Galaxies are some of the largest clearly defined structures in space. There are trillions of them, and many are clustered around each other. But how does that clustering affect them? That’s been a question for a while, and older papers have yielded contradictory results. Now, a new paper analyzing millions of galaxies from researchers at the University of Washington, Yale, and several other institutions shows a clear pattern that had been debated before – galaxies surrounded by other galaxies tend to be larger.
The path to that conclusion was a long one. Several other surveys showed that galaxies in “dense environments” were both larger and smaller. However, these studies were only conducted using a relatively limited dataset of hundreds or thousands of galaxies. So, the researchers in the new paper, led by Aritra Ghosh, a postdoc at UW, thought, “Why not get more data?”
So they did, using the Subaru telescope’s Hyper Suprime-Cam Subaru Strategic Program. This survey captured high-quality data on millions of galaxies for the first time, so the researchers were able to select 3 million galaxies with the best datasets from the cream of the crop.
Massive datasets on galaxies are becoming more common – Fraser discusses another survey of millions of galaxies.They then drew “circles” of about 30 million light-years around each of the three million galaxies and assessed how densely packed their local neighborhood was. Statistically, the correlation was obvious – galaxies in more dense spatial neighborhoods were larger than their more isolated cousins.
Saying the researchers drew circles around 30 million galaxies isn’t accurate, though – they used one of the myriad new machine-learning tools popping up in the astronomical community. This one, called the Galaxy Morphology Posterior Estimation Network, or GaMPEN, was the focal point of Dr. Ghosh’s PhD thesis at Yale. It specializes in estimating galaxy size and accounting for uncertainties in the measurement.
With the tool’s results, the question became, what does this mean? The idea that galaxies are larger in dense areas doesn’t fit well with astronomers’ current conception of how galaxies form. So, it’s time for a new theory to fit the data Subaru has collected, and the paper has analyzed.
Fraser discusses globular clusters, one of the dense states galaxies can find themselves in.Several theories put forward in a press release could explain the observations. One is that densely clustered galaxies are simply larger from the start. Another is that perhaps they are more effective at merging with closely proximate galaxies to create larger supergalaxies than the two originals.
A third, more intriguing possibility is that dark matter might be involved. But since scientists still don’t understand what dark matter actually is, this is akin to waving a magic wand to explain data that otherwise doesn’t fit the cosmological model.
Regardless of the reason, the study is an excellent example of how large datasets and AI-enabled tools will change astronomy shortly. In some cases, it will confirm existing theories, and in some cases, like the relationship between galaxy density and size, it will call for a new theoretical framework. Either way, it’s exciting to be around for all these new discoveries, whether AI-powered or not.
Learn More:
UW – Galaxies in dense environments tend to be larger, settling one cosmic question and raising others
Ghosh et al. – Denser Environments Cultivate Larger Galaxies: A Comprehensive Study beyond the Local Universe with 3 Million Hyper Suprime-Cam Galaxies
UT – This Distant Galaxy Cluster is Totally Relaxed, Unharassed for a Billion Years|
UT – A Collision Between Gigantic Galaxy Clusters. Too Big, Too Early
Lead Image:
Image of Abell 2218, a dense galactic cluster approximately 2 billion light-years from Earth.
Credit – NASA/ESA/Johan Richard
