Do We Need a New Model For Dark Matter? (script) (Patreon)

The stuff of the visible universe - stars, planets, people - are a cosmic afterthought. Just luminous sprinkling on top of the vast oceans of dark matter that dominate the gravitational universe. Although we don't know what dark matter actualy is, for a long time we thought we at least had it's behavior nailed. The so-called cold dark matter model had to be right because it worked so well in explaining structure in the universe--until it didn't. Distressing disagreements with newer observations sent many physicists back to the blackboard. But more recently, a new generation of sophisticated, high-power computer simulations may save our favorite dark matter model. With some unexpected help - it turns out that visible matter may be much more important in shaping the universe than we ever imagined.

INTRO

We have no idea what dark matter is, other than it’s some source of gravity that is completely invisible but exerts way more pull that all of the regular matter. More than all of the stars, all of the gas, all of the black holes…unless dark matter is black holes, then black holes are most of everything. Dark matter constitutes 80% or so of the mass in the universe, which means even our Milky Way galaxy is mostly a vast ball of dark matter that happens to have attracted a relative sprinkling of baryons—atoms in the form of gas, which lit up as starry glitter spinning in the middle of this invisible gravitational well.

But for all of dark matter’s mysteriousness, it’s remarkably simple stuff in terms of its behavior. Or at least so we have long thought. The mainstream model for dark matter is called cold dark matter, or CDM. This type of dark matter is described as a fluid of particles that don’t interact with each other or anything else except by gravity, and that have pretty low speeds, making them cold. If a universe starts out full of such a fluid, any tiny lumps in density of that fluid tend to attract more dark matter and grow. We do these incredible supercomputer simulations of this process, and based on the CDM model, this type of dark matter seems to lead to exactly the types of giant structures—galaxies, galaxy clusters, etc—that we see in the universe today.

Although we don’t know what this “CDM” might be made of, there are some seemingly very plausible types of particle that could do the job. Broadly, we have weakly interacting massive particles or WIMPs, which are predicted by various extensions to the standard model of particle physics, or another exotic particle, the axion, or even the frozen Planck-scale relics of evaporated black holes all behave like cold dark matter, and we’ve talked about all of these before.

The fact that there are many fairly natural ways to make cold dark matter is a point in favour of the idea. But another one is the so-called WIMP miracle. Most elementary particles in our universe were created soon after the Big Bang during a short period when the universe was hot enou  gh for matter-antimatter pairs to be created spontaneously from the extremely energetic radiation of that time. That creation process stopped when the universe expanded and cooled, and most of this stuff annihilated with itself. Regular matter, which is much more able to interact with itself, almost completely annihilated so that the matter that was left is just due to the 1-in-a billion overabundance of normal matter over antimatter.

But dark matter fitting the WIMP category wouldn’t have self-annihilated so quickly due to its weakly-interacting nature. Instead, it stopped annihilating only when the universe just got too big for dark matter particles to find each other. It turns out that if you calculate how much dark matter should have been left when that annihilation stopped you get pretty much the right number for the amount of dark matter we see in the universe today. That’s the WIMP miracle. Now this does require the existence of some unknown heavy non-standard-model particle like those predicted by supersymmetry, but actually its not tied to any particular theory.

Long story short, CDM was looking great. But then, starting around 20 years ago, we began to notice that our rapidly improving simulations and observations of the universe stopped matching quite so well. Let’s start with the simulation side. Because cold dark matter is cold, it should be able to form pretty small blobs with relatively little gravity holding them together. That’s because cold means slow-moving particles, which have trouble escaping even weak gravitational fields. So if we try to simulate, say, the formation of the Milky Way with this type of dark matter we get something weird. We get a big dark matter halo that attracts the gas that forms the stars of the main galaxy—that part is fine. But within this halo these CDM simulations also predict a lot of sub-haloes - smaller dents in the big gravitational field that themselves should presumably attract gas and form sort of mini galaxies - what we call satellite galaxies. CDM simulations predict that the Milky Way and really any big galaxy should be orbited by thousands of these things. But our observations of real galaxies don’t find anywhere near enough of them. This is the “missing satellites” problem, and it was a clue that our CDM model might not be as miraculous as we thought.

Another prediction of CDM is about the shape of big galaxies. The weakly interacting CDM particles don’t push each other apart. Add the fact that they’re moving slowly, and it becomes possible for these particles to form very dense, tight clusters. CDM models predict that inside galaxies, the density of dark matter should go up and up as you travel from the outskirts inwards, and should reach a very high density cusp near the very center. And regular matter should fall towards this dark matter concentration, resulting in lots of stars near the galactic center. But in real galaxies, the density usually flattens out within several thousand light years of the center, leaving a more rounded core. This is known as the “cusp-core” problem, and it’s the second strike against cold dark matter.

Image 2      Image 2.5

Of course, physics thrives on discrepancy, and these problems have inspired many years of work on alternatives to the standard CDM model.

The most obvious fix to cold dark matter is to change the temperature—change the “cold” part. We can’t heat up dark matter too much or its particles won’t form halos at all—the particles will just wizz randomly through the universe and be pretty useless. But what if we increase temperature just a bit to give us warm dark matter, or WDM? Sterile neutrinos could give us WDM, and we’ve discussed previously. However their abundance doesn’t “miraculously” come out just right from the particle physics and the cosmology in the way that WIMPs seem to.

If dark matter is made of relatively light particles that have a higher “temperature”, they can still form big haloes but have a harder time forming subhaloes. That’s simply because small haloes don’t have strong enough gravitational fields to hold these faster moving particles. So maybe that's why the Milky Way has very few satellite galaxies.

Another popular solution to this cusp-core problem is self-interacting dark matter (SIDM). It proposes that dark matter particles actually do interact with themselves a bit more than standard CDM would assume. That means they can repel and scatter off each other, which stops them becoming packed too tightly above a certain density.

Finally, there’s also fuzzy dark matter (FDM), which we talked about in our recent episode. In this model, dark matter is a superfluid of ultra-light axions with a de Broglie wavelength thousands of light years long. It can’t form structures smaller than this wavelength, so may neatly eliminate tiny satellites and very concentrated galactic cores, perhaps solving both the “missing satellites” and “cusp-core” problems.

OK, so we have a few reasonable solutions to our dark matter woes. But which is right? Well, not so fast. While all of this new physics was being concocted, there was a parallel effort to see if the problems with cold dark matter could be fixed while keeping dark matter cold. Remember, CDM is appealing as a model for more reasons than its behavior as dark matter. Many believe that it is the most natural type of dark matter to have formed in the early universe. That theoretical convenience isn’t enough for us to accept CDM if it doesn’t explain our observations. But it turns out that despite hints to the contrary, CDM may still be our best bet.

The CDM simulations I’ve been talking about so far are “dark-matter-only”, which means that they don’t include ordinary matter; they’re just simulating how dark matter behaves in the absence of gas and stars and all that visible stuff. That sounds a bit … forgetful on the part of the simulators. But remember that there’s more than 5 times more dark matter than ordinary matter in the universe. Almost all the gravitational oomph in the universe is from dark matter, so it’s actually not crazy for these simulations to ignore the relatively piddling contribution from atoms. But actually, we’re now discovering that there are a couple of reasons why ordinary matter really does matter.

First, ordinary matter is the stuff we actually see, so when you’re trying to use a simulation to predict what it is we should see, it’s helpful to have that stuff in your simulation; otherwise, how do you predict what you’re going to see through your telescope? For example, we guessed that all of the dark matter sub-halos around a system like the Milky Way should contain visible satellite galaxies. But that turns out to be wrong. Sub-halos that are less than around a billion times the mass of the sun are simply incapable of capturing gas from the intergalactic medium, because the gas is too hot__its particles moving too quickly—to become trapped in the shallow gravitational wells of these low-mass sub-halos. So most of these sub-halos have few to no stars, and therefore we can’t actually see them. If this is right then the satellites aren’t missing, they’re just invisible.

Also, our telescopes are becoming more and more sensitive and we’re starting to see a lot of very faint galaxies and substructures, so we’re getting closer to the predictions of the CDM models.

And now the second reason that regular matter matters. Although so-called “baryonic matter” represents only a small fraction of the mass in the universe, this stuff is way more interesting than dark matter because it can form stars, planets and people. It can also influence dark matter in unexpected ways. Because it can create stars, it can also produce supernova explosions when the most massive of those stars die. Every time a supernova goes off, it blows away a huge chunk of the gas surrounding it. That gas can be forced out of the galactic core, or even out of the galaxy entirely. And this outgoing gas also drags a little bit of dark matter with it. The dense galactic core of a young galaxy will experience burst after burst of star formation, which lead to multiple waves of supernovae. Over time, quite a bit of dark matter can get dragged out of the core of the galaxy. This can ultimately flatten out what started as the very high density core, eliminating the cusp predicted by the CDM model.

But CDM isn’t totally out of the woods yet. In recent years, we’ve found some counter examples that seem to defy these needs fixes for CDM’s problems. For example, there are several galaxies that have the typical low density cores but don’t seem to have had enough star formation in their pasts to do that supernova-driven smearing. And strangely, also some systems with lots of star formation that actually have high-density cusps. Like for some reason the dark matter ignored the supernovae in these cases. So what started as a “cusp-core” problem has actually evolved into what’s being called a “density-diversity” problem. There’s a very  wide range of central densities in galaxies that doesn’t appear to be correlated with star formation or the mass of the surrounding dark matter halo. Currently it’s hard to see how the CDM model can accommodate this kind of diversity.

The missing satellites solution also has a new problem. While it’s true that small satellites are going to be invisible due to not holding onto enough gas, those CDM simulations do predict a reasonable number of larger sub-haloes within a galaxy, with masses more than a billion times that of the sun. These things should have no problem holding onto enough gas to make stars, so we should be able to see them. But we don’t really — at least not in the numbers predicted by CDM simulations. So the “missing satellites” problem has evolved into what’s known as “too-big-to-fail” problem: we expect to not see the smaller subhaloes, but we still seem to be missing some of the bigger CDM halos—the ones that are too massive to have failed to form galaxies and stars.

Both the “density-diversity” and “too-big-to-fail” problems are challenges to the CDM model because they’re difficult to solve by including regular matter. But they aren’t quite the hypothesis-killers for CDM that the cusp-core and missing satellite problems were once thought to be. It may be that improved simulations and observations will solve these too.

As the prospects of CDM recovered, alternative models like warm or fuzzy dark matter are coming under strain. Our improving telescope sensitivity is finding more and more small-scale structures that conflict with the predictions of  WDM and FDM, which are all about stopping small-scale structure. There are ways to tweak these models to better match our observations, but that kind of defeats their original purpose, so they;re perhaps not as parsimonious as they once were.

The bottom line is that there is still a lot of uncertainty and debate about both the microscopic constituents and the macroscopic behavior of dark matter. But one thing is clear: ordinary matter is not nearly as insignificant to cosmology as we once thought. In order for us to finally solve one of the greatest mysteries in all physics—in order to figure out the nature of dark matter, we’ll need to continue to improve our understanding of the complex ways in which all types of matter shape the structures in Space Time.