Editorial NexoScience
How do you measure the age of a star? Nothing about the way one looks gives it away, and there is nothing to cut open and count. For stars roughly the size of the Sun the best trick available is spin: young stars turn fast, old stars turn slowly, and the long wind-down in between has served as a clock for decades. A study published in The Astrophysical Journal finds that the clock stops keeping honest time about halfway through a sun-like star’s life, which is exactly the stretch astronomers most wanted to measure.
A star slows itself down using its own magnetic field. The churning outer layers of a sun-like star act as a dynamo and generate a magnetic field, and that field takes hold of the wind of charged particles constantly escaping the surface. Every particle that leaves carries a little of the star’s rotation away with it, a process called magnetic braking. Because the braking looked steady and predictable, astronomers built an age-dating method on top of it, gyrochronology, which reads a star’s age from its rotation period much as the rings in a trunk give away the age of a tree. For a lone star with no cluster and no companion to help, it has long been the most practical clock available.
The stars that broke the method were timed by their own vibrations. Sound waves rattle around inside a star and make its surface pulse, and the pattern of that pulsing, visible in the flicker of the star’s light, encodes what the interior is like. Rotation leaves its own fingerprint there too, splitting the oscillation frequencies in a way that reveals how fast the star’s envelope is turning. That matters because the standard way to clock a star is to watch dark starspots rotate in and out of view, and old, quiet stars barely have spots to watch. Asteroseismology, the study of these stellar oscillations, works even on stars whose magnetic activity has faded, so it reaches precisely the population the starspot method loses.
The team fitted every star in the sample at once. Nicholas Saunders, an NSF graduate research fellow at the Institute for Astronomy at the University of Hawaiʻi at Mānoa, and his colleagues assembled 49 Kepler field stars with rotation periods measured this way, a small sample beside the thousands a starspot survey can gather, but one that reaches ages the starspot method never gets to. To anchor the early and uncontroversial part of the story they added the Sun and stars from three open clusters, Praesepe, NGC 6811 and NGC 6819, young enough to still be braking in the classical way. Running full stellar evolution models for every combination of mass, composition and braking strength would have been impossibly slow, so they trained a neural network to imitate the model grids, then fit the whole population in one hierarchical framework, with the braking law shared by every star and mass, age and composition left free for each one.
The brakes let go before a star reaches the stage the Sun is at now. The transition is governed by the Rossby number, the ratio between how long a star takes to turn once and how long a parcel of gas takes to complete one convective circuit beneath its surface, a dimensionless quantity that puts stars of different masses on the same magnetic footing. Below a critical value the star brakes normally. Above it, in this model, the loss of angular momentum simply stops. The fit places that critical value at 0.91 times the Sun’s own Rossby number, sitting clearly below the solar value of one rather than on top of it. A star like the Sun, in other words, has already crossed the line.
Past that point the textbook picture misses, and it always misses the same way. Averaged over the whole sample, standard spindown overpredicts rotation periods by 0.72 days, which sounds like nothing at all. Restricted to stars beyond the halfway mark of their main sequence lifetimes, the same model overpredicts by 6.00 days. The weakened braking model errs in the opposite direction and by roughly half as much across those same old stars, and a goodness-of-fit test came out near the value expected of a good model for weakened braking against something more than an order of magnitude worse for standard spindown. Old stars are turning faster than they are supposed to, and the gap widens with age rather than closing.
The clock stops right where astronomers needed it most. Once braking weakens, rotation changes only slowly, drifting with the star’s swelling size rather than with its magnetism, and a star can hold the same rotation period for billions of years. The authors put the practical limit for gyrochronology at around four billion years for sun-like stars, and nearer two and a half billion for stars a little heavier than the Sun, which reach the threshold sooner and lose the clock earlier. They are careful to add that a rotation period still brackets a star’s age after the transition. What changes is that the answer has to come with a far wider margin, and their measurement of the threshold is what makes that margin calculable rather than guessed.
The authors are explicit about what their model does not capture. Shutting angular momentum loss off completely at a critical value is, as they put it, the simplest case rather than a description of what the dynamo is really doing, and they caution that it likely does not capture how the transition unfolds over time. Calibrator stars near the threshold are sparse, which the researchers say leaves the precise onset relative to the Sun’s own evolution hard to pin down. Their model also begins underpredicting rotation periods near the very end of the main sequence, where a star’s expanding radius takes over. Three stars were dropped from the sample outright because they spin too fast for their age, a pattern the team suspects means they are unrecognized binaries rather than single stars behaving strangely.
Repeating the fit with a different set of stellar models moved the answer. When the researchers retrained their emulator on grids built by a second stellar evolution code, YREC in place of MESA, the inferred braking parameters shifted at about the one-sigma level, and the difference traced mainly to how each grid handles the amount of helium a star was born with, a quantity that is notoriously hard to measure in sun-like stars. They conclude that this helium uncertainty has to be carried as a systematic uncertainty on the threshold itself, and it is the first time the effect of that modeling choice has been quantified for this problem. Worth being plain about, too: nobody watched a magnetic field weaken here. This is a comparison of models against measured spins, and the physical reason the braking should falter, possibly a shift in the complexity of the field itself, is still unresolved.
What would settle it is more old stars measured close to the moment the brakes let go. With enough of them, the researchers write, they could test a gentler version of the transition, one where braking fades gradually or weakens without ever switching off, instead of the on and off switch they were forced to assume. Direct measurements of stellar magnetic field strength would give an independent handle on the same question. Until then, the safest thing to assume about an old sun-like star is that it has been coasting for a while, at a speed that no longer says how long it has been out there.
Source: “Stellar Cruise Control: Weakened Magnetic Braking Leads to Sustained Rapid Rotation of Old Stars,” by Saunders, van Saders, Lyttle and colleagues, published in The Astrophysical Journal in 2024. DOI: 10.3847/1538-4357/ad1516
