Friday, October 10, 2014

More Problems With Dark Matter

In a new preprint by Wu and Kroupa of October 8, 2014, accepted for publication in MNRAS and entitled "Galactic rotation curves, the baryon-to-dark-halo-mass relation and space-time scale invariance" these advocates for MOND and related theories in lieu of dark matter raise multiple new serious concerns about both cold dark matter and warm dark matter theories.  They also use statistical methods applied to large sets of galactic rotation curve fits to determine which of the MOND interpolation formulas is more accurate.

The highlights:

* Empirically, the fit of galactic rotation curves to baryonic mass distributions is much more tight than in the galaxies produced by CDM and WDM simulations, despite the fact that the observational evidence has measurement errors which are absent in the simulation data.

The "standard" interpolation function in MOND models outperforms the "simplified" version, although both are very good fits.  The WDM model has about twice as much scatter as the "simplified" MOND theoretical prediction line.  The "tamed" CDM models that make unphysical assumptions to replicate reality are twice as bad as WDM simulations.  And, the fully realistic CDM models have massive scatter despite the fact that the real world galaxies behave almost precisely as the standard interpolation function from MOND predicts.

A non-parameteric statistical test disfavors all of the CDM and WDM models and the simple MOND model when compared to the observational evidence by 99.99%, while finding it 80%-90% likely that "standard" MOND and another sophisticated MOND variant called SID are a fit to the data.  This is particularly stunning given the amount of measurement error involved.

The link between baryonic matter distributions and apparent dark matter halo distributions is so tight with such a simple formula that it is very hard to see how any DM theory could achieve this result.  The "scatter" of different possible dark mater halos for given baryonic matter distributions in DM theories are fundamental, built in, parts of the theory.

* All DM model simulations systemically over predict the amount of dark matter in galaxies of a given baryonic mass, while the amount of dark matter that would be necessary to reproduce the MOND results is right on the money.

* Kroupa acknowledges to important facts that one has to correct for when using MOND to predict galaxy dynamics: (1) the non-luminous interstellar gas to star mass ratio, and (2) the impact of fields from nearby objects (e.g. the impact of gravity from a central galaxy on the dynamics of a satellite galaxy) aka the "external field effect".

In general, more non-luminous interstellar gas simply requires an adjustment in the estimated amount of ordinary baryonic matter in a system.

Outside gravitational fields, in general, reduce the apparent amount of dark matter in the system in the eyes of a Newtonian observer.

Bonus trick:  If you know the amount of luminous matter and interstellar gas in a system, and how fast it is rotating, you can use this slightly more sophisticated version of MOND to determine how far the influenced gravitational system is from the outside object created gravitational system.

* Dark matter halos inferred by Newtonian observers when MOND is used to govern dynamics have the correct isothermal shape the observational evidence implies, rather than the NFW distribution that CDM models naturally produce.

* About 70% of disk galaxies are "bulgeless" and MOND can easily reproduce these dynamics in the face of various galaxy formation scenarios.  But, WDM and CDM models create far, far too few bulgeless galaxies, because those models tend to create serious bulges whenever DM halos or galaxies merge, which happens a lot.  Some WDM/CDM simulations cheat to prevent relatively equal galaxy collisions like the Bullet cluster and get results more like real life, but there is no physical mechanism to cause that to happen and we know that big collisions are far more common than the "tame" simulations permit.

* These very accurate results are achieved with the usual one MOND parameter (plus external field effects and gas-star mass ratios in particular galaxies), and very easy math compared to many competing theories.

* It sounds like some progress has been made in physically motivating the equations and in reducing the amount by which apparent DM is underestimated at cluster scales from the "old days" although these continue to be problems.

I'll admit that I've definitely flirted with WDM models, which are definitely closer to observational evidence than CDM models that perform just dismally.  But, this analysis makes it extremely hard to believe that there are really any dark matter particles out there at all.


Tienzen said...

Andrew: “But, this analysis makes it extremely hard to believe that there are really any dark matter particles out there at all.”

Now, you are talking. Can I say {I told you so}?

By all means, the dark matter (in terms of ‘particle’) is still a speculation. That is, it is OK to be wrong and be total nonsense.

On the other hand, there are ‘four established knowledge’ which are totally wrong and totally nonsense.

The first one is the Higgs ‘mechanism’ regardless of the hype for naming the newly discovered boson as the Higgs boson.

The second one is the Darwin mechanism, totally wrong and totally nonsense. Fortunately, this situation is a bit better than the Higgs fiasco as someone is now rethinking the issue. See, .

Anonymous said...

I'm glad that you scan the web for interesting papers that some times fail to get highlighted by the mainstream science media. The post of W boson helicity was quite nice. However, in this post, you seem to be covering MOND theories without really digging into the many significant problems associated with MOND.

The authors of the paper you reference make the same mistake that proponents of CDM make when they attempt to critique WDM. They forget to treat the keV dark matter as a fermion, quantum degenerate particle. This, along with possible consumption of the keV particle in the center of galaxies by black holes, can explain the lack of a dark-matter-bulge near the center of galaxies. MOND is not required, and hence WDM is in my opinion the best explanation for all of the data. (And I mean all of the data.)

MOND and MOND-like theories are only attempting to fit the data at the galaxy level. They completely fail to explain all of the other necessary properties of dark matter:
(1) First, dark matter is the original source of clumpiness when the CMB was generated. Without dark matter, you would have a very different shape to the CMB curve.
(2) With dark matter, the total energy of the universe is near zero (as predicted by inflationary models.) Without dark matter, what do MOND theories predict the total energy of the universe to be?
(3) How does MOND respond to gravitational lensing by dark matter?
(4) How does MOND explain the Bullet Cluster? (and all other cases when the matter and the dark matter are not collocated...which is very often.)

As such, it seems like you (and MOND proponents) are cherry picking the data and ignoring the other failed predictions of MOND. I suggest that you be more discerning in your coverage of MOND papers.

andrew said...

Re WDM theories - the key point is that WDM does not particularly tightly track baryonic matter distributions, while empirical evidence does. This predicted scatter is distinct from the core-cusp issue which you correctly note that WDM addresses.

Re: Points 1-4 about MOND.

(3) MOND and MOND-Like theories generally predict identica gravitational lensing to what is observed.

(4) MOND has issues with clusters in general. But, other gravitational approaches like that of Deur and of Moffat explain the Bullet Cluster.

(2) I'm not sure that I understand what you are asking in point (2). Perhaps you could spell out what you mean.

(1) MOND and MOND-like cosmologies can produce cosmologies that match observation, although they get their in different ways. MOND itself doesn't do a great job of reproducing the CMB curve and there hasn't been a huge amount of cosmology work on modified gravity theories generally. But, at least heuristically, there are mechanisms in Deur's analysis, for example, that explain both clumpiness and much of dark energy as a corollary of that clumpiness in a way that also resolves the cosmic coincidence problem.

andrew said...

Also, WDM does not solve the bludgeless galaxy frequency issue that it shares with CDM.

Ryan said...

I think Eddie is asking with (2) whether or not the universe is still flat or nearly flat under MOND, as that is a very elegant outcome with dark matter.