A supernova signals a star's death throes. Having exhausted its fuel for nuclear fusion, the star collapses, producing a gigantic explosion of matter and energy that can be seen from 10 billion light-years away. The supernova shines for a few months, then fades. All that remains after the cosmic light show is either a dense, smoldering core, called a neutron star, or a gaping black hole.

At least, that is what's supposed to happen.

Some 500 million light-years away, in a galaxy so distant it looks like little more than a smudge, a star exploded five times over the course of nearly two years, spewing the contents of 50 Jupiters and emitting as much energy as 10 quintillion suns.

This isn't even the first time this star has gone supernova: Astronomers believe this same body was seen exploding 60 years ago.

Somehow, this “zombie” star has managed to survive one of the most powerful, destructive events known to science — multiple times. It should make us question, researchers wrote Wednesday in the journal Nature, how much we really know about supernovas.

The discovery was made by scientists working on the Intermediate Palomar Transient Factory, which uses a telescope near San Diego to survey the night sky for ephemeral events like supernovas. Iair Arcavi, an astrophysicist at the University of California at Santa Barbara and Las Cumbres Observatory who worked on the project, was mainly interested in stars in the early stages of explosion. So, in September 2014, when the survey captured a fading supernova near the constellation Ursa Major, he didn't give it much thought. The event looked like a garden variety star well on its way toward oblivion.

Five months later, an intern who had been assigned to look over old data asked Arcavi to look at something weird. The intern pulled up a plot of the supernova's emissions over the past 137 days — bizarrely, the explosion was getting brighter.

Figuring that this must be a fluke — maybe just a star in our galaxy twinkling weirdly — Arcavi broke the light from the explosion into its component wavelengths. This “spectrum” contained all the signatures of a supernova.

Even stranger, it looked like a nova that was only 30 days old — though the scientists had concrete proof that it had in fact been going on for months.

“That's when it became very puzzling,” Arcavi recalled.

The event, dubbed iPTF14hls, was put on 24/7 watch. The eyes of the Las Cumbres Observatory — a robotic network of telescopes positioned all over the world — followed the supernova as it brightened, then faded, then brightened again. The nova hit five peaks of brightness before finally seeming to dwindle in summer 2016. But at 600 days old, it was already the longest-lived supernova ever observed.

Meanwhile, one of Arcavi's colleagues started searching astronomy archives in the hopes that someone else might have observed this renegade explosion. It turned out that the very same Palomar telescope had examined this portion of the sky twice before, in 1954 and 1993.

No supernova was visible during the more recent observation. But in 1954, there was a noticeable bright spot in the galaxy that's home to iPTF14hls: another supernova.

Because the galaxy is so distant, earthbound telescopes can't distinguish the light of individual stars within it. There is a small chance that the 1954 explosion is a coincidence — another star that happened to blow up in the same galaxy. But given what astronomers know about the frequency of supernovas, Arcavi put this possibility at no more than 5 percent.

“We’ve never really caught two supernovae happening in the same place before,” he said. “It would be really weird if the first time we see it is also the strangest supernova we’ve ever seen.”

According to Arcavi, no model of stellar evolution can fully explain astronomers' observations of iPTF14hls. One scenario that comes close is known as a “pulsational pair instability supernova” — a kind of “impostor” supernova that happens when a star 100 times bigger than the sun is destabilized and begins to blow off some of its outer layers.

“This process can repeat every few years or every few decades,” said Arcavi. “It’s an attractive theory because it explains why we may have seen this thing explode multiple times.”

But even that model can't keep up with the truly bizarre behavior of this supernova. For one thing, the energy released in the 2014 explosion is greater than what the model predicts will be the total output of all explosions a star might go through. Analysis of the types of light coming from the explosion indicates that its chemical composition is different from what astronomers might expect.

“Either this might be the first pulsational pair instability supernova, but the theory needs to be modified,” Arcavi said. “Or it might not be that, in which case the theory needs to be something completely new.”

Arcavi and his colleagues are scheduled to make follow-up observations with the Hubble Space Telescope next month. The powerful space-based observatory should be able to peer into iPTF14hls's host galaxy at some of the supernova's neighbors, which may illuminate details about the star that spawned this strange explosion. They'll also be keeping an eye on the supernova, since Arcavi is by no means convinced it won't act up again.

In an analysis for Nature, Stan Woosley of the University of California at Santa Cruz, who was not involved in the research, wrote that a better understanding of iPTF14hls could lead to revelations about the evolution of massive stars, the emergence of extremely bright supernovas and, maybe, the origins of the kind of black holes we've detected with gravitational waves.

“For now,” he concluded, “the supernova offers astronomers their greatest thrill: something they do not understand.”

Read more: