Mysterious Bright Flashes in the Night Sky Baffle AstronomersNEWS | 18 December 2025Long, long ago a cloud of stars circled a galaxy-size black hole, safely at a distance. Then about 200 million years ago one member of the cloud bumped another, a sun-size star, and sent it toward the black hole. The black hole was a million times more massive than the sun-size star, and its gravitational pull proportionately stronger, so the star was drawn closer and closer—until it got too close. Some of the star’s gas was pulled into an orbiting stream around the black hole that widened into a flat pancake called an accretion disk. The rest of the star came apart in a sudden and great flash of light.
On September 19, 2019, just before noon, the flash reached the 1.2-meter mirror of the Zwicky Transient Facility in southern California. Astronomers named the flash AT2019qiz and noted that they hadn’t seen it three days before. On September 25, 2019, the 10-meter Keck I telescope in Hawaii identified AT2019qiz as a so-called tidal disruption event—a flare-up that occurs when a black hole’s gravitational tides rip a small object apart. The star the size of the sun exploded with 10 billion times the sun’s luminosity.
But AT2019qiz wasn’t finished yet. An entirely unrelated star, maybe from the same cloud, was on an orbit that intersected AT2019qiz’s newly created disk. Each time this other star splashed into the disk, it flashed, though less brilliantly than the original, pulled-apart star. In December 2023 the brightness of AT2019qiz (now the name of the disrupted star, the accretion disk and the flaring star that ran into them) peaked, dimmed down and then shot up again—a pattern that repeated nine times. Each flash marked a pass of the interloper through the disk, which occurred every 48 hours. Between 2019 and 2024, astronomers observed AT2019qiz with telescopes on the ground and in space, at wavelengths from x-ray through ultraviolet, optical and infrared. The multitelescope, multiwavelength data together confirmed that AT2019qiz was first a tidal disruption event and then a “quasi-periodic eruption.” Both are examples of phenomena astronomers call transients. Both involved unspeakable violence on unearthly scales. Neither could have been identified 20 years ago.
On supporting science journalism
If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
Transients, which are astronomical objects that appear suddenly from nowhere and usually disappear soon after, contradict the standard truth that the universe changes predictably and slowly over billions of years. They include what the typically staid National Academy of Sciences called “the most catastrophic events in spacetime.” They are astronomically sized objects that change on human timescales—in seconds, hours, days—which is a combination of size and speed that seems impossible. If we didn’t observe them, says astronomer Vikram Ravi of the California Institute of Technology, “you’d never know that physics allows these things to exist.”
But physics says everything not forbidden will, sooner or later and with some probability, happen. And astronomers, noticing these improbable things and knowing that nothing is one of a kind, began to find many more, all at the far reaches of physics. Between 1976 and 2012, the number of transients listed on the International Astronomical Union’s official Transient Name Server was around five each year. Between around 2013 to 2015, that number jumped to about 100. Since 2019, scientists have seen roughly 20,000 a year. At press time, the total was 175,953 transients. Chart this rise, and it looks like a long tail with an elephant attached.
The growth has been the result of a large number of astronomical surveys, most still ongoing, “vacuuming the whole sky,” says experimental physicist Christopher Stubbs of Harvard University. For instance, the Zwicky Transit Facility, which started the jump in detections in 2019, scans the entire northern sky every two nights and compares each evening’s images with the ones taken two nights before. And the Vera C. Rubin Observatory in Chile, which came online in 2025, will soon survey the entire southern sky every three nights, identifying changes within 60 seconds of their detection to create near-real-time movies of the sky and finding 10 million changes every day. The elephant will go seriously nonlinear.
With such a large amount of data, astronomers can begin to study credible demographics: that is, they can move from just finding these wild, unlikely creatures to figuring out what they are. Because things that happen once and disappear are hard to study, the transients’ identities—the physics that drives them, the processes that produce them—are still speculative. Most of their names are just adjectives, and “when the transient’s name is a description,” says astrophysicist Raffaella Margutti of the University of California, Berkeley, “that tells you we know nothing intrinsic about them.” That’s about to change.
Scientists sort transients into two main groups: events involving the deaths of stars and events around supermassive black holes in the centers of galaxies. The first known transients fell into the former category: they were supernovae, or massive stars that blow up. Before the 1600s, astronomers confidently knew of five of them; now they count tens of thousands. Supernovae fit into two general categories. One kind is the dead core of a star pulling gas from a nearby star, piling up mass until nuclear fusion restarts and goes critical and the whole thing pops off like a 20-billion-billion-billion-megaton thermonuclear bomb, which it is. It explodes in a day, stays bright for days to weeks, and fades out over months.
The other type of supernova is called a core collapse: A star burns through enough of its fuel and is massive enough that the outward push of its radiation loses to the inward pull of its gravity. Its core collapses in on itself so thoroughly that its electrons meld with the nuclei of its atoms until the star is made mostly of neutrons—a neutron star—and it shrinks in the space of one second from a radius of about 6,000 kilometers to about 10 kilometers. The collapse causes a shock wave that breaks out of the star’s remaining atmosphere with a flash called a shock breakout, and minutes later the star is as bright as 10 billion suns. It fades out over months; the remnant is called a neutron star.
Beyond these two main categories, though, are many variants—the Transient Name Server identifies 31 types so far. One new kind, called a gap transient, is dimmer and probably less massive than other supernovae, and nobody knows why it explodes. Another is a superluminous supernova, twice as luminous as a core collapse supernova; it has the light of 20 billion suns, and nobody knows why it’s so bright. Supernovae are by far the most numerous of the stellar-death transients, but, as astronomer James E. Gunn of Princeton University points out, stars have “a vast number of interesting ways to die.”
Ron Miller (illustrations) and Jen Christiansen (graphic)
In 1967, for instance, the U.S. Vela satellites detected surprising flashes of extremely energetic gamma rays that could have been (but weren’t) illegal nuclear tests in Earth’s atmosphere; the National Enquirer thought similar flashes seen later might be a space war between alien civilizations. Eventually astronomers pooled data from the U.S. and the U.S.S.R. to identify the flashes, which were the first known gamma-ray bursts—a class of transients now understood to be “the brightest of the brightest,” says astrophysicist Peter Jonker of Radboud University in the Netherlands, who observes space in high-energy wavelengths. Their light rises in seconds to the brightness of a trillion suns, and they last for seconds to hours. The fastest ones might be massive stars going supernova, collapsing so thoroughly that they don’t stop at neutron stars and instead condense into star-size black holes that aim high-intensity jets of plasma at Earth.
Gamma-ray bursts may or may not be related to other high-energy stellar deaths called fast x-ray transients. Discovered in 2008, they number only around 70, although this tally will soon change. China’s Einstein Probe, an x-ray satellite telescope that began collecting data in mid-2025, should find 50 to 100 fast x-ray bursts a year. “The next few years could be dramatic,” says astronomer Mansi Kasliwal of Caltech. Meanwhile, because fast x-ray transients are still rare, no one is ready to say what they are—maybe massive stars exploding, maybe neutron stars colliding before disappearing into black holes.
Another dramatic rarity is called a fast optical blue transient, or FBOT—“fast” because although it explodes at the same outrageous brightness as a superluminous supernova, its light rises and falls not in months but in days. The first FBOT, found in 2018, is officially named AT2018cow and is called Cow for short. Since then, scientists have seen 12 more Cow-like FBOTs. Astronomers know they’re not supernovae—“the energy source of the normal supernovae doesn’t work” for Cows, Margutti says—but aren’t sure what they are. Maybe they flash when a nearby star’s mass piles up onto a neutron star or a modest-size black hole, or maybe they represent shock breakouts from a star that puffed up in its later years. “Whatever they are,” says astronomer Anna Y. Q. Ho of Cornell University, who helped to find the original Cow, “they’re interesting.”
In 2007 radio astronomer Duncan Lorimer and astrophysicist Maura McLaughlin, who are colleagues at West Virginia University and married, were looking in the archives of a radio telescope survey at a small galaxy 200,000 light-years away. They were interested in pulsars, which are rotating neutron stars that release jets of radio light from their magnetic poles. These lighthouselike jets sweep the sky so that whatever is in their path is exposed to a metronomically regular radio pulse every few seconds to milliseconds.
In the course of their search, Lorimer and McLaughlin found a radio spike that lasted a few milliseconds, but it didn’t pulse and was so bright it saturated the telescope’s instrument. Lorimer calculated its distance as seven billion light-years away. “Oh,” he thought, “it’s really far.” Anything that distant and still that bright had to be sending out a billion times more energy than nearby pulsars.
This odd find is now called the Lorimer Burst. Surveys have since identified several thousand of these so-called fast radio bursts scattered throughout other galaxies, emitting in one millisecond the radio energy sent out by the sun in 100 years. “These things are weird,” Lorimer says.
Some of these possible stellar death transients could be related to a deeply strange object called a magnetar. Magnetars existed only in theory until they were observed in 1998. Their weirdness quotient is high even among transients. A magnetar is a neutron star that “rotates ridiculously fast,” making a full turn in milliseconds, says Daniel Kasen of the University of California, Berkeley, “but with a ridiculously high magnetic field.” The strength of the sun’s magnetic field is somewhere around 10 gauss; a magnetar’s is 1014 gauss or higher. That field is “so high it’s unstable,” Ravi says. “It chaotically reconfigures itself.”
Transients are astronomically sized objects that change on human timescales—in seconds, hours, days.
The object’s magnetic field lines twist and snap and reconnect, and in the process they send out flares. The combination of absurdly strong magnetic fields and absurdly fast rotation leads to lots of explosive physics, Kasen says. In 2004 a flare from one magnetar halfway across the Milky Way ionized the upper layers of Earth’s atmosphere. Astronomers know of around 30 of them in our galaxy so far.
“Magnetars are invoked to explain a lot of things we don’t understand,” says Brian Metzger of Columbia University, a theoretical astrophysicist who specializes in stellar-death transients. For instance, different transients might be different phases of a magnetar’s life. Magnetars might be born in the core collapse of the same massive stars as superluminous supernovae. A supernova might then condense into a pulsar and send out jets that are seen as gamma-ray bursts. Later, when the magnetar’s spin period has slowed from milliseconds to seconds, its flares may be seen as a fast radio burst. Magnetars might even explain FBOTs, Ho says, but so far FBOTS are too distant for scientists to be sure.
The stellar-death transients are dying in ways intrinsic to stars. But stars can also die because they’re just in the cosmically wrong place, in the nuclei of galaxies with supermassive black holes. These “nuclear transients,” the second overall category of transients, have turned up only in the past decade. They’re rare and barely understood.
One reason for that is that nuclear transients “are a minefield of contamination,” says Suvi Gezari of the University of Maryland, College Park. Astronomers must distinguish the flashes of nuclear transients from supermassive black holes whose behavior varies. One percent of supermassive black holes, the quasars, are furiously, actively accreting gas and shine so brightly they can be seen near the beginning of the universe. Most of the rest are inactive and just flickering; they have gravitationally cleared out much of the space around them, and their brightness varies by just 10 to 45 percent. And another, unknown fraction are not accreting at all; they’re completely black and invisible.
Nuclear transients are not active quasars, and they don’t flicker—they’re cosmic flash-bangs. One kind is a tidal disruption event such as AT2019qiz, a star trapped in a supermassive black hole’s gravitational field and torn to smithereens. Astronomers have found around 100 tidal disruption events, each visible for a few months in the x-ray, optical and ultraviolet ranges, each with its own small accretion disk that lasts for a few tens of years. Maybe one in 10 tidal disruption events do what AT2019qiz did and become the site of another kind of nuclear transient, the quasi-periodic eruption. In these cases, an errant star passes through the tidally disrupted star’s accretion disk and flares up in x-rays to the brightness of a billion suns. Such flares last minutes and repeat in hours to weeks.
Other nuclear transients may not involve stars at all and may reflect odd behavior of the black holes. One kind of transient discovered in the past decade is called a changing-look quasar (CLQ). It has the brightness of a normal quasar but rapidly changes its appearance in unexplainable ways. It should take thousands of years for a quasar to switch off and go from brilliantly active to quietly inactive. Yet astronomers have found dozens to hundreds of CLQs that change their looks by 200 percent in months—they change so much and so quickly that “they’re not theoretically explainable,” says astrophysicist Paul Green of the Center for Astrophysics | Harvard & Smithsonian. Maybe they’re the aftermath of a long-gone tidal disruption event, or maybe, he says, “we haven’t watched long enough to see a change of state that’s lasting.”
As if CLQs weren’t improbable enough, astronomers also find ambiguous nuclear transients (ANTs), whose problem is in their name: “They’re ambiguous,” says astrophysicist Philip Wiseman, who studies nuclear transients at the University of Southampton in England. They are a diagnosis of exclusion, a flash that isn’t any other transient. ANTs are brighter than all transients except gamma-ray bursts. Their light rises slowly over months and lasts for two or more years. They’ve been found in data archives in numbers from a few to hundreds, depending on who’s defining them. “We can find them, but we don’t know what they are,” says astronomer Matthew Graham of Caltech, another nuclear-transients specialist.
These events are flashes of inconceivable amounts of energy.
One ANT discovered in 2020 became famous: At first astronomers thought it was an actively feeding supermassive black hole in the center of a galaxy, but they couldn’t find the galaxy. The lonely supermassive black hole, like a kind of negative island, is somewhere between 10 and 1,000 times the size of the one in the Milky Way. One of its names is ZTF20abrbeie; astronomers call it Scary Barbie.
ANTs could be outsize tidal disruption events—that is, instead of sun-size stars being torn apart by black holes with the mass of a million suns, they might be 10-solar-mass stars torn apart by black holes with the mass of a billion suns. Or they could be supermassive black holes moving from inactive flickering to active fiery accretion—black holes “turning on,” Graham says. Researchers are still looking for Scary Barbie’s galaxy. “We’re guessing at half this stuff,” Graham adds.
The obvious question is, Are some of these transients somehow aspects of the same thing? For stellar-death transients, the answer is not exactly no. Several of them may be related to one another or to magnetars; in general, they’re a menu of the variables that determine how stars end their lives. For nuclear transients, the answer is unsatisfying: either a captured star or a black hole’s accretion disk is brightening. For a better answer, astronomers need to collect many more nuclear transients.
Nor can stellar and nuclear transients be put together into a single grand unified theory. Such a picture should be based on their physics—specifically, the source of energy for their outbursts. “The holy grail is understanding what produced the transient,” says Eliot Quataert of Princeton, a theoretical astrophysicist studying nuclear transients. Theorists want to be able to slot energy sources into a few categories, such as radioactive decay, shocks and gravity, although some transients don’t seem to fit into any of these boxes.
To figure out the energy sources and maybe unify transients, astronomers need to compare what they see in different wavelengths, which each reflect different physical processes. In supernovae, for instance, ultraviolet light comes from shock breakouts, and x-rays and radio waves come from collisions between matter ejected in the explosion and the surrounding gas. Collecting every possible photon from every physical process allows astronomers to assemble a complete picture of the event.
Accordingly, telescopes now operating in optical, ultraviolet, x-ray, gamma-ray and radio-wave bands are about to be joined by a series of new telescopes in space. Among them are NASA’s Nancy Grace Roman Space Telescope, which will launch by mid-2027 and observe in the infrared; the Einstein Probe in x-ray; and NASA’s Ultraviolet Explorer, which will launch in 2030.
You might wonder whether this is a lot of telescopes and effort just to learn about 100,000 one-offs in a universe full of 10,000 billion billion stars in 100 billion galaxies. Understanding transients is important partly for answering other astronomical questions. Supernovae are used as distance markers to enable calculations of the universe’s acceleration. Both tidal disruption events and quasi-periodic eruptions hold evidence about supermassive black holes that are quiescent and therefore invisible, as well as about the all but theoretical class of black holes whose masses are between those of stellar black holes and supermassive ones. And fast radio bursts, because they are visible in the distant universe, can be used as searchlights to map the distribution of regular matter, of which only 10 percent is known.
But transients are also interesting for their own odd selves, for their ability to teach us what physics doesn’t forbid. Kasen says they are “laboratories for fundamental physics and extreme conditions”; they are “physics at the extreme,” Margutti says, “and I can’t probe that on Earth.” Transients show “the range of phenomena possible in the universe,” Ravi says.
These events are flashes of inconceivable amounts of energy released in the time it takes to buy groceries, drink a glass of water or snap your fingers. A supernova shock breakout travels the distance from Baltimore to Western Australia in half an eyeblink. A magnetar passing 160,000 kilometers away could demagnetize every credit card on Earth. A neutron star compresses a massive star to the length of a leisurely two-hour walk. The study of transients is certifiable science, but if it weren’t, it would still be reason for near-holy astonishment.Author: Clara Moskowitz. Ann Finkbeiner. Source
