Claims of the non-existence of dark matter are a staple of astro-kookery, but Physics World today has a news story with the provocative title "Galaxy survey casts doubts on cold dark matter," which makes it sound like people from reputable collaborations are questioning the existence of dark matter.
So what's the deal? Well, here's the explanation of the results, which come from a survey of 200-odd galaxies identified by both radio and visible telescopes:
Using data from both telescopes, the team classified the galaxies in terms of six independent properties. These were two optical radii (which define the sizes of the regions of a galaxy that emit 50% and 90% of the object's light); the luminosity; the mass of neutral hydrogen in the galaxy; the dynamical mass (which includes dark matter); and the colour of the galaxy. The team then carried out a statistical analysis of the data and found five correlations between these six properties -- leading them to conclude that the structure of these galaxies is controlled by just one parameter. Although the team was unable to conclude exactly what this parameter is, Disney says that it seems to have a strong relationship with the mass of the galaxies.
Disney argues that the team's finding is at odds with the hierarchical theory of galaxy formation, according to which the structure of a galaxy would be strongly influenced by the nature of the collisions that formed it. "If this were the case we would have expected to see 4-5 independent parameters", he says. And because the hierarchical theory has CDM at its root, Disney believes that the team's survey provides strong evidence that CDM does not exist. "Maybe our observations could be explained by CDM, but I wouldn't bet on it", he said.
That sounds a whole lot squishier than the title would make you think. A complicated statistical analysis suggests that there are correlations between a bunch of parameters, which might invalidate a model of galaxy formation that might predict less of a correlation, and because that model involves cold dark matter, it might invalidate the CDM hypothesis. That's a lot of mights.
I'm not really in a position to evaluate this, though. One of my colleagues at work probably is, but I'm not going in today, so I'll throw this out to my astro-savvy readers: what do you make of this story?
- Log in to post comments
More like this
Looks like this is the paper, and Julianne over at Cosmic Variance is on it as well. My initial guess is that this isn't actually a Problem for Lambda-CDM, but it makes for better spin in an article if it is.
I think the problem with dark matter is its name. Maybe changing the name to something like "solid existing matter" would make people question its existence less often.
And how do they explain the observed rotational velocities in galaxies?
It has been known since the 1970s that the velocity of rotation of luminous matter (mostly stars) in galaxies falls off much more slowly (or not at all) than the r-1/2 dependence predicted from Keplerian orbital mechanics. The implication is that a halo of mass which we cannot see surrounds these galaxies.
Since then there have been other studies that point to the existence of non-baryonic dark matter, but the rotational velocity curves were one of the early clues.
They may have invalidated a particular model of galaxy formation (I haven't read the paper in question, so I can't speak to how good their result is). They have definitely not invalidated the CDM hypothesis.
I really don't know about this new paper, but at least J. Bekenstein is no kook - and here:
http://xxx.tau.ac.il/PS_cache/astro-ph/pdf/0701/0701848v2.pdf
He gives an overview of an/the alternative for CDM cosmology.
The answer to Eric's question is given in figure 5.
non-luminous matter?
Could someone please explain (or give a link to an explanation) why CDM would form halos around galaxies rather than distributing its mass the same way throughout a galaxy in the same way as normal matter - dense in the center and less dense as the distance from the center increases?
Tophe: I guess the word "halo" is a bit misleading. The dark matter is distributed (roughly) smoothly around the galaxies, with the dark matter being denser at the center (where the gas+stars of the galaxy is) and much more diffuse further away... we don't mean anything like a "ring" or something like an angel's halo would be :)
I thought that dark matter's density was more uniform than ordinary matter because it doesn't lose energy through interaction with other matter, and that's why the visible matter seems to violate Kepler's law. Did I misunderstand this?
Thanks mollishka :)
Riesz Fischer @ 8:
Yes, the dark matter is thought to be less centrally concentrated than ordinary matter. It is still centrally concentrated, however.
I'm on the paper, and I think dark matter exists. My collaborator would really enjoy it if dark matter did not exist.
I think the trend described in the paper suggests that there is an awful lot of self-regulation going on, and while Mike thinks that makes it hard for CDM, I just think it makes it interesting for self-regulation and feedback.
Doesn't the Cold in CDM refer to the tendency of the mass to clump? At least if you think of CDM as being composed of classical particles (as opposed to say fuzzy particles with a very broad mass distribution), then the average velocity should be related to clumpiness, if they were all traveling very slowly, they could clump into a small space. But of course average velocities are presumably determined by cosmological determinants, like the escape velocity of galaxies and galaxy clusters. What would be the difference between CDM, and warm dark matter?
bigTom @ 12:
The "cold" in CDM refers to the average velocities of the particles, which in turn depends inversely on their mass: the more massive the particles, the lower their average velocity. The velocities depend mainly on conditions very early in the Big Bang, not on "local" things like galaxy escape velocities.
You're right that the "temperature" of the particles affects their ability to clump: CDM particles clump relatively easily on small scales, so small, low-mass halos are predicted to form first, and then galaxy-size halos are built up through mergers of smaller halos. That's why CDM models of galaxy formation suggest that galaxies are built hierarchically, through mergers of smaller sub-units. "Hot" DM particles (e.g., neutrinos) would form cluster- or supercluster-sized halos first, which would then gravitationally fragment into smaller halos. But this leads to lots of wrong predictions for when galaxies form and how they appear to cluster together, and wrong predictions for cosmic background radiation fluctuations.
I told my college Astronomy students, and tell my high school students, that modern Cosmology is a shaky 4-storey building in an earthquake zone.
1st floor: Universe is expanding. Hence Big Bang.
2nd floor: Galaxies rotate in apparently nonkeplerian way, hence putative cold matter, for which there is other evidence (cf. bullet galaxy).
3rd floor: universe is more uniform than can esily be explained, hence Inflation at greater than the speed of light for unknown reasons that stopped for unknown reasons. Maybe.
4th floor: expansion of universe might be accelerating, hence something we don't understand at all, called Dark Energy.
Penthouse: universe being gravitationally attracted by something humongous ourside the universe, and other science fiction.
When the paradigm collapses, how much of the Tower of Babel will still be standing?
The recent work by the Galactic Black Hole Survey Team (the had a whacky name, can't remember) posited the theory that the unexpectedly high orbital speed of outlying stars in galaxies is directly related to the black holes at their cores. There survey showed an exact correlation beween galactic black hole mass and orbital speed of outlying stars. Their (surely oversimplified) layman's explanation was that, in a young universe, there were only large active black holes and their massive accretion disks. The black holes imparted their own rotational speed to the disks, which eventually coalesced into galaxies.
Don't know if that damages the CDM theory, there must be other lines of evidence?
If the question is whether the correlations among the observables could be accounted for by their all being correlated with Something Else, the right tool to use to answer the question is factor analysis, which (from a quick scan of arxiv:0809.1434) they don't seem to use. If they only reported the full correlation matrix...
Just so we are talking about the right paper: The article Chad quoted is mainly referring to a recent Letter to Nature which hasn't yet appeared on the arxiv. Disney et al. Nature 455, 1082 - 1084 (23 Oct 2008)
If you have a subscription:
http://www.nature.com/nature/journal/v455/n7216/pdf/nature07366.pdf
As the journalist who wrote the story for physicsworld.com, I would like to clarify that I talked to Mike Disney and assumed that he spoke for all the authors on the Nature paper as regards to the existence of cold dark matter. As Julianne Dalcanton points out, Disney has strong opinions on its non-existence, with which she doesn't apparently agree!
Hamish Johnston
Editor, physicsworld.com
What shape galaxy are we talking about? After all, the Milky Way may be a Barred Spiral... I love the phrase "the possible destruction of the halo cusp" which sounds ambiguously Hard SF and High Fantasy.
arXiv:0810.4925 [ps, pdf, other]
Title: Anatomy of the Bar Instability in Cuspy Dark Matter Halos
Authors: John Dubinski, Ingo Berentzen, Isaac Shlosman
Comments: 22 pages, 29 figures, 4 animations available at this http URL
Subjects: Astrophysics (astro-ph)
We examine the bar instability in galactic models with an exponential disk and a cuspy dark matter (DM) halo with a Navarro-Frenk-White (NFW) cosmological density profile. We construct equilibrium models from a 3-integral composite distribution function that are subject to the bar instability. We generate a sequence of models with a range of mass resolution from 1.8K to 18M particles in the disk and 10K to 100M particles in the halo along with a multi-mass model with an effective resolution of ~10^10 particles. We describe how mass resolution affects the bar instability, including its linear growth phase, the buckling instability, pattern speed decay through the resonant transfer of angular momentum to the DM halo, and the possible destruction of the halo cusp. Our higher resolution simulations show a converging spectrum of discrete resonance interactions between the bar and DM halo orbits. As the pattern speed decays, orbital resonances sweep through most of the DM halo phase space and widely distribute angular momentum among the halo particles. The halo does not develop a flat density core and preserves the cusp, except in the region dominated by gravitational softening. The formation of the bar increases the central stellar density and the DM is compressed adiabatically increasing the halo central density by 1.7X. Overall, the evolution of the bar displays a convergent behavior for halo particle numbers between 1M and 10M particles, when comparing bar growth, pattern speed evolution, the DM halo density profile and a nonlinear analysis of the orbital resonances. Higher resolution simulations clearly illustrate the importance of discrete resonances in transporting the angular momentum from the bar to the halo.