Questioning Assumptions: Have Binary Stars Been Tricking us into Overestimating the Age of Clusters?

For decades, astronomers have puzzled over the age of globular clusters, heavenly objects made up of hundreds of thousands of stars, living and dying together as they travel through their galaxies. They tend to shine red, indicating that their stars are ancient; in fact, their accepted age is somewhere between 10 and 14 billion years. This is only slightly younger than the Universe itself (13.7 billion years)—which begs the question, how could such complex objects form so soon after the Big Bang? Stars need time to form and drift together into clusters, and gravity works slowly at large scales.

In a binary star system, the smaller of the pair can siphon matter from its larger counterpart, making them both progress through their lifecycles faster.
Image Credit: University of Warwick

A team of astrophysicists think that the discrepancy lies in how we calculate the age of globular clusters. One of the most common methods used to calculate a cluster’s age is to examine its properties (luminosity, color, and composition, which all change as a star evolves over the course of its lifetime) and compare them against the result of a computer simulation. By varying initial conditions on the simulation, researchers can glean an understanding of how the cluster was formed and how long ago.

However, as Dr. Elizabeth Stanway of the University of Warwick points out, this method is only as good as the stellar evolution model you use. “The ages and properties derived for stellar populations are very dependent on the models adopted to interpret them,” she explains. And to date, virtually all studies of globular clusters have worked under the assumption that they are composed of solitary stars.

But as it turns out, this assumption may not be entirely valid. Binary star systems—where two stars orbit a common center of mass and interact throughout their evolution—seem to be a lot more common than formerly believed. According to one study published in 2017, nearly all massive stars have at least one companion. It even appears that up to 40% of low-mass stars begin as binaries.

What does this mean for our models of globular clusters? Well, it turns out that binary star systems evolve quite differently from single stars. As the stars age, the smaller companion can strip material from the larger giant, effectively causing them both to age faster than they would on their own, and potentially fooling researching into overestimating their ages.

Enter Stanway and Dr. JJ Eldridge of the University of Auckland. The two had previously worked together to develop a model that mimics the evolution of binary stars, called Binary Population and Stellar Synthesis, or BPASS for short. After fine-tuning it for younger star systems, they set their sights on aging stars, developing a second version that increased the accuracy for older populations. Now all they needed to do was try it out, and globular clusters were a natural choice for this study because all the stars were known to be billions of years old.

First, they used the light from simulated globular clusters to estimate which types of stars populated it and in what quantities; the colors, chemical compositions, and masses of each star all had to add up to that of the entire cluster. As you can imagine, this population synthesis is incredibly resource-intensive, often taking up to several days on a supercomputer cluster.

Once they’d determined the stellar populations of each cluster, the team used BPASS v2.2 to determine their ages. To represent a variety of initial conditions, they ran the simulation for a representative sampling of the whole space of possibilities. To their surprise, they found that when they incorporated the effect of binary stars, it seems that globular clusters might have formed as recently as 5-8 billion years ago—a whopping 4 billion years younger than the commonly accepted age!

Of course, these findings are the result of a single analysis which needs to be verified and replicated. But if they are valid, it could greatly affect our understanding of star formation and galaxy evolution. Eldridge comments, “It may change for example how quickly we think the amount of heavy elements (i.e. elements other than hydrogen and helium) built up in the Universe.” It could also change our timeline for galaxy formation.

Stanway is quick to point out that they aren’t claiming that our entire galaxy evolution paradigm is completely wrong. Instead, it “relieves some of the tension” in our pieced-together history of the Universe. She sums up, “the main consequence of this result is that we don’t require so many of the stars in today’s massive galaxies to form very early in the evolution of the Universe, but that they can form in a more leisurely manner over cosmic time.”

Stanway, pictured here next to the inner workings of a sophisticated observatory—this kind of astronomy can’t be done by squinting through an eyepiece.
Image Credit: E. Stanway

However, the biggest takeaway for astronomers is that it’s always worth taking a second look at the assumptions they’re making. In Stanway’s words, “The very fact that scientifically plausible models, constructed using our best current understanding of binary star populations and their evolution, can give such different values for these parameters suggests that this is an area that needs to be studied in more detail, and where some assumptions may need to be questioned.”

It’s worth noting that an important next step will be checking the population synthesis (the calculation of stellar populations based on the overall light emitted). In the short term, Stanway and Eldridge plan to examine the age and mass ranges of nearby clusters, where they will be able to resolve the individual stars well enough to check their work. Another important sanity check will come with the arrival of comprehensive surveys of “astronomical transients”: comparatively brief stellar flare-ups like supernovae, gravitational wave chirps, and gamma ray bursts. Using tools like the Large Synoptic Survey Telescope (LSST), astronomers will be able to compare the actual number of transient events associated mostly with binary stars to the number predicted by the population models Stanway and Eldridge have developed.

Whether or not future studies confirm Stanway and Eldridge’s findings, two points are clear. The first is that physicists must always question the assumptions their models are based on. The second? You never know what will pop out of a research project. As Eldridge remarks incredulously, “We were only really trying to verify and test our models, not find such a big difference in cluster ages!”

—Eleanor Hook

You may also read these articles