Strange Dark Galaxy Puzzles Astrophysicists

The surprising discovery of a massive, Milky Way-size galaxy that is made of 99.99 percent dark matter has astronomers dreaming up new ideas about how galaxies form.










Among the thousand-plus galaxies in the Coma cluster, a massive clump of matter some 300 million light-years away, is at least one — and maybe a few hundred — that shouldn’t exist.

Dragonfly 44 is a dim galaxy, with one star for every hundred in our Milky Way. But it spans roughly as much space as the Milky Way. In addition, it’s heavy enough to rival our own galaxy in mass, according to results published in the Astrophysical Journal Letters at the end of August. That odd combination is crucial: Dragonfly 44 is so dark, so fluffy, and so heavy that some astronomers believe it will either force a revision of our theories of galaxy formation or help us understand the properties of dark matter, the mysterious stuff that interacts with normal matter via gravity and not much else. Or both.

The discovery came almost by accident. The astronomers Pieter van Dokkum of Yale University and Roberto Abraham of the University of Toronto were interested in testing theories of how galaxies form by searching for objects that have been invisible to even the most advanced telescopes: faint, wispy and extended objects in the sky. So their team built the Dragonfly Telephoto Array, a collection of modified Canon lenses that focus light onto commercial camera sensors. This setup cut down on any scattered light inside the system that might hide a dim object.

The plan was to study the faint fringes of nearby galaxies. But the famous Coma cluster — the collection of galaxies that long ago inspired astronomer Fritz Zwicky’s conjecture that such a thing as dark matter might exist — beckoned. “Partway through, we just could not resist looking at Coma,” Abraham said. “You could argue that this discovery emerged from a lack of discipline.” They planned to study the Coma cluster’s intracluster light — the faint glow of loose stars floating between the cluster’s galaxies.

Instead, they found 47 faint smudges that wouldn’t go away. These smudges seemed to have diameters roughly the same size as the Milky Way. Yet according to the commonly accepted models of galaxy formation, anything that big shouldn’t be so dim.

In these theories, clumps of dark matter seed the universe with light. First, clouds of dark matter coalesce into relatively dense dark-matter haloes. Then gas and fragments of other galaxies, drawn by the halo’s gravity, collect at the center. They spin out into a disk and collapse into luminous stars to form something we can see through telescopes. The whole process seems to be reasonably predictable for big galaxies such as our Milky Way. Having measured either a galaxy’s dark-matter halo or its assortment of stars, you should be able to predict the other to within a factor of two.

“It’s not just dogma. It’s basically that there are no exceptions that we knew of,” said Jeremiah Ostriker, an astrophysicist at Columbia University.

After Abraham and van Dokkum realized that they appeared to be looking at 47 exceptions, they did a search through the literature. They found that similar fuzzy blobs have been on the edge of discovery since the 1970s. Van Dokkum thinks astronomy’s transition from photographic plates — which were perhaps better suited to picking up extended, diffuse objects — to modern digital sensors may actually have hid them from further attention.

Abraham and van Dokkum first noticed their smudges in the spring of 2014. Since then, similar “ultra-diffuse galaxies,” or UDGs, have been discovered in other galaxy groupings like the Virgo and Fornax clusters. And in the Coma cluster, one study suggested, there may be a thousand more of them, including 332 that are about as large as the Milky Way.

Meanwhile, the Dragonfly team has been advancing the case that these new dim galaxies really are oddballs that challenge current theory. They’re failed galaxies, this argument holds. Dark matter planted the seeds of a spiral disk and stars, but somehow the luminous structure didn’t sprout.

That argument has convinced outside experts like Ostriker, who finds van Dokkum’s prior record highly credible. “There are many, many other people who could have ‘discovered’ this where I’d be much more skeptical,” Ostriker said. “The simplest way of putting it is: His papers aren’t wrong.”

Not everyone is so convinced. While these UDGs may be large, they’re not necessarily massive, argue some astronomers. One idea is that UDGs might be lightweight galaxies that look puffy because they are in the process of being torn apart by gravitational tides from the rest of the Coma cluster.

Michelle Collins, an astronomer at the University of Surrey, argues that “the only other place we’ve seen things that are that extreme or more extreme are a handful of galaxies around the Local Group,” referring to small, dim “dwarf galaxies” that frequently orbit larger galaxies such as our Milky Way. “They are all things that are currently being ripped apart.” That would make most UDGs just large dwarf galaxies in the process of being ripped to shreds.

Another possibility hinges on the idea that galaxies can “breathe.” At the end of 2015, Kareem El-Badry, who was at the time an undergraduate student at Yale University, proposed that galaxies can swell out and then collapse in size by over a factor of two. In this process, gas first falls into the galaxy, forming massive stars — the breathing in. The stars quickly end their lives in supernova explosions that blast the gas outside the galaxy — the breathing out. The gas eventually cools, and gravity pulls it back toward the galactic center. In a lone galaxy, this rhythm can continue indefinitely. But in the harsh environment of the Coma cluster, where hot gas fills the space between galaxies, the gas after the galaxy exhales could be stripped away, leaving the whole galaxy stuck in a puffy state.

Yet another interpretation, suggested in March 2016 by Harvard University astrophysicists Nicola Amorisco and Avi Loeb, is that UDGs are ordinary galaxies that are just spinning fast. “In our scenario, it’s very natural,” Loeb said.

That idea piggybacks on standard theories of galaxy formation, in which gas pours into a dark-matter halo to build a galaxy. As the material falls, it begins to rotate. The amount of rotation determines the size of the final galaxy. Without much spin, gravity pulls the galaxy into a compact shape. But galaxies that get a big rotational push can spin themselves out into large, lightweight disks.

It could be, according to this model, that the UDGs are natural examples of the very fastest spinners. If so, their stretched-out disks wouldn’t be dense enough to form as many stars as a slower rotator like the Milky Way, explaining why they look so faint.

These ideas may well explain some of the UDG population, according to Abraham. “Probably this is going to evolve into a mixed bag of things,” he said. But according to his team’s latest data, obtained from observations that spanned a total of 33.5 hours on the 10-meter Keck II telescope in Hawaii, there is no evidence that the Dragonfly 44 galaxy is rotating. In addition, they argue that the total mass of the galaxy is around a trillion suns — massive enough to prevent it being ripped apart like a dwarf galaxy, and heavier than the galaxies thought to periodically puff up.

That mass measurement is the real sticking point, said Philip Hopkins, a theoretical astrophysicist at the California Institute of Technology who is preparing several papers on UDGs. It comes from two observations of different parts of Dragonfly 44. First, the motions of stars in the galaxy’s inner regions suggest that the area is massive, filled with dark matter. Second, the outskirts of the galaxy are home to a number of globular clusters — tight, ancient balls of stars. Just as the number of stars in a galaxy is ordinarily linked to the amount of dark matter, observations show that the more globular clusters a galaxy has, the higher the mass of its dark-matter halo. Dragonfly 44 has Milky Way-level clusters. Other UDGs seem to have lots of globular clusters, too.

Because of this, even if these UDGs don’t have heavy dark-matter haloes, researchers will still be left to explain why they have far more globular clusters than the known relationship suggests they should. Something is weird about these things,” Hopkins said. “Either way, it’s really cool.”

The discovery has generated enough interest to earn the team precious time on the Hubble Space Telescope to study Dragonfly 44’s globular clusters. “The thing I find hilarious is we’re using humanity’s most powerful telescope in space to follow up a bunch of telephoto lenses,” Abraham said. To fully understand the relationship between dark matter and the globular clusters, though, they have to measure the motions of the clusters — for which they’ll need to wait until the James Webb Space Telescope launches in 2018.

In parallel, they’re looking to find and characterize more Dragonfly 44s, preferably a few located both outside of a cluster — and thus free of the harsh cluster environment — and closer to us. It’s an open question as to whether they exist elsewhere and, if so, what form they take. “The resolution of whether the UDGs are what we argue they are, or something else, would come from finding them outside of clusters of galaxies and seeing how they look there,” Loeb said. A few candidates have emerged, van Dokkum said, and they are now being followed up with Keck and Hubble.

For theorists like Ostriker, that’s an exciting prospect. If the motion of stars in a galaxy like Dragonfly 44 can be studied up close, it would be a make-or-break test for current dark-matter theories, which make different predictions about how the missing mass should be distributed. The leading theory, called cold dark matter, suggests dark matter should surge at the heart of a galaxy. Right now, though, the dark-matter-dominated galaxies we have to study are nearby dwarf galaxies, and they don’t exhibit that characteristic. “Many of the properties that dark matter is supposed to have … these little galaxies don’t show,” Ostriker said. “But we say, ‘We don’t really know how these things were formed anyway,’ and we just change the subject.”

By contrast, an otherwise normal-but-dark Milky Way would eliminate that loophole. In the universe’s other Milky Way-size galaxies, stars and gas can outweigh dark matter in the central regions by a factor of five to one. That makes disentangling the gravitational pull of dark matter alone tricky. But the center of Dragonfly 44’s disk is 98 percent dark matter, meaning a map of its central mass would give unprecedented insight into dark matter’s properties, Ostriker said.

The way forward to understand UDGs isn’t clear yet, Abraham said, but hopefully at least some of the ideas now being proposed will persist through the next few years of observations. “In astronomy, it’s still valid to be just an explorer. In the case of Dragonfly, we’re like Leif Eriksson,” he said. “You’ve been on the ship for months, and suddenly somebody said, ‘Land ho!’ And it’s not on the map.”

This article was reprinted on



Comment on this article