MOND vs. Dark Matter

The first serious advocate of modifying Newton's laws instead of postulating unseen (or dark) matter was Moti Milgrom, from whom today a new article appears on the astrophysics preprint archives.

In particular, Milgrom asserts the following:

MOND predictions imply that baryons alone accurately determine the full field of each and every individual galactic object. These predictions are contrary to the expectations in the DM paradigm in view of the following:

a. the haphazard formation and evolution of galactic objects,

b. the very different influences that baryons and DM are subject to during the evolution, and

c. the fact that the baryon-to-DM fraction in galaxies is much smaller than the cosmic value.

All this should disabuse DM advocates of the thought that DM will someday be able to reproduce MOND...

Now, I understand that, as the founder of MOND, Milgrom is obliged to give it the best defense that is available to it. And indeed, there are problems with the dark matter paradigm; many such problems were examined in detail in this paper. MOND's basic tenet is that, at very very small accelerations, the laws of Newtonian gravity need to be modified. There are a whole host of problems with MOND, perhaps the most glaring ones are as follows:

1. MOND relies on all Baryons and no Dark matter, which is in horrendous conflict with observations of Large-Scale-Structure and the Microwave background. The Large Scale Structure data is even worse, but not as facile to generate. Take a look at the difference between the CMB data with both dark matter and baryons:
   and the CMB data with baryons alone:
   
   Now, let's compare this to the ACTUAL data available:
   The amplitude and location of the first peak alone should be enough to convince you that we need dark matter to reproduce this.
2. We can *see* that gravity doesn't line up with where the baryons are in different places! MOND can't explain this at all.
3. Dark matter is predicted to be in a halo, while MOND relies on the baryon distribution to create the gravitational signal necessary. There are examples where these two don't align, and they favor dark matter, not MOND.
4. There is still an unexplained mass discrepancy in galaxy clusters with MOND, whereas numerical simulations of Dark Matter yield the right mass distribution for galaxy clusters.

I can write a lot about why Dark Matter is highly supported by the evidence, but there are problems with Dark Matter that we don't understand. For example, we don't understand why there is such an apparent variety between the baryon/dark matter ratios in galaxies, particularly in low-luminosity dwarf galaxies. Also, MOND is far superior to all models of dark matter in predicting the observed rotation curves of galaxies. But really, that's about it.

As for Milgrom's points,

a. Yes, we are still researching how galaxies form, merge, and evolve. There's a lot we don't know, because it's an active area of research. MOND doesn't fix any of the problems associated with galaxy formation and evolution, though.

b. Dark Matter interacts only gravitationally, while baryons interact with photons and themselves through electromagnetic and nuclear interactions. We know this. It's tricky to model it, but again, it's an active area of research. Just because there are caveats we don't understand doesn't mean that we don't understand anything.

c. "Much smaller"? This isn't true. Within the errors, and within the variations observed, the ratios are pretty consistently, in all but the smallest galaxies, about 15% baryons to 85% dark matter. There was a great (technical) paper a few years ago called

the cosmic energy inventory

, where this was articulated explicity, and the non-luminous baryons were accounted for.

In summary, MOND explains galactic rotation curves better than dark matter does, and pretty much nothing else. Dark matter explains the cosmic microwave background, large scale structure, galaxy clusters, gravitational lensing, the Bullet cluster and the cluster Abell 520, and also is consistent with a host of cosmological observations (e.g., nucleosynthesis, supernovae data, cluster estimates of the matter density) that MOND isn't. All that really remains is to detect dark matter directly.

Until then, however, MOND will continue to be tweaked to be as consistent as possible with observations, and will continue to tout its one success that dark matter hasn't been able to match. And this isn't a bad thing; MOND has encouraged the dark matter people (of which I'm one) to step up their efforts to explain rotation curves and solve the open questions associated with dark matter, and finally, to detect it. We're working on it...