Credit: Illustration: NASA, ESA, and JPL-Caltech; Science: NASA, ESA, and D. Apai and E. Buenzli (University of Arizona)
Study of Substellar Objects Using Bridges-2 Sheds Light on Globular Cluster Age and Early Development of Universe
Collections of stars called globular clusters are some of the oldest known structures in our Galaxy, and their properties preserve the record of conditions in the early Universe. But estimating their exact ages is very challenging, and different dating techniques often disagree with each other. An international team used data from the new James Webb Space Telescope (JWST) and simulation on supercomputers including PSC’s Bridges-2 to identify and date for the first time three brown dwarfs in a globular cluster. These objects, too small to be stars and too large to be planets, cool off over time, providing a new way to estimate the age of the parent cluster.
WHY IT’S IMPORTANT
Scientists studying the origins of the Universe must glean clues about how stars and, well, everything formed, using weak signals from faraway astronomical structures. Globular clusters are an important source of this information. These are collections of stars that look a little like large spheres (“globes”) of light through a pair of binoculars. Most of them are almost as old as the Universe itself and may offer clues as to the formation and evolution of the earliest stars and galaxies.
The challenge is that globular clusters are all extremely far away, occupying the outlying regions of our Milky Way Galaxy. Their member stars are very faint from Earth and can only be studied with very large telescopes. Brown dwarfs within these clusters are even fainter. They couldn’t be observed at all prior to the launch of the JWST.
“I am trying to understand the process that took the universe from the primordial soup of hydrogen and helium to this incredibly chemically diverse environment that can sustain complicated objects like astronomers. In order to do that, I look at very old stars because their chemical composition would capture the very first steps of this process.”
— Roman Gerasimov, University of Notre Dame
Roman Gerasimov, now a postdoctoral researcher at the University of Notre Dame, studied how the JWST and its sensitivity to low-energy infrared light could help identify and characterize brown dwarfs within the globular cluster NGC 6397. Then a graduate student in Professor Adam Burgasser’s group at the University of California San Diego, Gerasimov did this work as part of an international team led by Luigi Bedin at the National Institute for Astrophysics in Italy. NGC 6397, a small collection of stars, is the second-closest of about 160 globular clusters in our Galaxy. To analyze the data, the team turned to supercomputers available through the NSF’s ACCESS program, including Expanse at the San Diego Supercomputer Center and Bridges-2, PSC’s flagship system.
HOW PSC HELPED
One of the problems with gauging how globular clusters develop and evolve is that methods for determining their age, all based on the evolution of their stars, give different answers. Dating brown dwarfs by their brightness promised to be a critical tiebreaker, because unlike the other methods, it doesn’t depend on how a star evolves. Brown dwarfs’ brightness reflects only how long they’ve been cooling down, and so is an independent measure of age.
Before the team could identify and date brown dwarfs in NGC 6397, though, they’d need to know exactly how these objects dim with age. This wasn’t straightforward, as brown dwarfs have relatively cool outer layers with temperatures at which relatively complicated molecules, such as water and methane, can form. These molecules act as a sort of “blanket” over the brown dwarf, affecting its cooling rate in a very complicated way. The team would need sophisticated computer simulations to capture all of the intricacies of the associated physics and chemistry.
The creation of their new set of models, SANDee, would require testing as many models as possible all at the same time, in a computer that had good I/O — the ability to move data back and forth between the processors and the disk storage. This is a particular strength of both Bridges-2 and Expanse.
“This is primarily a high-throughput computing problem. We needed to compute a large grid of models for all possible combinations of chemical composition and temperature. Overall, we needed about a million CPU hours on Bridges-2 and Expanse to get all the models we need. The code that we used also relies heavily on disk I/O to store intermediate output, so we needed a parallel high-performance file system.”
— Roman Gerasimov, University of Notre Dame
SANDee allowed the team to identify three brown dwarfs in new JWST images of the cluster, BD1756, BD1628, and BD1388, and measure their temperatures. These objects’ ages set the age of the cluster at the older end of the earlier predictions. The new measurements further indicated that the brown dwarfs may have a distinct abundance of light elements compared to the more massive stars in the cluster, offering clues as to how the cluster had developed over time. The team reported their results in The Astrophysical Journal in August, 2024. Gerasimov’s related work on the chemical composition of ancient stars received an International Astronomical Union Stars and Stellar Physics Division Award for outstanding scientific achievement of astronomy.
Three brown dwarfs won’t be enough to get really good statistics on the age and development of NGC 6397. In particular, the uncertainty was large, and so the results aren’t yet competitive with other, more established dating methods. But the report served as a proof of concept for SANDee, which the team has made available to the scientific community. They plan to use their tool for further identification and dating of brown dwarfs, hundreds of which should be detectable in future JWST imaging campaigns.