Astronomers directly track a star spiraling inside a giant in Andromeda
Astronomers say they have, for the first time, directly tracked over several years a star spiraling inside the swollen envelope of a giant companion, offering a rare observational test of a central idea in how binary stars evolve. In a paper led by Viraj Karambelkar posted to arXiv on Nov. 17 and reported as published online by Nature on Nov. 18, researchers argue that the event, seen in the Andromeda Galaxy, captures a long-theorized phase known as common-envelope evolution.
The object, called AT 2025abao, erupted on Oct. 17, 2025, after what the team says was a roughly seven-year slow brightening seen in archival and survey data. The outburst was detected including by the Zwicky Transient Facility, and the source was later classified as a luminous red nova, a kind of stellar-merger-type transient. Karambelkar said in NASA’s Nov. 18 summary of the work: “Stellar mergers are very, very difficult to observe in real time because they are hard to predict… However, we noticed something unusual was going on in 2023 and continued to monitor the system to see what it would do. Then, in 2025, it erupted!”
Common-envelope evolution is the stage when one star swells enough to engulf its companion in a shared gaseous envelope. Friction inside that envelope drains orbital energy and angular momentum, pulling the two objects closer together. Astronomers have long relied on that process in models explaining how initially wider pairs can become tight binaries containing white dwarfs, neutron stars or black holes. But direct observations of the inspiral itself have been scarce.
That is what makes AT 2025abao potentially important. Karambelkar’s team identifies the primary as an asymptotic giant branch, or AGB, star — a late-life giant star — with a mass of about seven times the sun’s and a radius about 350 times the sun’s. In the team’s preferred interpretation, a companion about 0.7 times the sun’s mass spiraled into that giant envelope, and frictional drag powered the years-long precursor brightening before the final eruption.
“When a companion star begins to interact with the outer surface of the primary star, it produces a subtle effect — a slow brightening, which we clearly saw in Bao’s data,” Kishalay De said in NASA’s summary.
The study draws on a broad set of observations, including data from NEOWISE, WINTER, ZTF and SPHEREx, along with later spectroscopy and follow-up from the James Webb Space Telescope. NASA said Webb obtained mid-infrared observations on Dec. 24, 2025, and that those data were important for constraining the dust produced in the event and the system’s overall energetics.
From those observations, Karambelkar and colleagues argue that the final phase was a rapid merger between the companion and the primary’s core over fewer than 10 orbits. They say the event ejected the outermost roughly 10% of the envelope at speeds about 10 times the primary star’s escape speed. The paper also derives an empirical range of about 0.01 to 0.4 for the common-envelope energy-unbinding efficiency parameter, known as alpha, a poorly constrained quantity that affects how efficiently orbital energy can expel a star’s envelope.
That parameter matters well beyond this one outburst. Because common-envelope evolution is used throughout binary-star theory, even a rough observational constraint could help sharpen models of how close compact systems form. The authors present AT 2025abao as the first direct multi-year observational characterization of a common-envelope inspiral inside a giant star, distinguishing it from earlier merger cases such as V1309 Sco, which involved a main-sequence binary rather than an inspiral inside a giant or AGB envelope.
The observation itself — a years-long brightening followed by eruption in Andromeda — is not in dispute. But some of the key physical inferences are. An independent modeling paper by Chen and colleagues, posted to arXiv on Oct. 5, fits the same object differently, inferring a companion of about 1.4 solar masses and finding scenarios in which the companion survives in a tight, eccentric orbit rather than necessarily merging. That means the merger interpretation, as well as the precise masses and ejecta estimates, remains model-dependent.
Even with that caveat, the system offers astronomers something they have rarely had before: a chance to watch the theorized early warning signs of common-envelope evolution unfold in real data. As De put it, “When a companion star begins to interact with the outer surface of the primary star, it produces a subtle effect — a slow brightening, which we clearly saw in Bao’s data.”