For seven years, a giant star in the Andromeda galaxy grew steadily brighter. It expanded to more than twice its original size while its surface temperature barely changed. Then, in October 2025, it erupted. The slow brightening and sudden outburst together tell the story of a smaller star orbiting inside a massive companion—and what ultimately became of the pair, researchers report November 18 in Nature.
In work published last year, astronomers analyzed nearly 30 years of telescope data and found that the giant star had begun brightening in 2018. They proposed that a smaller star was circling ever closer to the giant and had even slipped inside its outermost layers—a fleeting phase in a binary star’s life known as common-envelope evolution.
Now the astronomers have examined the system’s dramatic subsequent developments, culminating in the October 2025 eruption. The stars are nicknamed Little Bao and Big Bao, a nod to the event’s official designation, AT 2025abao.
First, astrophysicist Viraj Karambelkar of Columbia University and colleagues revisited the years of brightening in light of the imminent eruption. One possible cause was gas flowing away from the stars and emitting light as it slowed. But even if all the energy of that moving gas had been converted into light, it would not have been enough to explain the glow.
Instead, the researchers found that their original binary hypothesis fit better: as Little Bao orbited inside Big Bao’s gaseous outer layers, the gas pushed back against the smaller star. That drag caused Little Bao to spiral inward and transfer energy to the surrounding gas. Some of that energy escaped as light, making Big Bao shine brighter.
The team then investigated what caused the eruption. If the outburst had thrown enough of the shared gas into space, the individual stars could have survived in an even tighter orbit. But the team estimates that only about a tenth of the giant’s outer gas escaped. The energy carried by that fast-moving gas also suggests that Little Bao reached Big Bao’s core. Together, those clues point to a merger.
Too much dust still clouds the system to directly observe what remains, and it will not settle for years or even decades. Karambelkar and his colleagues plan to keep watching.
They will also search for more examples of this type of merger. “Now we can go and try to look for similar systems,” Karambelkar says, “and hopefully aim to catch more of these systems even before they erupt.”


