as the two meet. They did this within ordinary general relativity, without the exotic modifications to gravity that such black hole alternatives often require, which is part of why the result drew notice.
In the model, the dark energy bubble starts from zero size and grows outward while the dust falls inward. As the bubble approaches the Schwarzschild radius, the size at which an object of that mass would tip over into a black hole, its expansion naturally slows. If it comes to rest just as it meets the infalling matter, the inward and outward motions cancel and a static gravastar is born. The authors compare the seed to a small Big Bang going off inside the dying star.
There is a delicate handoff where the growing bubble meets the falling dust. In the model a surface tension along that boundary keeps the dark energy from simply engulfing the matter; instead the dust is nudged outward as the region expands against it. Because the dust is treated as collisionless, the two do not slam together in a shock, they just pile up slightly at the seam. That is what lets the configuration settle rather than tear itself apart.
Daniel Jampolski worked out the solution in his master’s thesis, supervised by Luciano Rezzolla, a professor of theoretical astrophysics at the university. To find configurations that actually balance, the pair did not run the collapse forward. They started from a finished gravastar and integrated the equations backward in time to learn which starting conditions could have produced it, a standard trick when only a narrow set of inputs yields the configuration they were after.
Where the picture is still idealized
This is a result on paper, not a sighting in the sky, and the authors are careful about what it does and does not establish. The balance that produces a gravastar is one of only three possible outcomes in their model. The other two are a plain black hole or an object that never settles into equilibrium. Landing on the gravastar requires fine-tuning the starting conditions, which means the scenario is possible in the mathematics rather than obviously common in nature.
There are other simplifications baked in. The collapsing matter is treated as idealized dust with no internal pressure, and the collapse is assumed to be perfectly spherical, conditions no real star fully obeys. The dark energy bubble is assumed to begin at exactly zero size, which the authors describe as a mathematical convenience that a fuller quantum treatment would have to replace. And the model finds a ceiling on how compact the original star can be: above an initial compactness of three-eighths, in the units physicists use, the collapse to a black hole becomes unavoidable no matter what.
Rezzolla has been blunt that none of this is a case against black holes. Looking for alternatives, he has said, should not suggest skepticism toward black holes, which remain the most natural and simplest explanation for what gravitational collapse produces. The point is to test whether the exotic alternatives can even form in principle, and now one of them has at least a formation story.
How we would ever tell the difference
If gravastars look so much like black holes from the outside, the obvious question is whether anyone could distinguish them. Using light to tell the two apart is still a matter of open debate among physicists. The more promising route runs through gravitational waves. When two compact objects merge, the new object left behind rings like a struck bell as it settles, and a gravastar with no horizon should ring at different frequencies than a black hole with one.
Those tremors are exactly what detectors have been recording since 2015, when the first gravitational waves were observed. So far the signals fit ordinary black holes well, and the difference a gravastar would make is subtle enough that picking it out would take more sensitive instruments than exist today.
For now the work adds a worked example to a long-running argument about what really sits at the end of a star’s life. The equations permit something other than a singularity, at least in an idealized case, which is not the same as saying nature chooses it. Whether the universe ever takes that option is a question the math cannot answer on its own.
On WASP-121 b, a gas giant more than 1.5 times the width of Jupiter, the line where day turns to night is not the same temperature all the way around. The edge of the planet where evening is setting runs hotter than the edge where morning is breaking, and on the hotter side the air is fierce enough to tear water molecules into their separate atoms.
A team of astronomers reported the difference in Nature Astronomy on June 10, and the way they found it is almost as striking as the finding itself. They did not photograph the planet or fly anything near it. They watched WASP-121 b cross the face of its star with the James Webb Space Telescope and noticed that the planet’s shadow changed shape as it slowly rotated during the few hours of the crossing. That change is the first time anyone has measured a planet’s rotation during a single transit and read its weather from it.
How you watch a planet turn from light-years away
WASP-121 b is what astronomers call an ultrahot Jupiter, a gas giant with an equilibrium temperature above roughly 2,000 kelvin, which is hotter than some stars are cool. It hugs its star so tightly that a full year there lasts about 30 hours. At that range the planet is tidally locked, meaning one face is held toward the star permanently, the way the same side of the Moon always faces Earth. One hemisphere bakes in endless daylight; the other never sees the star at all.
The usual way to study such an atmosphere is transmission spectroscopy. When the planet passes in front of its star, a sliver of starlight filters through the ring of atmosphere around the planet’s edge, and different gases absorb different colors of that light. Read the colors that go missing and you can tell what the air is made of around the planet’s rim.
The new twist is timing. Because WASP-121 b is so close to its star, it turns far enough during a single transit that the slice of atmosphere backlit at the start of the crossing no longer overlaps the slice backlit at the end. So as the planet rotates, fresh longitudes swing into view, and the amount of starlight the planet blocks shifts from one moment to the next. The team, led by Cyril Gapp, built a light-curve model that allowed the planet’s apparent size to change during the transit, and tested it against the older assumption that a planet’s silhouette stays fixed. Across two separate Webb observations, the changing-size model won by a wide statistical margin.
The two stares were years apart and used different instruments: one in October 2022 with Webb’s near-infrared spectrograph, another in October 2023 with a separate slitless imager. That spread mattered, because WASP-121 b is an awkward target. It circles a hot, fast-spinning star on a tilted orbit badly misaligned with the star’s spin, close enough that the star’s gravity is slowly pulling the planet apart, and the star itself is brighter at its poles than at its equator. Each of those quirks can smear a transit on its own, so the team had to model them out before trusting that the leftover asymmetry belonged to the planet’s weather rather than to its star.
Why the evening edge runs hotter than the morning edge
The shape of the change told the team which way the temperatures ran. As WASP-121 b rotated, the side rotating into view absorbed more and more carbon monoxide while its water signal stayed flat or dipped slightly. The carbon monoxide feature grew by about 200 parts per million relative to the rest of the spectrum.
That pattern points back to heat. Water survives in cooler air but breaks apart on a scorching dayside, a process called thermal dissociation, while carbon monoxide holds together even when it is roasting. So a stretch of atmosphere rich in carbon monoxide but poor in water is a stretch that is extremely hot. The team found that signature strengthening across the evening terminator, the dividing line where the dayside rotates around toward night.
The reading lines up with a picture other instruments have been sketching for years: WASP-121 b’s dayside is lopsided, with the eastern half hotter than the western half. On planets like this a deep circulation carries heat from the permanently lit dayside toward the night side, and a strong equatorial jet drags that flow eastward, piling heat onto the evening side. Webb has now caught that lopsidedness not as a static map but in motion, by letting the planet’s own spin carry the hot air across the telescope’s line of sight.
What this does and does not prove
The detection is statistically strong, and the team did the unglamorous work of ruling out the boring explanations. They showed the asymmetry was not an instrument artifact, and not just an illusion cast by the host star’s uneven brightness, which can mimic a lopsided planet on its own.
Even so, the result comes with honest caveats, and the paper does not hide them. One of Webb’s two detectors recorded a symmetric transit while the other recorded the asymmetric one; the signal lives at the longer infrared wavelengths, which is consistent with the physics the team proposes but is a reminder that the effect is subtle. The interpretation also leans on a three-dimensional climate simulation of the planet, and that simulation does not reproduce the observed asymmetry unless its morning side is made colder by hand. In other words, the model and the data agree on the direction of the temperature difference but not yet on its size.
There is also a number worth not over-reading. The simulation puts WASP-121 b’s dayside at around 2,800 kelvin, but that is a modeled value, not a thermometer reading, and the study deliberately stops short of quoting a precise temperature gap between dawn and dusk. The honest claim is about which side is hotter and why, not by exactly how many degrees. A faint hint in the data even suggests the planet may be slightly squashed by its star’s gravity, but the team declined to make that measurement, judging the evidence too shaky to stand on.
A new way to read distant weather
What makes the work matter beyond one planet is the method. Until now, the difference between a planet’s morning and evening had to be teased out either from the static shape of its silhouette or from ground-based instruments sensitive enough to catch the Doppler shift of moving air. Using the planet’s rotation during a single transit gives astronomers a third handle, one that works at the lower resolution that space telescopes deliver.
The team points to other planets where the same trick should work, fast-rotating ultrahot worlds such as WASP-33 b and KELT-9 b, though those orbit rapidly spinning stars whose own distortions will have to be untangled first. For now, WASP-121 b stands as the test case: a world where the sky on the evening side and the sky on the morning side are different climates, close enough in the data that a telescope more than a million kilometres from Earth could tell them apart by watching the planet quietly turn.
For five years, the leading story about Kamoʻoalewa, a small asteroid that travels alongside Earth, was that it might be a stray piece of the Moon. A new peer-reviewed study argues it is probably something far more ordinary: a fragment of a distant asteroid family, weathered until its surface only looks lunar.
The timing is sharp. On or around June 7, after a 13-month cruise, China’s Tianwen-2 spacecraft slipped into orbit around Kamoʻoalewa, a maneuver tracked by independent radio astronomers even before China’s space agency formally confirmed it. Sample collection is expected within weeks. The asteroid that may decide a long-running debate about its own origin is now being circled by the machine built to bring a piece of it home.
Why the Moon theory took hold
Kamoʻoalewa was found in 2016 by the Pan-STARRS survey in Hawaiʻi, and it is one of a small set of asteroids called quasi-satellites: rocks that orbit the Sun but stay close enough to Earth to seem like loose companions. This one is the most stable of the bunch, expected to linger near us for a few hundred years. It is small, somewhere between roughly 40 and 100 metres across, and it spins fast, turning about once every 28 minutes.
Quasi-satellites are faint and awkward to observe, which is part of why this one drew such outsized attention once a clue to its makeup appeared. It sits close, it stays put for centuries, and it is bright enough to study with large telescopes during its annual passes. For an object that might preserve a record of the early Solar System, that combination is rare.
In 2021, astronomers reported that its reflected light was a closer match to lunar-like silicates than to any common asteroid type, with an unusual extra redness. That resemblance was the spark for the lunar-fragment idea. A 2024 study went further, proposing a specific birthplace: the young Giordano Bruno crater on the Moon’s far side, from which an impact could have flung debris into an Earth-like orbit.
It was a tidy and romantic picture. A souvenir of the Moon, looping around its parent planet, waiting to be collected.
What the new study actually did
Writing in Nature Communications, a large international team led from the Chinese Academy of Sciences reopened the case from three directions. First, they reanalyzed the asteroid’s spectrum and pinned down a key absorption feature at a wavelength of 1.001 micrometres, a value they report as consistent not with lunar material but with LL chondrites, a common type of stony meteorite.
Then they tested whether ordinary meteorite could be made to look like Kamoʻoalewa. In the lab, the team blasted LL chondrite samples with laser pulses to mimic space weathering, the slow darkening and reddening that solar wind and tiny impacts inflict on any airless surface over millions of years. Heavily weathered chondrite powder, they found, reproduced the asteroid’s spectrum. Solid chips of the same meteorite did not.
Finally, they traced where such an object could have come from. Their modeling points to the ν6 secular resonance, a gravitational escape route in the inner asteroid belt, and more specifically to the Flora family, a cluster of asteroids born from an ancient collision. Kamoʻoalewa’s makeup, they write, resembles the well-studied asteroid Itokawa and seven Flora-family members, only more weathered.
The resonance matters because it offers a plausible delivery route. Rocks that drift into it can have their orbits stretched until they cross Earth’s path, which is one way an inner-belt fragment could end up as a quasi-satellite rather than a lunar one. The argument, in short, is that you do not need the Moon to explain Kamoʻoalewa. An ordinary asteroid nudged out of the Flora family and aged in sunlight can account for the same colour, the same orbit, and the same odd companionship with Earth.
Why a powder, not a slab, is the key detail
That powder-versus-chip result is the quiet heart of the paper. A fresh asteroid surface and a deeply weathered one can be the same rock and still reflect light differently, and grain size matters as much as chemistry. The team’s reading is that Kamoʻoalewa is coated in fine, long-exposed dust, which is exactly what would redden an ordinary asteroid until its colour drifts toward the lunar range.
If that holds, the redness that first suggested the Moon is not a fingerprint of lunar rock at all. It is what time does to a small, slow-tumbling asteroid that has been left out in the solar wind for ages. The comparison with Itokawa carries weight here, because Japan’s Hayabusa mission already returned grains from Itokawa and confirmed it as an ordinary chondrite body. The new study places Kamoʻoalewa in that same broad company.
What this does and does not prove
This is a competing explanation, not a verdict. The authors are careful to say Kamoʻoalewa probably came from the Flora family, and the hedge is doing real work. Reflectance spectra are notoriously slippery; weathering, grain size, viewing angle and composition can all nudge an asteroid’s colour in similar directions, which is why two capable teams can look at the same faint object and reach different conclusions.
The lunar-origin case has not been withdrawn, and the dynamical argument for a recent lunar escape still has its defenders. It is also worth noting that the new paper is an early-access version of an accepted manuscript, released ahead of final editing, so small details may still change. What the study does is weaken the single strongest pillar of the Moon theory, the matching red spectrum, by showing that a plain meteorite can be pushed to look the same.
In other words, the resemblance to the Moon is real. The new work simply argues it is a coincidence of weathering rather than a sign of shared birth.
The sample that will end the argument
What makes this debate unusual is that it has an expiration date. Most arguments about an asteroid’s origin are settled, if at all, by ever-finer remote measurements. This one will be settled by holding the rock.
The stakes are not trivial. If the grains turn out to be lunar, Kamoʻoalewa becomes a free sample of the Moon’s far side, delivered to Earth’s neighbourhood without a crewed landing. If they are ordinary chondrite, the asteroid becomes a well-preserved relic of the early Solar System and a test case for how badly space weathering can disguise a familiar rock. Those are two very different prizes, and the same capsule settles which one we get.
Tianwen-2 is expected to collect material from Kamoʻoalewa within weeks and send a capsule back toward Earth in 2027, after which the spacecraft heads onward to a comet. A returned sample can be measured for the isotopes and minerals that distinguish lunar rock from a chondrite asteroid, the kind of test no telescope can perform from afar. For once, a question about where a space rock came from will be answered by the rock itself, and the answer is already on its way.
Login
