In March of 1437, Korean astronomers in Seoul saw what they thought was a new, bright star appear in the night sky. Now, nearly 600 years later, astronomers have figured out what those stargazers actually saw: a thermonuclear explosion caused by the interaction of two distant stars. The new research pinpoints the location of those two stars in the sky, solving a mystery that’s plagued astronomers for decades and providing clues about what happens to pairs of stars centuries after they explode.

The event that the Korean astronomers saw lasted 14 days, leading modern astronomers to suspect it was something known as a classical nova. This is a type of explosion caused by an ordinary star similar to our Sun and a white dwarf — a small, super-dense stellar corpse left over when a regular star uses up all its nuclear fuel. The white dwarf is so dense that it sucks hydrogen off the other star, building up a layer of gas that eventually blows outward like a giant bomb. The process turns the Sun-like star into a red dwarf — a much smaller, cooler, and fainter type of star.

Astronomers have long thought that the 1437 event could help solve an astrophysical puzzle: what exactly happens to the two stars in a classical nova after the explosion? “In order to answer that question, it isn’t good enough to look at nova stars 20 or 30 or even 100 years after the eruption,” Michael Shara, an astrophysicist at the American Museum of Natural History who discovered these stars, tells The Verge. “It looks pretty much the same for the century before an eruption and after an eruption.” But there is an idea that many centuries after a classical nova, the two stars get much fainter and undergo much smaller explosions periodically over time.

Now, it looks like that theory may be true. The resulting red dwarf / white dwarf pair, described today in the journal Nature, are indeed fainter and undergoing periodic eruptions. This is obviously just one example, and more novas should be observed to confirm the theory. “But it’s very strongly supportive [of the theory],” says Shara.

Shara’s discovery is the culmination of a decades-old search; he started looking for the 1437 nova back in 1980, in an area of the sky where he thought it might be located. Based on the Korean astronomers’ notes, historians deduced that the event probably occurred between two stars in the tail of the constellation Scorpius. Shara started looking in that general area, using huge astronomical databases and star catalogs, but had no luck for decades.

Then everything changed a year and a half ago, when he decided to expand his search. “I asked the question ‘What if it isn’t those two stars? What if it’s the next two stars over?’” he said. “And then in about maybe 90 minutes I found [it]. Really it was just sitting over there in plain sight waiting to be found.” The red dwarf and white dwarf he spotted seemed to fit the bill.

But one thing that was still missing was proof: how could Shara know for sure that these were the exact same stars that caused the classical nova in 1437? He and his team came up with a way to check when the stars may have exploded.

Whenever classical novae occur, they send out a round, glowing shell of hydrogen gas speeding outward in all directions. Eventually, that hydrogen shell slows down and stops when it runs into the particles and dust that exist in the space between stars, called the interstellar medium. But the two stars that created that shell don’t stop moving. The pair that once began at the center of the expanding shell eventually becomes off-center. “Imagine you’re in a convertible and you throw confetti out of the car as you drive down the highway,” says Shara. “The confetti will stop very quickly, because it runs into the surrounding air. But your car doesn’t slow down, just because it’s so massive.” Similarly, the nova stars kept on going, while the nova shell stopped.

Shara and his team decided to use this displacement to determine when the stars first exploded. They dug up an old photographic plate from Harvard, taken in 1923, that had captured the star pair. (The stars had been mapped before, but no one had connected them to the 1437 event.) They then compared the position of the stars from nearly 100 years ago to their positions today, which allowed them to calculate where the stars would have been in 1437. Sure enough, they would have been exactly in the center of the hydrogen shell, confirming that they were the source of the explosion.

Now, observations of the star duo show that they’re behaving a lot like dwarf novae — explosions caused by a pair of stars that are similar to classical novae. But unlike classical novae, eruptions in dwarf novae are fainter and occur repeatedly every few days to decades, under different mechanisms. But the findings suggest that classical novae and dwarf novae may actually be the same star systems, just at different points in their lives. It’s something Shara and his team wouldn’t have been able to know if the stars hadn’t been examined over such a long timescale.

“Astronomy is concerned with both the large scale and the long term, and historical observations are often important for resolving evolutionary questions,” writes Steven Shore of the National Institute for Nuclear Physics in an article published alongside the Nature paper. He notes that this discovery is “a lovely piece of historical scholarship.”

But Shara says he is equally excited about how his team verified that this star system was indeed the one from 1437. Now, other astronomers might be able to do the same with other star systems. “The most important step forward here has been demonstrating you can use this proper motion clock,” says Shara. “Getting the date of something is very hard, measuring distances and how old they are are the greatest challenges we have. This is a new tool we’ve inserted into astronomers’ tool chest.”