A link between rotation and age for Sun-like stars has long been known, but a stringent test of it for older stars has been lacking. The Kepler mission helps to fill this gap with observations of an old star cluster. See Letter p.589
The clocks of the cosmos tick constantly, but too softly for us to hear, and so we cannot directly measure the ages of stars. Yet we need good ages, because nearly every aspect of astrophysics deals with how things evolve with time. Even the Sun is silent on this score, and it is only from being able to study Solar System material — meteorites — in the laboratory that we know the Sun's age so exquisitely: 4,567±1±5 million years1. In this issue, Meibom et al.2 (page 589) describe a key step needed to obtain better estimates for the ages of cool, low-mass stars like the Sun.
Many of the stars for which we would like to know the ages are like the Sun. They live for 10 billion years or more, and so represent the entire age of our Galaxy's disk, which is where most of the Galactic stars reside. Sun-like stars are the ones that naturally excite the most interest in our quest for other Earths, and, indeed, the first question that will be asked when someone reports signs of life on an exoplanet will be, 'how old is the host star?', because we will want to place the discovery in an evolutionary context.
What makes the Sun and stars like it favourable for life-bearing planets is that they change very slowly over time. But that also makes them inaccessible to age estimation by conventional means, which is by determining a star's temperature, luminosity and composition and then comparing them with stellar models. Those models themselves are calibrated against the Sun, the only star with well-established fundamental properties.
Given this problem of estimating ages, we settle for what we can get3, and it has been suspected for some time that, for Sun-like stars, rotation declines with age on something close to a power-law relation4 — a nice straight line in a log–log plot. This is convenient, in that our concern with accuracy roughly scales as the age itself, and if we could reliably derive ages that are good to 5% or better, that would improve our knowledge of stellar and Galactic processes substantially.
But why does the Sun spin so slowly, as much as 100 times more slowly than some young Sun-like stars, and is the spin rate a reliable clock? As was shown5 in 1967, slow rotation is a property shared by all Sun-like stars, and the mechanism underlying the phenomenon starts right where the outer layers of the stars begin to become convective. Thus, it is convection that makes the Sun and stars like it Sun-like. In the Sun, convection and rotation lead to complex motions of the conducting plasma of the convective outer layer. Those motions generate a magnetic dynamo that is based in or just below the outer layer. The magnetic field produced by the dynamo can then grip the solar wind of high-speed charged particles beyond the Sun's surface, and so angular momentum is steadily lost.
The loss of angular momentum is a slow process for the present-day Sun, but we also know that, because of their faster spin rate, young Suns generate much stronger magnetic fields. The magnetic dynamo provides a feedback mechanism that causes convergence in the spins of stars that have the same age but different initial rates. Observationally, that convergence seems to occur by about the age of the nearest loosely bound star cluster, the Hyades — that is, 600 million years or so.
Star clusters are fundamental to studies of angular-momentum loss because they provide good-sized samples of stars sharing the same composition and age. But the problem is that there are few clusters older than the Hyades. Clusters get ripped apart as they orbit in our Galaxy owing to tidal forces from objects such as giant molecular clouds or black holes (the source of the effect is still poorly understood). Such break-up accounts for why stars are spread all over the sky, as opposed to being concentrated in clusters, but it also means that clusters older than 0.5 billion years or so are inherently rare, and thus few are near the Sun. Compounding the difficulty, the faint stars in these more distant clusters have fewer and smaller star spots. These are regions of lower temperature than the surrounding surface that cause the star to dim and brighten when they rotate in and out of view, and so can be used to measure the star's spin. The few small spots on older stars yield variations in starlight that are hard to detect using ground-based telescopes.
That is why the capabilities of a mission such as NASA's Kepler satellite, with its exquisite measurements of stellar brightnesses, is essential, and so why Meibom and colleagues' work, which is based on Kepler data, matters. The authors used the satellite to measure the rotation period of 30 cool stars in the star cluster NGC 6819 (Fig. 1), which is about 2.5 billion years old. NGC 6819 therefore fills the large gap in age between the Sun and existing cluster observations. By using methods such as the study of rotation, Kepler has revolutionized stellar physics as much as it has the study of exoplanets. It means that we can consider using the slow, steady spin-down of Sun-like stars as a way of determining their ages.
But this method of estimating a star's age has limitations. The main one is that we do not understand the physics of rotation and angular-momentum loss in Sun-like stars, and the rotation–age relation remains purely empirical. This is fine if we can calibrate that relation well and if there is a tight correspondence between rotation and age. But stars can acquire extra angular momentum late in their lives by swallowing a companion, such as another star or a planet. Orbiting objects have much more angular momentum than has a star, and so even small bodies can be significant. We know of no means by which a star can have its angular momentum stolen, and so these accumulations add a systematic uncertainty.
Also, as noted, older stars have at best only weak variations in their light. With even the highest-precision measurement, we cannot always see the signal of the Sun's rotation. Likewise, not all stars reveal their rotation to us even when we badly want them to; nature is indifferent to our curiosity. But persistence can pry out those secrets. Meibom and colleagues' study shows exactly why we develop new capabilities: there is the good reason (in the case of Kepler, it was to find Earth-like planets, a very good reason), and then there is the real reason, which is to enable clever people to do what was not foreseen at the start.
About this article
Astronomy & Astrophysics (2016)