A giant star fakes a supernova every 113 days because a collapsed star feeds on it

Roughly every hundred and thirteen days, a giant star in a nearby spiral galaxy flares so brightly that it passes for a supernova. It has done this at least twenty-seven times, and it is still there. X-ray observations from NASA’s Chandra observatory now point to what drives the blasts: a black hole or a neutron star on a stretched orbit. It swallows gas pulled off the giant each time the two swing close.

That makes the star, catalogued as AT 2016blu, the first star of its unstable class known to have a collapsed stellar remnant as a partner. It also links two puzzles astronomers have mostly studied apart: supernova impostors, stars that erupt like supernovae without dying, and the brightest X-ray sources in other galaxies, whose engines are still debated.

The star is enormous. Hubble images taken while it was quiet put it at no less than 33 times the mass of the Sun and at least half a million times its brightness. It sits in the galaxy NGC 4559, about 29 million light-years away, and belongs to the same unstable family as Eta Carinae. Its steady rhythm had already told the team it orbits something.

At the peak of an outburst it shines in red light like roughly ten million Suns: far fainter than a true supernova, but bright enough to land in supernova surveys. “This star’s behavior was confusing at first because it kept acting like a supernova, but clearly it’s not because it keeps surviving these outbursts about three times a year,” said co-author Núria Torres-Albà of the University of Virginia.

How Chandra caught the star mid-outburst

Two massive stars with colliding winds give off modest X-rays. Gas falling onto a black hole or neutron star can shine thousands of times brighter. To tell them apart, Chandra had to look during an outburst, which meant knowing when one was coming.

So the team watched. Telescopes at Lick Observatory and the Las Cumbres Observatory network checked the star almost nightly, and an alert from the American Association of Variable Star Observers brought in more than thirty observers, many of them amateurs. When the star brightened by more than a full magnitude, the team triggered an urgent Chandra request. The telescope observed it five times over four days, for about 100,000 seconds in total, and detected it every time.

“We suspected that the bright repeating flashes at optical wavelengths might actually be powered by periodic blasts of high-energy X-rays,” said co-author Nathan Smith of the University of Arizona. “The dense gas surrounding the system is heated by the X-rays and then glows in the optical.”

Only Chandra could do it. A much brighter X-ray source, NGC 4559 X7, sits about ten arcseconds away, and Europe’s XMM-Newton telescope blurs the two into one spot. Chandra had also imaged the field by chance more than two decades earlier, while AT 2016blu was calm, and saw nothing at its position. The new detection is at least 100 times brighter than that limit. The odds that a background quasar happened to sit inside the tiny search circle are about six in 100,000.

Why only a black hole or a neutron star fits

The paper weighs four ways a star like this can make X-rays. Shocks inside a single star’s wind top out more than ten million times fainter than what Chandra measured. Colliding winds from two massive stars fall short by a factor of about 10,000. Shells of gas thrown off on successive orbits could crash into each other, but at the speeds seen in the star’s light that would give soft, low-energy X-rays, and Chandra recorded a hard, high-energy spectrum.

That leaves accretion, matter falling onto a compact object. The X-ray output equals the theoretical ceiling, called the Eddington limit, for an object of about 3.5 solar masses, which is already above the ceiling for a neutron star. It falls just short of the cutoff for an ultraluminous X-ray source, a class of extreme emitters found in other galaxies, and Chandra may simply have missed the true peak. The team calls AT 2016blu a candidate member.

The visible outbursts carry about a hundred times more energy than the X-rays. If both come from the same feeding, about 99% of the energy is soaked up by surrounding gas and re-emitted as visible light, as Smith describes.

How it compares with Eta Carinae and other impostors

At its X-ray maximum, Eta Carinae, whose X-rays come from colliding winds, reaches roughly a two-thousandth of what AT 2016blu produces. Most luminous blue variables in our galaxy show no X-rays at all.

The nearest match is SN 2010da, an X-ray-bright impostor in the galaxy NGC 300 whose regular pulses revealed a neutron star. Its big star, though, is a red supergiant, not a luminous blue variable.

That matters because of SN 2009ip, a luminous blue variable that erupted for years and then appears to have exploded for real, which single-star models do not expect. A growing set of models blames a compact partner for such pre-death outbursts, and AT 2016blu is the first observed luminous blue variable that fits.

What the X-rays cannot tell yet

The biggest open question is what the partner is. Neutron stars can be identified by regular X-ray pulses, but Chandra collected too few X-ray photons to search for them, and XMM-Newton’s view is swamped by X7. Weighing the partner would need an orbit solution, and the speeds measured in the star’s hydrogen light do not track the orbit, so no mass comes out.

The feeding rate leans toward a black hole without proving it. Taking the visible outbursts into account, the companion may be swallowing gas at more than a hundred times the rate a typical neutron star is supposed to handle. Some neutron stars have been seen exceeding that limit about a thousandfold, so the authors stop short of a verdict.

The star’s spectrum also shows features that come and go without lining up with the orbit, which could mean a jet, a disk or clumpy gas is involved.

Then there is why the eruptions started at all, after years of a stable star. The team favors a giant that has slowly swollen with age until it began to reach its partner. The alternative is a temporary puffed-up phase known in these stars, in which case the outbursts should eventually stop. So far they have not.

Common questions about supernova impostors

What is a supernova impostor?

It is a massive star that brightens enough to be mistaken for a supernova but survives, often a luminous blue variable such as Eta Carinae. Real supernovae destroy the star and are usually far brighter. Impostors are often logged by supernova surveys and reclassified once the star turns up again.

How can X-rays reveal a black hole or a neutron star?

Gas falling onto a very compact object heats to millions of degrees and glows in hard X-rays at power levels ordinary stars cannot reach. AT 2016blu shines at roughly 100,000 times the Sun’s total output in X-rays alone, far beyond what colliding stellar winds can produce. Telling a black hole from a neutron star needs extra evidence, such as pulses or a measured mass.

Will AT 2016blu explode as a real supernova?

Nobody knows. The star is massive enough to end as a supernova eventually, and the impostor SN 2009ip shows that a star with a similar history can go on to explode. Nothing in the new data says when.

The next close pass

Chandra observed the star from March 24 to 27, 2026, after archival looks in 2001 and 2002 and a first recorded eruption in 2012. The results were published on November 9 in The Astrophysical Journal. “For 15 years, astronomers have been watching this star get really bright and then dim again on a regular basis,” said lead author Mojgan Aghakhanloo of the University of Virginia. “Until we were able to get Chandra data at the right time, we couldn’t figure out what was making this happen.”

The authors argue that finding more systems like this one is essential to learning how impostors, luminous blue variables and real supernovae connect. AT 2016blu keeps to its schedule, so each close pass is another chance. What would close the case is X-ray pulses, which only a neutron star can make, or an orbit that finally weighs the partner.

Reference: Aghakhanloo et al., “AT 2016blu: Accretion-powered Outbursts in a Luminous Blue Variable and Compact Object Binary,” The Astrophysical Journal, 2026. DOI: 10.3847/1538-4357/ae9a86

Tags: astronomy, Supernova, black hole, AT 2016blu, Chandra X-ray Observatory, neutron star