scilogs Dark Matter Crisis

State-of-the-art cosmology: the current status

Pavel Kroupa | 25. August 2010, 10:00

Last week's contribution "Is LambdaCDM or standard cosmology a 4th order speculation, and ought it be further researched?"  was concerned with the recent suggestion by Prof. Abraham Loeb that alternative approaches should be followed to advance science. But at the same time he proposes the alternative approach "MOND plus netrinos" to be a second order speculation not worth the effort.

Following this logic of Loeb, it becomes immediately apparent that LCDM, or standard/concordance cosmology, is at least a 4th order speculation, with the corresponding implications.

Dr. Garry Angus is a very talented young cosmologist currently at the University of Torino, Italy, but shortly moving to the University of Cape Town, South Africa, who has been working in the field of "MOND plus neutrinos". Dr. Angus is the recipient of the Cormack Bequest Prize for his 2007 publication on the topic of the Bullet Cluster, Neutrino Dark Matter and Alternative Gravity. This prize is awarded annually to the most outstanding postgraduate student contribution to astronomical research in Scotland.

Below he directly addresses Abraham Loeb's assertion concerning his field.

It should be noted, before reading Garry's text, that in the LCDM field (4th order speculative science) whole armies of researchers (hundreds?) have been toiling over the past decade to improve the computations and observations. It is the accepted model of cosmology, and over the past 10-15 years the very major professorships in cosmology or extragalactic astrophysics have been filled with experts in this one specific field. In contrast,  "MOND plus neutrinos" (2nd order speculative science according to Abraham) has been worked on by not more than about 2 researchers, while the other alternative, Modified Gravity (MOG), has been worked on by not many more researchers than that as well.

It nevertheless turns out that LCDM sort of works on large scales, and MOND plus neutrinos does at least as well as far as the existing work allows us to judge. Indeed, as Dr. Garry Angus shows below, the cosmic microwave background (CMB) power-spectrum is fitted perfectly well in MOND + neutrinos. On  scales smaller than about ten million light years LCDM fails however, while non-Newtonian/Einsteinian gravity works brilliantly (Kroupa et al. 2010). 

Thus, MOND plus neurinos seems to be the astrophysically most modern and successful cosmological description we have currently. 

Dr. Garry Angus writes:

I'd like to just make a comment on why MOND+neutrinos is not a 2nd order speculation. I  don't know how familiar you are with the literature on MOND+neutrinos, but no one, to my knowledge, has ever suggested that the CMB can be fit by MOND plus the active neutrinos - be they 2.2eV or 0.1eV. Skordis et al. (2006) clearly showed the apparently high 3rd peak is not compatible with even 3x2.75eV in TeVeS, even if the critical MOND acceleration (a_0) is boosted by a factor of 4.

It should be a well known fact to all cosmologists that replacing  Omega_CDM x h2  with the same energy density in sterile neutrinos gives as good a fit to the CMB. The proviso is that this energy density comes in the form of a single, thermal sterile neutrino species. Given a reasonable mixing angle, it is perfectly possible for these sterile neutrinos to be thermalised in the very early Universe. This means the neutrino has a mass of ~11eV. A figure of the MOND + neutrino CMB calculation can be seen here where we fit both WMAP 7 and ACBAR data. As can be seen, the theortical fit is near to perfect to the CMB data.

This has nothing to do with MOND. In fact, it requires MOND to have no influence at redshift z>1000 and a cosmological constant is still required. It just so happens that an 11eV sterile neutrino would resolve all problems MOND has in clusters of galaxies. At 100~kpc (about 300 thousand light years) in basically all clusters there is a 10:1 ratio of DM:baryons (after accounting for MOND), it is only at distances like 1Mpc that there is a 2:1 ratio. Angus, Famaey & Diaferio (2010) looked at 30+ groups and clusters and made the intriguing observation that the Tremaine-Gunn limit for the 11eV neutrinos is reached in every system, but never need be exceeded for very sensible values of the brightest cluster galaxy's mass-to-light ratio.

These neutrinos would free stream out of Milky Way type galaxies, so all the successes of MOND at galaxy scales would be unaltered. The ramification of this is that the galaxies must collapse under their own gravity (enhanced by MOND) without the aid of a cold dark matter halo (see Sanders 2008). Linked to this, we have run preliminary cosmological simulations that incorporate MOND and 11eV sterile neutrinos and the conclusion is that they form roughly the correct number of clusters of galaxies as a function of cluster mass. It could just as  easily have ended up in a big black hole or with no structure forming at all.  If we run numerical simulations with the 11eV neutrino and no MOND, then no structures form i.e. MOND is essential for massive neutrinos to work.

Whether the correct number of galaxies form is an incredibly difficult question to answer and the numerical tools are nowhere near ready - basically because the 2 or 3 people with the necessary expertise to develop the codes can't get jobs for love nor money. However, based on the successes of MOND at galaxy scales, we do expect that MOND+sterile neutrinos can reproduce that observed properties of galaxies with no effort, unlike CDM. For example, as long as a galaxy forms, we know trivially that it will conform to the baryonic Tully-Fisher relation. Furthermore, the highly organised distribution of satellite galaxies surrounding the Milky Way (see Kroupa, Theis & Boily 2005; Metz et al. (2009); Kroupa et al. 2010)  will immediately be explained to be tidal dwarf galaxies. Currently they have no explanation in LCDM.

For these reasons I don't see how MOND+neutrinos is a second order speculation. If one has MOND, then there is an incredibly high chance that 11eV sterile neutrinos must exist and by the same token, if sterile neutrinos at 11eV exist then MOND is needed. In the former case it is possible that a deeper theory of MOND will spring a surprise which conspires at cluster scales, for the expansion history, during the formation of the acoustic peaks of the CMB and during structure formation to resemble an 11eV sterile neutrino. Abraham Loeb mentioned briefly that the baryonic acoustic oscillations appear at the scale predicted by LCDM. I don't believe this is fully accurate. They appear at the scale defined by a scale factor that evolves as if it has a dominant dark matter component, hot, cold or warm. The peaks themselves do not require the dark matter to be cold.

A lot of people talk about neutrinos being against the design, spirit or original intention of MOND. I feel this is never very helpful. As Professor Milgrom clearly states above, the state of observational astronomy was very different in the early 80s. That MOND works at all highly disfavours the need for any type of Warm or Cold dark matter, but if that dark matter is hot and hot enough to free stream from galaxies then MOND (in the 80s) made no predictions about its presence. Nowadays, MOND without sterile neutrinos and MOND with sterile neutrinos are two different models with very different predictions for cosmology.

In addition to the cosmological evidence for 11eV sterile neutrinos, there also exists tentative particle physics evidence from Miniboone (see Giunti & Laveder 2008). And, more experiments are in the pipeline, for instance the T2K experiment will be able to put excellent constraints on 11eV sterile neutrinos, with results probably released in early 2012. Unfortunately, Planck will not offer any evidence for the specific mass of sterile neutrinos because the CDM model with a very low mass (say <0.1eV) thermal sterile neutrino would generate an identical power spectrum. It can, however, rule out the existence of further thermalised neutrino species (i.e. if N_eff=3).

I should add that although 11eV sterile neutrinos is my preferred solution that extends MOND to cosmology, there are others. HongSheng Zhao & Baojiu Li are working very hard testing a model that combines the MOND effect, the cosmological dark matter and dark energy into the same field that behaves differently depending on environment, which is a very nice idea. It boils down to the same essential ingredient for cosmology, however, and that is hot dark matter plus MOND.

by Anton Ippendorf, Pavel Kroupa and Marcel Pawlowski (25.08.2010, "State-of-the-art cosmology: the current status" in "The Dark Matter Crisis - the rise and fall of a cosmological hypothesis" on SciLog. See the overview of topics in  The Dark Matter Crisis.

Posted in General Comment: (0). Trackbacks: (0). Permalink

Is LambdaCDM or standard cosmology a 4th order speculation, and ought it be further researched ?

Pavel Kroupa | 18. August 2010, 10:00

On August 9th, 2010  Prof. Abraham Loeb from Harvad University  published a stimulating paper on the electronic preprint server with the title "Taking "The Road Not Taken'': On the Benefits of Diversifying Your Academic Portfolio".

In this paper he takes issue with which type of research ought to be supported, making the much noted suggestion that innovative projects not following the main stream should be invested in by young researchers, in addition to following more secure research directions.  Abraham points out that even if one in a million new ideas bear fruit, this may completely transform our understanding of reality justifying the entire effort of using a certain fraction of funding for new, risky ideas.  To achieve this goal, he recommends that each researcher should spend up to about 50 per cent time on non-standard research, while individuals will naturally develop their own strategy based on personal and social factors. His suggestion is further that senior members of the community should find better strategies for rewarding innovation. (More)

Posted in General Comments: (2). Trackbacks: (2). Permalink

The Train Wreck Cluster - an "anti-Bullet-Cluster": disproof of Cold or Warm Dark Matter

Marcel S. Pawlowski | 11. August 2010, 10:00

This is the final part of our small series on the Bullet Cluster (and galaxy clusters in general). In the first part we have already argued that the Bullet Custer can not be used as a "smoking gun" for dark matter and even poses a problem for concordance cosmology. The second part laid out that theories of modified gravity can account for galaxy clusters and expecially for the Bullet Cluster, too.

There is one cluster which, in some respects, resembles the Bullet Cluster: Abell 520 (see also: Cosmic 'train wreck' defies dark matter theories).  

Source: Chandra X-ray observatory site, Harvard University.

Similar to the Bullet Cluster two galaxy clusters have collided recently. Consequently, the hot gas is again found in the middle of the clusters. And again one expects Dark Matter, as it is collisionless in contrast to the gas (which, when atoms collide, radiates its energy in the form of light and thus cools and slows down), to be centered on the two galaxy clusters. But to the surprise of the Dark-Matter community, Mahdavi et al. (2007) found a “Dark Core in Abell 520”, that is, there is Dark Matter in the center where no galaxies are.

This object therefore looks like the inverse of the Bullet Cluster. Things look messed up, that's why the object got the name “Train Wreck Cluster”. We did not find an explanation for it in the literature and one of us, Marcel Pawlowski, even discussed it's case with standard cosmologists. Up to now, they all agree that we do not understand it in Standard Cosmology. Interestingly, the alternative gravity community has come up with an explanation, such as Moffat and Toth (2009) for MOG.

After finding out about the existence of this “Train Wreck Cluster”, one question cames to mind: How is it that everybody mentions the Bullet Cluster as a proof of Dark Matter, but (almost) nobody ever talks about the Train Wreck Cluster? Does an object for which the theory gives a good explanation have more "evidence-value" than an object which seems to be at odds with the theory?  Isn't that a bit too selective for scientists? In fact, while during discussions everybody points at the Bullet Cluster, many people and even a lot of astronomers do not even know about Abell 520!

We have to be really careful here. Always pointing at one object as the ultimate proof for dark matter and not mentioning a counter-example isn't good science. In fact, this selective reporting distorts the evidence especially towards people who do not and cannot acquire the objecitve information - the public gets a wrong impression. 

And stating that galaxy clusters can not be explained in modified gravity theories while there are peer-reviewed papers doing exactly that is very bad style and positively unscientific. The whole problem of the existence of Cold or Warm Dark Matter should not be about opinions, but about science. And the evidence is defintely not in-favour of its existence.

A Radical Conclusion

Why can we make such a radical statement depite the vast majority of fellow-scientists expressing the oposing view?

Well, given the material on galaxy clusters presented here it is very clear that the Cold- or Warm-Dark-Matter hypothesis has problems with galaxy clusters, particularly with the Bullet and Train-Weck Clusters. Non-Newtonian approaches on the other hand seem to easily account for them. And, the Local Group of galaxies (and thus us humans) cannot really exist in a Cold- or Warm-Dark-Matter universe.

Putting this together we get a positively dark view of Dark Matter Cosmology, while the alternative models (MOND or MOG or ... ?) yield a notably bright window towards a much more superior description of cosmological reality.  

by Anton Ippendorf, Pavel Kroupa and Marcel Pawlowski (11.08.2010): "The Train Wreck Cluster - an anti-Bullet-Cluster: disproof of Cold or Warm Dark Matter" in "The Dark Matter Crisis - the rise and fall of a cosmological hypothesis" on SciLog. See the overview of topics in  The Dark Matter Crisis.

Posted in General Comments: (4). Trackbacks: (0). Permalink

The Bullet Cluster and galaxy clusters in modified-gravity theories

Anton Ippendorf | 04. August 2010, 10:00

This is our second post on the topic of the Bullet Cluster and galaxy clusters in general. Here you can find the first part: "But the Bullet Cluster ...", in which we argue that it can not be used to proof the Dark Matter Hypothesis. The third and last part, about an 'anti Bullet Cluster', will be published in a few days.

Modified gravity theories do well on galaxy scales. Even Ethan Siegel accepts that they are better than Dark Matter at these small scales. But can they cope with the Bullet-Cluster observations? Ethan says no. Let us explain why this statement is wrong.

As mentioned before, the relative velocities of the two galaxy clusters are too fast to be consistent with the standard cosmological framework. In theories of modified gravity, in contrast, high relative velocities of galaxy clusters occur because the effective gravitational acceleration is stronger. And this does take into account that there is less mass because there is no dark matter. The increase in acceleration out-weighs this. High impact velocities naturally occur in MOND, as Angus and McGaugh (2008) have shown, while Moffat and Toth (2010) were able to resolve the infall-velocity issue using MOG.

Thus, there is at least one problem on galaxy cluster scales where modified gravity theories do better than Dark Matter. They not only convincingly triumph on the small scales (as even Ethan Siegal submits to and as is demonstrated scientifically in our research paper). Seen this way, one can just as well use the Bullet Cluster as an example against the Cold Dark Matter Hypothesis.

But how to explain the missing mass in galaxy clusters? It can not all be in hot gas because of the observed offset between the lensing mass and the gas emission in the Bullet Cluster. Assuming the validity of Newtonian Dynamics, the mass missing in clusters of galaxies is about a factor of 4 more than what can be seen in luminous galaxies and gas. Life is made much easier in Modified Newtonian Dynamics (MOND) as Sanders (1999) has shown, because in MOND less mass can produce the same acceleration. This reduces the factor of missing mass to only two times the visible mass. This amount of missing mass might be found in neutrinos, or it might not even be real: a factor of two is not much in astronomy. And a systematically larger mass may be obtained through biases hitherto not taken into account.

We know that neutrinos oscillate, therefore they must have a mass. That mass is small. This makes them a form of hot dark matter that we most definitely know to exist. In order to explain the oscillations, particle physics even predicts the existence of more massive, sterile neutrinos, which only interact by gravity. If they exist they might be massive enough to account for the missing mass in galaxy clusters in MOND (and they can fit the first three acoustic peaks in the CMB).

For Moffat's theory of Modified Gravity (MOG) the situation is different still: Analyzing the Bullet Cluster, Brownstein and Moffat (2007) realized that in MOG, no Dark Matter is needed at all:

"Using Modified Gravity (MOG) theory, the ‘normal’ matter in the Bullet Cluster is enough to account for the observed gravitational lensing effect."

So, please, do not say the Bullet Cluster or the high speeds of galaxies in clusters kill all alternative gravity theories. They don't. In fact, they might just be arguments for non-Newtonian gravity.

by Anton Ippendorf, Pavel Kroupa and Marcel Pawlowski (04.08.2010): "The Bullet Cluster and galaxy clusters in modified-gravity theories" in "The Dark Matter Crisis - the rise and fall of a cosmological hypothesis" on SciLog. See the overview of topics in  The Dark Matter Crisis.

Posted in General Comment: (0). Trackbacks: (1). Permalink