Home » Science » Dark matter vs. dwarfs: Issues for our best model of the Universe | Ars Technica

Dark matter vs. dwarfs: Issues for our best model of the Universe | Ars Technica





Dark matter has had incredible explanatory power. After the introduction of the concept in the early 1980s, dark matter quickly became a central feature of our cosmological picture of the Universe. The current leading dark matter model is called Lambda Cold Dark Matter, or ΛCDM, and its predictions have consistently been borne out. A host of observational evidence confirms that our galaxy, as well as every other, sits within a halo (basically a spherical blob) of invisible, weakly or non-interacting matter, with a mass that far outweighs all the stars that sit within it.

There has been some skepticism, of course. Some have argued that there may be no dark matter and that instead, our understanding of gravity doesn’t apply on galactic scales, which could eliminate the discrepancies that we explain with dark matter. But as time goes on, the evidence that comes in continues to support dark matter. Those alternative models are still possible, but no convincing evidence has come to light in support of them.

But that doesn’t mean the ΛCDM model is perfect. Two problems have arisen in the form of predictions that don’t match certain observations. It may be the case that these issues can be solved by tweaking the model or by taking into account other processes that may be taking place involving normal, “baryonic” matter. A group of researchers has written a piece for the journal PNAS summarizing the various possible resolutions of the issues.

Cusps, cores, and satellites (oh, my)

The first problem came to be known as the “cusp-core problem.” It is predicted that dark matter should accumulate close to the core of its galaxy and that the high abundance of dark matter toward the core should detectably increase the speeds of visible objects like stars near the centers of galaxies. This is not what is observed, however. ΛCDM predicts far more dark matter in galactic cores than we’ve seen so far. The problem is detectable in typical galaxies like our own Milky Way, but it becomes much more blatant in smaller galaxies.

At first, scientists wondered if this discrepancy could indicate that the problem was essentially an optical illusion—an artifact of objects in the smaller galaxies not following circular motions, leading us to misinterpret their speeds. But as better and better measurements came in, it became clear that there was no such simple solution in sight; the problem really exists.

And then there’s the “missing satellite problem.” Just as the Earth has a satellite in the Moon, so the Milky Way has a number of satellite galaxies, smaller bodies that orbit the Milky Way. These satellite galaxies rest in their own dark matter halos, which themselves are within the boundaries of the Milky Way’s large halo. For this reason, they’re considered “sub-halos” of the Milky Way halo.

Before the year 2000, only nine satellite galaxies were known, including the Large and Small Magellanic Clouds. But in addition to these satellite galaxies, the Milky Way also has a number of “empty” dark matter subhalos. These blobs of dark matter could have been host to satellite galaxies, but they aren’t. Therein lies the problem: according to predictions, there should be up to 20 more satellite galaxies than we’ve observed, with only the mysteriously empty subhalos making the numbers look reasonable.

There were a few proposed solutions to this. It could be the case that early in the history of the subhalos, when galaxies were beginning to form in them, background UV light may have driven the gas out, preventing the galaxies from forming. Another possibility is that stellar winds, charged particles originating from the first generation of massive stars or from supernovae, drove out the accreting gas.

Things got more complicated after the Sloan Digital Sky Survey (SDSS). About 15 new satellite galaxies were detected, faint enough that they were missed previously. That survey only covered about 20 percent of the sky; if similar dwarfs were further out within the Milky Way’s halo, they would be too dim to be detected, and there could be several hundred dwarf galaxies orbiting the Milky Way in total.

This makes for a confusing picture in terms of dwarf galaxy formation. But in light of the newer findings, the missing satellite issue no longer presents a problem for ΛCDM. “Despite the gaps in understanding,” the authors write in their paper, “it seems reasonable for now to regard the relation between low-mass subhalos and ultrafaint dwarfs as a puzzle of galaxy formation physics rather than a contradiction of CDM.”

But just when we thought it was safe to go back in the dark matter subhalo, a new form of the missing satellite problem reared its ugly head. Because when comparing the empty, or dark, subhalos to the bright ones—the ones hosting dwarf galaxies—a difference became apparent. The mass of the dark matter in the center regions of the dark subhalos exceeds the mass in the centers of bright ones by about five times.

Some researchers suggested that this could be a coincidence—the Milky Way’s satellite galaxies just happen to reside in the least massive dark matter subhalos. But this was a somewhat unsatisfying solution, and the discrepancy has been called “too big to fail” by one researcher. Even with the lowest estimate of the Milky Way halo’s mass, which would maximize the chances of the dwarf galaxies forming in the least massive subhalos by chance, the discrepancy is still too big to be a statistical fluke. Not only that, but the Andromeda galaxy has the same problem, making it even less likely that it’s a coincidence.

In its new form, the missing satellite problem begins to look a lot like the cusp-core problem. Both problems involve galactic cores having far less mass than they’re predicted to. Therefore, the authors argue that the two problems have effectively merged into one.

The question, then, is how the issue can be resolved. Is there some unknown or unaccounted for process involving baryonic matter causing the discrepancies? And if not, what does it mean for our models of dark matter—will the properties of the mysterious substance have to be changed?

Solutions in baryonic physics?

After the cusp-core problem first became apparent, it was thought that baryonic physics could offer no solution. After all, one of the properties of dark matter is that it only interacts with baryonic matter through gravity (and perhaps through the weak force, but such interactions would be few and far between and would make little difference in this context).

But if one were to place a lot of baryonic matter, like a galaxy, inside a dark matter halo, the added gravity should draw more dark matter toward the center, making the problem worse, not better. If, instead of dark matter, the galaxy was suspended in a cloud of gas particles of similar mass to the dark matter halo, it might be possible that the galaxy could help keep the center relatively clear of gas by heating it. When gas heats up, it expands—meaning it doesn’t clump in the center. But with dark matter, there’s no way to transfer heat from the baryonic matter, so that can’t happen.

But this common wisdom didn’t stop scientists from looking for ways around the problem: complex scenarios involving baryonic matter that could affect dark matter. One such solution, considered extreme at the time, involved supernova explosions. When a massive star explodes, it blows off its outer layers, which in turn can push nearby matter away. This surge of outgoing baryonic matter can have an effect on dark matter.

These supernovas are part of a feedback mechanism—the more rapid star formation that takes place, the more gas is expelled from the galaxy due to supernovas. This in turn prevents more stars from forming, and over time, drives baryonic matter out.

This feedback affects the dark matter. During the phase of the system’s evolution in which a lot of baryonic matter is still present, the added gravity draws dark matter deeper into the galaxy’s center, just as common wisdom predicts. But then, when the feedback mechanism clears out most of the baryonic matter, the dark matter finds itself suddenly released. This speeds up the dark matter particles quite a bit, widening their individual orbits and reducing the clumping in the galaxy’s core.

“Extreme” as this idea may have sounded at first, it began to bear fruit as computer simulations improved and became able to model complex situations like this one with ever-increasing accuracy. The simulations confirmed that the feedback mechanism could produce the dark matter distribution observed in galaxies. Additionally, strong gas outflows have been observed in certain galaxies, some of which likely represent an early stage in the evolution of a higher-mass galaxy with a similar cusp-core dark matter distribution.

However, for smaller-mass (below about ten million solar masses) galaxies, the supernova feedback mechanism isn’t enough, at least not on its own, to explain away the cusp-core problem. In conjunction with a few other baryonic-matter mechanisms, it gets close, but probably not close enough to solve the problem, though some researchers have pointed out that certain situations might be able to change the playing field just enough to make up the difference.

The ultimate test, of course, is to look at actual galaxies in the right mass range and see how they’re behaving. Isolated dwarfs in particular make excellent candidates for testing because they’re low enough in mass and because they’re not close enough to other galaxies that they could be stealing matter from their neighbors, so most of their baryonic matter is their own.

One study has already looked at such galaxies and concluded that the problem persists even there—the baryonic mechanisms aren’t enough to account for the dark matter not clumping. This study hasn’t yet been reproduced, but if the result remains consistent, it could indicate that this challenge to ΛCDM hasn’t gone away.

Solutions in dark matter physics?

That being the case, the solution could lie in the basic premises of the ΛCDM model. One alternative is that the dark matter might not be as cold as we think. Warm Dark Matter (WDM) is so called because its particles are moving around a lot faster than CDM but not reaching the relativistic speeds necessary to be termed Hot Dark Matter (HDM).

WDM has the advantage that it would not clump toward the center as easily because of its speed. While this doesn’t solve the problem entirely, the clumping isn’t as bad as it is with CDM. This sounds promising at first, but WDM suffers from a rather significant flaw: it conflicts with observations. For one thing, WDM predicts fewer subhalos than there are observed satellite galaxies. And even if these problems are ignored, WDM still can’t solve the cusp-core problem, at least not on its own.

Another possibility is that dark matter is cold and does weakly interact with baryonic matter but can interact with other dark matter. If the individual dark matter particles are colliding with each other on a regular basis, it could redistribute energy among the particles, leading to the core having a roughly constant density.

Early on, studies suggested that this model was ruled out by gravitational lensing and other observations. But more recent studies have not borne out these objections, instead showing that the self-interacting dark matter could produce all these observations.

Best of all, the self-interacting dark matter works in a wide variety of scenarios. It can account for the cusp-core problem in both spiral galaxies like the Milky Way and dwarf galaxies like the satellites. And unlike WDM, the self-interacting dark matter model predicts the right number of dark matter subhalos. All in all, self-interacting dark matter looks like a better candidate than WDM; the next step is to test it and either rule it out or discover some evidence for it.

Besides those two, there are also a few other possible dark matter mechanisms that could solve the cusp-core problem. These include dark matter particle decay, dark matter particles and anti-particles annihilating each other, or escape from flavor-mixed quantum states.


Despite the evidence suggesting that baryonic physics isn’t enough to account for the discrepancies, more detailed simulations may reveal a different picture. The evidence is by no means conclusive, and a lot of work targets the problem from the “baryonic solutions” angle. But the tension could turn into a serious problem for the ΛCDM model.

There are a number of ways that tension could be resolved. For one thing, the self-interacting dark matter models make predictions about galactic halo shapes, and future observations could look to see if those predictions are borne out. Some of the other models mentioned here make testable predictions as well.

But the most conclusive way to address the issue would be to further test ΛCDM itself. It makes a specific prediction about the number of low-mass subhalos which should form around galaxies like the Milky Way—about 20,000—which, if confirmed, would substantially strengthen the model.


The next generation of telescopes, especially the James Webb Space Telescope and the Atacama Large Millimeter Array, should be sensitive enough to begin looking deeper into this issue. In the meantime, we hope that with the wide variety of experiments looking to find the dark matter particle, it will be found sometime within the next decade. That could also shed some light on the issues—the nature of the particle may reveal the behavior of dark matter as a whole, such as whether it interacts with other dark matter, for example.



Dark matter vs. dwarfs: Issues for our best model of the Universe | Ars Technica.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: