Dark matter on the mind

Bookmark and Share

There are a few interesting experimental results and analyses from the physics world this week, mostly having to do with dark matter. Probably the biggest of these is a fairly detailed paper on the local density of dark matter by the team of Moni Bidin, Carraro, Méndez, and Smith. As you may know, dark matter is astrophysicists’ favorite method to explain how the tangential velocity of stars in large galaxies can be nearly constant all the way from the center out to the (visible) edge, despite the fact that a simple model would tell you that the velocity should be slower for stars further out. It explains a bunch of other observations too, including measurements of gravitational lensing by large galaxy clusters, so we’re pretty confident that dark matter exists.

Dark matter density plotWith that in mind, it’s kind of surprising that the analysis done by Moni Bidin, Carraro, Méndez, and Smith finds no dark matter at all within a few kiloparsecs of the solar system! Basically, what they’ve done is apply Newtonian gravity (which applies fairly well on these scales), along with ten reasonable-sounding assumptions, to find a formula which relates the velocities of stars in some region of the galaxy to the local density of matter (regular and dark) within that region. They then took measurements of the velocities of 400 red giants in the vicinity of Earth, extrapolated to the entire stellar population using the known statistics of stellar motion, plugged the velocities into the formula, and came out with a density of \(\Sigma(\SI{1.5}{kpc})= (55.6\pm 4.7)\mathrm{M}_\odot\,\si{pc^{-2}}\), which exactly matches the density of visible stars — no dark matter needed. This is shown in the plot at the right: the dark matter density calculated from the formula is the solid black line, and the gray lines are various theoretical predictions. The line labeled “VIS” means “visible matter only.”

Of course, there are a number of ways in which this model could be inaccurate; for example, maybe the velocities of the red giants don’t reflect that of the overall stellar population as well as we think, or perhaps the measurements of the dimensions of the galactic disc are a little bit off, or perhaps one or more of the ten assumptions isn’t quite right. That’s why the entire second half of the paper is devoted to an analysis of how the result for \(\Sigma(Z)\) would change if a measurement is incorrect or an assumption is wrong. And the conclusion from that part is that, within the constraints that we definitely know from other measurements, there’s pretty much no combination of changed parameters or invalid assumptions that would make the result match any of the common models of dark matter. The only way people are seeing to make sense of the data is to assume that the dark matter is somehow clumped in particular regions of the galaxy, and we just happen to be in the middle of a pretty big dark-matter-free zone. That doesn’t seem very likely, but it is possible. It’s probably about as likely, though, that there’s some new dark matter model nobody’s thought up yet which will make more sense out of all this.

Another result that’s getting a fair amount of attention is a possible actual detection of dark matter, discovered by Christoph Weniger in data collected by the Fermi Large Area Telescope. The FLAT has been pointed at the sky to collect gamma rays for almost four years now. There are many different sources for these gamma rays, but one possible source is the annihilation of dark matter particles with each other, which could happen if dark matter particles and their antiparticles both exist in large amounts in the same region (or if dark matter particles are their own antiparticles, as is predicted by several models). Now, if you assume that the dark matter particles are moving slowly relative to each other, then if two of them annihilate into a pair of photons, each of those photons will have the same energy as the mass of the original dark matter particle. And in fact, the standard model of cosmology, the Λ-CDM model, does specify that dark matter particles should be moving slowly. So if dark matter particles and antiparticles annihilate to produce photons, we should be able to detect a bunch of photons all at roughly the same energy, which will in turn tell us the mass of whatever particle constitutes the dark matter. This is kind of similar to what happens in particle accelerators: if the particles being collided have just enough energy to make, say, a heavy quark-antiquark pair, then that pair might decay into two photons, each of which has the same energy as the mass of one of the quarks.

Potential annihilation peakThis is just the sort of phenomenon that Weniger has discovered (or at least that he’s claimed to have discovered; from reading the paper, it’s a little hard to see an effect as strong as what he reports). In gamma rays detected from a region near the center of the galaxy, there is a little bump in the photon spectrum around \(\SI{130}{GeV}\). This suggests that there may be dark matter particles with that mass annihilating in this region. It makes sense that the dark matter would cluster near the galactic center, since it responds to gravity more than to any other force. But the bump is still very small, and as Weniger himself points out, it can only be considered a tentative discovery at this time — not even a discovery, really, more like an observation. Still, this is exciting because, if the result turns out to be true, it would represent the first definitive, direct evidence that dark matter interacts in any way other than by gravity, and the first indication of what sort of particle might be making up this matter.

Yet another interesting result comes courtesy of Sean Carroll at Cosmic Variance, specifically in reference to this paper which reanalyzes data from CDMS. This paper is pretty technically dense, so I haven’t been able to properly read skim it, but between Sean’s blog post and what I can pick up from the paper, the claim is that the original analysis done by the CDMS collaboration is not sensitive enough to pick up the signal that would be generated by the dark matter at the density predicted by the common models. Specifically, the original analysis excludes a signal greater than \(\SI{0.06}{events/keVnr kg day}\) (that’s detected events per day, per kilogram of detector material, per unit energy bin width), but Collar and Fields say that the signal from WIMP dark matter should be \(\SI{0.035}{events/keVnr kg day}\). What’s more, they run a different analysis and find that the CDMS data actually do show some events which can’t be explained by the known (regular matter) interactions, but which do seem to match a signal found by their “competitor” CoGeNT. If this analysis is correct, it lends support to the idea that there is a substantial population of dark matter particles in the vicinity of the Earth, in contrast to what Moni Bidin et al. concluded. So one or the other of these results is probably going to be wrong, although it’s likely going to be a very subtle correction. It’s definitely going to be a very interesting time for dark matter detection research over the next few years.

For more information on these results, I’ll point you to blog posts by Sean Carroll and Matt Strassler.