MOND and the missing baryons problem

[wolram]

http://arxiv.org/pdf/1508.04001v2.pdf

Although i am not a proponent of MOND this paper gives an alternative to the missing baryons problem

The colliding ‘Bullet Cluster’ is often adduced as strong evidence for DM, with the understood – 3 – implication that it is evidence for the non-standard-model particle DM that is invoked by the mainstream to fill the Universe. In the case of the Bullet, most of the baryons (in the form of hot gas) are found in the collision zone, while the ‘DM’, as deduced from weak lansing, is mainly in two regions flanking the collision zone, where the galaxies from the colliding clusters are found, after they went through, hardly affected by the collision. The claim then goes, that modifieddynamics alternatives to DM perforce predict that the ‘phantom DM’ should appear where most of the baryons are, unlike what is seen in the Bullet . This statement, in itself, is not correct. Modified dynamics theories, including MOND, do not predict that the discrepancies should follow the baryons. It is true, though, that purist MOND does not account for the observed geometry of the Bullet without invoking some yet undetected matter in the system. However, just as everything that glitters is not gold, everything that is dark is not the DM. What is ‘seen’ in the Bullet might well be just an inkling of small amounts of yet undetected baryons indigenous to clusters.

 

[DEvens]

Have not read the preprint paper you cite. But IIUC, it is saying we need MOND *and* dark matter. It seems a long journey.

 

[wolram]

From the paper:

MOND is a paradigm of dynamics that strives to explain away the mass discrepancies in galactic systems, and the Universe at large, without dark matter (DM). It departs from standard (e.g., Newtonian) dynamics at accelerations that are lower than the MOND acceleration constant a0 ≈ 1.2 × 10−8 cm s−2 . Recent reviews of MOND are Famaey & McGaugh (2012) and Milgrom (2014).

 

[Chalnoth]

I find it really sad that so many are still harping on MOND. Those ideas should have been dead and buried quite some time ago. MOND never did very well with clusters, even before the Bullet Cluster observations, and I don't think they've come close to explaining the CMB anisotropies.

 

[DEvens]

I find it really sad that so many are still harping on MOND. Those ideas should have been dead and buried quite some time ago. MOND never did very well with clusters, even before the Bullet Cluster observations, and I don't think they've come close to explaining the CMB anisotropies.

I have a different view.

I think MOND is probably just about as dead as you indicate. It has been a while since my admittedly hasty reading about MOND, but I think it has even other problems besides and in addition to the ones you list. But it still has potential value.

Dead ideas can be useful in pedagogical ways, even if for no other purpose than "this is what a wrong theory looks like."

Dead ideas can be thought provoking. The fact that a theory met some, hastily admittedly limited and possibly illusory, success, means the ideas in it might be instructive in thinking about how to deal with the problem at hand. The fact that things "look like" MOND, in some limited way under some highly restricted conditions, is interesting in what it can tell us. I'm expecting that it does not tell us that anything about MOND is right. But it could tell us, for example, how to seek to better resolve what a better answer is. "It looks like particles in this region are departing from Newton by this tiny amount" might be an interesting place to start working on the problem.

Single light of science and all that.

 

[ohwilleke]

MOND is important, even if it isn't true, because it illustrates that it is possible to phenomenological model all dark matter phenomena from solar system scale to large galaxy scale with a single additional degree of freedom, and you can extend it to model cluster scale phenomena with just one or two more degrees of freedom.

This, in turn, implies that even if dark matter is the answer, that it ultimately have to be a very simple phenomena and that there is far more order in the data set than dark matter theory would naively expect that has to come from somewhere. If your dark matter model can't substantially reproduce the galactic scale predictions that MOND does with a single new parameter, then it is wrong.

Further, MOND is a good way to develop intuition about the way that our observed universe differs from a no-DM universe governed only by GR. The really interesting deviations between expectation from a no-DM GR universe and observation are predominantly in extremely weak gravitational fields, not in the strong field limit, and are more pronounced in clusters than in galaxies. (MOND and DM both predict no measurable solar system scale effects).

 

[Earnest Guest]

I find it really sad that so many are still harping on MOND. Those ideas should have been dead and buried quite some time ago. MOND never did very well with clusters, even before the Bullet Cluster observations, and I don't think they've come close to explaining the CMB anisotropies.

Isn't this a bit like the pot calling the kettle 'black'. ΛCMD has been disproven by LSB galaxies as much as MOND has been disproven by the Bullet Cluster. Oh, and there's also the issue that there's no dark matter particle. And the energy that comes from empty space. And the ad-hoc field that starts and stops for absolutely no reason other than to make the universe flat. I think there are enough ridiculous ideas for everyone here. Take the splinter out of your eye and then you can see clearly to take the splinter out of your neighbor's eye.

 

[rootone]

I remain to be convinced that the MACHO hypothesis is completely discountable as well, although I'm not an enthusiast for it.
It could still be a possiblity if the universe contains a large amount of black dwarf like objects or wandering remnants of long dead solar systems that are too dark and small to be detectable.

 

[ruarimac]

Isn't this a bit like the pot calling the kettle 'black'. ΛCMD has been disproven by LSB galaxies as much as MOND has been disproven by the Bullet Cluster. Oh, and there's also the issue that there's no dark matter particle. And the energy that comes from empty space. And the ad-hoc field that starts and stops for absolutely no reason other than to make the universe flat. I think there are enough ridiculous ideas for everyone here. Take the splinter out of your eye and then you can see clearly to take the splinter out of your neighbor's eye.

Is it though? While it's certainly true LambdaCDM is not perfect but in my opinion it is simply in a different league. MOND sets out to explain just dynamics and I don't think that has been achieved in a consistent way even across all galaxy types never mind moving up to the scale of clusters. But then there is everything else, for example does MOND describe galaxy clustering? No. MOND isn't really a cosmology because it just doesn't work on cosmological scales. LambdaCDM on the other hand is a cosmology, it describes the CMB, the evolution of structure and the dynamics of galaxies. One of the great success of the CDM paradigm is the NFW profile, it showed that from simple initial conditions profiles would form which explain why galaxies have flattened rotation curves. MOND simply declares it to be so, no explanation as to why that should be. You say there is no dark matter particle? Correct, there is no physical explanation behind MOND arbitrary function either. CDM at least shows how flattened rotation comes about.

The cosmological constant is a another leap it's true but MOND doesn't work at these scales, LambdaCDM's success should not be held against it. It's true faint dwarfs are a sore for concordance cosmology but at these scales baryonic effects are not negligible, it's difficult to say it disproves CDM without making grand assumptions.

 

[Earnest Guest]

Is it though? While it's certainly true LambdaCDM is not perfect but in my opinion it is simply in a different league. MOND sets out to explain just dynamics and I don't think that has been achieved in a consistent way even across all galaxy types never mind moving up to the scale of clusters. But then there is everything else, for example does MOND describe galaxy clustering? No. MOND isn't really a cosmology because it just doesn't work on cosmological scales. LambdaCDM on the other hand is a cosmology, it describes the CMB, the evolution of structure and the dynamics of galaxies. One of the great success of the CDM paradigm is the NFW profile, it showed that from simple initial conditions profiles would form which explain why galaxies have flattened rotation curves. MOND simply declares it to be so, no explanation as to why that should be. You say there is no dark matter particle? Correct, there is no physical explanation behind MOND arbitrary function either. CDM at least shows how flattened rotation comes about.

The cosmological constant is a another leap it's true but MOND doesn't work at these scales, LambdaCDM's success should not be held against it. It's true faint dwarfs are a sore for concordance cosmology but at these scales baryonic effects are not negligible, it's difficult to say it disproves CDM without making grand assumptions.

I hesitant to put myself as a defender of MOND, because I agree with you about it's theoretical deficiencies. However, with regard to an objective view of the track record: MOND doesn't describe EVERY galaxy type out there, but the track record is way, way better than ΛCDM. The best DM models still do not resolve the issue with the central density of galaxies: if dark matter collected like regular matter, then we'd see flattening all the way to the center of the galaxies. We don't.

You mention it's success as a Cosmology. True. Except for the part that 95.5% of the universe is missing. If an economist missed by that amount and he called it a 'bookkeeping error', you'd either throw him in jail or fire him. I have an issue with anyone who can say that ΛCDM is a success with a straight face.

 

[ruarimac]

I hesitant to put myself as a defender of MOND, because I agree with you about it's theoretical deficiencies. However, with regard to an objective view of the track record: MOND doesn't describe EVERY galaxy type out there, but the track record is way, way better than ΛCDM. The best DM models still do not resolve the issue with the central density of galaxies: if dark matter collected like regular matter, then we'd see flattening all the way to the center of the galaxies. We don't.

You mention it's success as a Cosmology. True. Except for the part that 95.5% of the universe is missing. If an economist missed by that amount and he called it a 'bookkeeping error', you'd either throw him in jail or fire him. I have an issue with anyone who can say that ΛCDM is a success with a straight face.

I strongly disagree that it has a better track record. MOND claims the laws of physics are what flattens rotation curves, it just so happens the functions used to fit different types aren't the same. Are we expected to believe that fundamental physics is changing between red and blue galaxies? That's a stretch. It's not that MOND fails to describe every galaxy it's that it can't describe them at once. Furthermore it doesn't seem like it can describe the milky way and other galaxies simultaneously. That's more than a dent.

The reason we don't see flattening all the way to the core has been explained for decades, that is CDM. Because it doesn't have pressure it doesn't collapse as efficiently as normal matter and in the centers of big galaxies the baryons outnumber the dark matter. In modern hydrodynamical simulations it can even be shown that the baryons are so dominant in the core that they modify the central profile of the dark matter halo as compared to a dark matter only simulation.

The fact that the nature of dark matter and dark energy aren't known are physics problems, it's not a problem for the cosmology. It doesn't make any difference to the cosmology if I know what my particles are or not, only that I can calculate the observations I expect. LambdaCDM isn't perfect but it is by far the best we have. Yes it has added parameters like Lambda but that is fine if you're trading up in predictive power.

 

[Earnest Guest]

Yes it has added parameters like Lambda but that is fine if you're trading up in predictive power.

It's not fine if you hold Occam's Razor in any regard. Six is the maximum number of free parameters allowed because you can basically model anything in nature with seven free parameters. As a mathematical exercise, ΛCDM has been a success, but it has told us nothing about our universe.

And with regard to 'predictive power', the list of things the Big Bang failed to predict is much longer than the list of things it actually did predict. The 'power' you mention is in the ability to take a formula with six free parameters (seven if you include the M/L ratio) and retroactively fit any signal you acquire. Did you know that the Λ in ΛCDM was actually put there to counteract Gravity because Einstein thought the universe was static? It was largely ignored until 1998 when we discovered the expansion was accelerating, so we decided to flip the sign on the constant that was supposed to hold things in place, like that was a valid thing to do in science.

And before you mention 'concordance', let me remind you that the Ptolemaic Model could predict the time of sunrise at Easter, the phase of the moon as well as the retrograde of Mars. History has shown that 'Concordance' means nothing when you can go back and alter your 'predictions' with new data. Yet Occam's Razor has an incredible track record at predicting successful theories.

 

[wolram]

https://ned.ipac.caltech.edu/level5/Sept01/Milgrom2/Milgrom3.html

This what the MOND people say

One immediate result of eqs. (1)(2) is that at a large radius around a mass M, the orbital speed on a circular orbit becomes independent of radius. This indeed was a guiding principle in the construction of MOND, which took asymptotic flatness of galaxy rotation curves as an axiom (even though at the time it was not clear how definite, and how universal, this is). Second, this asymptotic rotational speed depends only on the total mass M via V4 = MGa0. This, according to MOND, is the fact underlying the observed Tully-Fisher-type relations, by which the typical (mean) rotational velocity, V, in a disc galaxy is strongly correlated with the total luminosity of the galaxy, L, in a relation of the form L[PLAIN]https://ned.ipac.caltech.edu/level5/New_Gifs/propto.gif[I]V[/I][PLAIN]https://ned.ipac.caltech.edu/level5/New_Gifs/alpha.gif. The power https://ned.ipac.caltech.edu/level5/New_Gifs/alpha.gif is around 3-4, and depends on the wavelength band at which L is measured. The close agreement between this TF relation and the prediction of MOND is encouraging; but, to test MOND more precisely on this count, one would have to bridge properly the mass-asymptotic-velocity MOND relation with the commonly presented luminosity-bulk-velocity TF relation. One should use the luminosity in a band where it is a good representative of the stellar mass, take into account not only the stellar mass, as represented by the luminosity, but also the contribution of gas to the mass, and use the asymptotic velocity, as opposed to other measures of the rotational velocity. It has emerged recently (see [Verheijen 2001] and reference therein) that if one does all this one indeed obtains a tight and accurate relation of the form predicted by MOND. etc, etc.

 

[Jonathan Scott]

I strongly disagree that it has a better track record. MOND claims the laws of physics are what flattens rotation curves, it just so happens the functions used to fit different types aren't the same. Are we expected to believe that fundamental physics is changing between red and blue galaxies? That's a stretch. It's not that MOND fails to describe every galaxy it's that it can't describe them at once. Furthermore it doesn't seem like it can describe the milky way and other galaxies simultaneously. That's more than a dent.

Where did you get that idea? I thought MOND had an excellent success rate in describing every type of galaxy with the same universal acceleration parameter, and that increasingly refined models based on the original rule (same formula but better ways of assessing the mass distribution) were making the fit even better.

I agree that MOND doesn't make sense as a physical theory, only as a formula which happens to match rotation curves for single galaxies.

One particularly unphysical aspect of MOND appears to me to be fixable, which is the arbitrary cut-off of the MOND effect above a threshold acceleration in order not to be detectable in laboratory and solar system effects. This idea never made sense to me, in that for example every star, planet or even atom in a star system at the edge of a galaxy would be subject to higher accelerations from internal gravitational effects within the system, yet somehow overall the system was supposed to be subject to MOND effects if the overall acceleration was small enough. This internal inconsistency could be fixed by changing the cut-off rule, for example (for illustrative purposes) to one where MOND effects only apply when the distance to the source exceeds the scale of the solar system.

 

[ruarimac]

It's not fine if you hold Occam's Razor in any regard. Six is the maximum number of free parameters allowed because you can basically model anything in nature with seven free parameters. As a mathematical exercise, ΛCDM has been a success, but it has told us nothing about our universe.

And with regard to 'predictive power', the list of things the Big Bang failed to predict is much longer than the list of things it actually did predict. The 'power' you mention is in the ability to take a formula with six free parameters (seven if you include the M/L ratio) and retroactively fit any signal you acquire.

And before you mention 'concordance', let me remind you that the Ptolemaic Model could predict the time of sunrise at Easter, the phase of the moon as well as the retrograde of Mars. History has shown that 'Concordance' means nothing when you can go back and alter your 'predictions' with new data. Yet Occam's Razor has an incredible track record at predicting successful theories.

I do but you're misquoting Occam's Razor. It says that of two models of equal explanatory power the simpler one is favored, or that simplicity should only be traded for increased explanatory power. It does not state complex=wrong. Take for example Kepler's model of the solar system of the Tychonian model, many more parameters to describe elliptical orbits but much improved explanatory power. That was just fine according to Occam and so is concordance cosmology.

Six is not the maximum number of free parameters, there is no maximum. Fit two guassians to a spectral line, that's 6 parameters and it describes a simple line. Try as hard as you like but those two guassians will never fit a compensated top hat well. 6 parameters is not enough to fit any 1D profile never mind half a dozen observational probes. 6 parameters is not enough to describe the orbit of a planet around the Sun (6 orbital elements and the solar gravitational constant). The maximum number of parameters is the fewest you can get away with, that is what Occam says.

Fun fact, with the CMB TT power spectrum if you ignore the low-l multipoles you can actually fit the spectrum with just a single free parameter, all others can be taken from other observations (or in the case of ns can be assumed). The free parameter is basically the amplitude. To fit 10 turning points with one free parameter is simply incredible. I'll see if I can dig that out. Concordance cosmology is healthy because it spans a vast range of observations simultaneously.

 

[PeterDonis]

Six is the maximum number of free parameters allowed because you can basically model anything in nature with seven free parameters.

Really? Please show your work.

 

[ruarimac]

Where did you get that idea? I thought MOND had an excellent success rate in describing every type of galaxy with the same universal acceleration parameter, and that increasingly refined models based on the original rule (same formula but better ways of assessing the mass distribution) were making the fit even better.

As far as I've read MOND's current trouble seems to be it's consistency (aside from the whole doesn't work with clusters thing). It is one acceleration parameter but I don't really think it's been shown to be universal. This is the paper on the consistency with the Milky Way and other galaxies.

http://arxiv.org/abs/1505.05181

I've also heard it alleged that the most popular interpolating function doesn't make sense relativistically but it's not my field, I wouldn't claim to know what argument they were making. If anyone more knowledgeable has any idea please feel free to enlighten me.

 

[PeterDonis]

the list of things the Big Bang failed to predict is much longer than the list of things it actually did predict.

Reference, please?

Did you know that the Λ in ΛCDM was actually put there to counteract Gravity because Einstein thought the universe was static?

As a matter of historical fact, this is true. So what? As a matter of physics, $\Lambda$ is not put in ad hoc; it should be there. The Lagrangian formulation of GR makes this obvious.

It was largely ignored until 1998

Reference, please? Please don't forget to account for all those relativity textbooks published between 1915 and 1998 that talk about $\Lambda$ and all the efforts to measure its value.

we discovered the expansion was accelerating, so we decided to flip the sign on the constant that was supposed to hold things in place

Incorrect. The Einstein static universe has a positive $\Lambda$, just like the current $\Lambda$ CDM model.

 

[Chronos]

Theories are always vulnerable to measurement uncertainty, observational bias and creative statistics. Science strives for models that best explain all observations, not some microcosm of observations. This is where MOND suffers - it does very well with certain classes of observations, but, fails the smell test. LCDM on the other hand is plagued by specific apparent exceptions, but, holds up admirably in the face of the totality of observational evidence. Our ancestors used tools like astrology and numerology to explain the natural world. They too enjoyed limited sucess, but, failed the smell test. Theories are tweaked to accommodate anomalous observations, not discarded because of them. GR was not a refutation of Newton's laws, merely an extension.

 

[mfb]

Six is the maximum number of free parameters allowed because you can basically model anything in nature with seven free parameters.

Reformulate the Standard Model of particle physics using just seven free parameters and you'll get a Nobel prize. Even if you can get it down to 15 it would be a huge theoretical success.

And with regard to 'predictive power', the list of things the Big Bang failed to predict is much longer than the list of things it actually did predict.

I'm interested in a reference here as well...

 

[Earnest Guest]

I'm interested in a reference here as well...

The Big Bang failed to predict:

1. A flat universe.
2. A homogenous universe.
3. An isotropic universe.
4. An exponentially expanding universe.
5. A universe of missing antimatter.
6. A universe without magnetic monopoles.

What Big Bang got right:

1. Black Body Radiation
2. Nucleosynthesis (for the most part, still some missing isotopes).

But, really, all you need to do is miss #1 above and you have enough to falsify the model.

 

[PeterDonis]

The Big Bang Failed to predict:

1. A flat universe.
2. A homogenous universe.
3. An isotropic universe.
4. An exponentially expanding universe.
5. A universe of missing antimatter.

All of these things are consistent with the Big Bang, so they don't rule it out. (In fact, #2 and #3 are assumptions of the Big Bang theory, so of course the theory didn't predict them--it took them as given.)

all you need to do is miss #1 above and you have enough to falsify the model.

No, to use #1 to falsify the Big Bang, you would have to show that it is inconsistent with the Big Bang, not that the Big Bang failed to predict it. No scientific theory ever invented has predicted every single future observation within its domain of validity.

Also, can you show me some alternative cosmological theory that did predict all of the above?

 

[Earnest Guest]

All of these things are consistent with the Big Bang, so they don't rule it out. (In fact, #2 and #3 are assumptions of the Big Bang theory, so of course the theory didn't predict them--it took them as given.)

1. The Big Bang predicts an infinite number of possible geometries. It turns out that the universe is completely flat. The odds of getting a completely flat universe out of BB are infinitesimally small. So while it doesn't strictly invalidate it, the odds against it being right (creating a flat universe) are so impossibly small that any reasonable person would reject it if they understood how probabilities work.
2. GR predicts an event horizon beyond which information can't be shared. BB predicts that new areas of space that become visible to us will have different densities and temperatures than other regions of space. However we see the same temperature and density in every direction we look. This is definitely a failure of BB.

 

[mfb]

The Big Bang failed to predict:

It also "failed" to predict the existence of physicsforums.
Is that a flaw of the big bang model?

• The Big Bang predicts an infinite number of possible geometries.

It does not. It is consistent with any curvature, so what? It is also consistent with the nonexistence of physicsforums.

BB predicts that new areas of space that become visible to us will have different densities and temperatures than other regions of space.

It does not (apart from random fluctuations, which we see). Inflation is a possible way larger regions can have a causal connection.

 

[Earnest Guest]

It also "failed" to predict the existence of physicsforums.
Is that a flaw of the big bang model?
It does not.

A good theory predicts things. When it fails to predict new things, it's not a very good theory. My point is, the alleged 'predictive' powers of ΛCDM are entirely in it's ability to look backwards.

 

[mfb]

A good theory predicts things.

There are tons of predictions of the big bang, it's not just the CMB and nucleosynthesis. As usual, Wikipedia has a list.
That does not mean every single observation we ever made (including the existence of physicsforums) has to be predicted by it.

 

[PeterDonis]

The Big Bang predicts an infinite number of possible geometries.

Not if inflation is included. With inflation, the Big Bang predicts a spatially flat universe.

Without inflation, the Big Bang makes no definite prediction about the spatial geometry.

GR predicts an event horizon beyond which information can't be shared.

Not as far as cosmology is concerned. The FRW spacetimes used in cosmology have no event horizons. The ones with nonzero dark energy have cosmological horizons, but that's not the same thing, despite certain similarities.

BB predicts that new areas of space that become visible to us will have different densities and temperatures than other regions of space.

Reference, please?

 

[PeterDonis]

A good theory predicts things. When it fails to predict new things, it's not a very good theory. My point is, the alleged 'predictive' powers of ΛCDM are entirely in it's ability to look backwards.

You have yet to give references for any of your claims about what the BB theory does or does not predict. Either give valid references or stop making these claims.

 

[rootone]

A good theory predicts things. When it fails to predict new things, it's not a very good theory. My point is, the alleged 'predictive' powers of ΛCDM are entirely in it's ability to look backwards.

BBT is a theory concerning the primordial conditions of the Universe and it's subsequent evolution.
As such 'looking backwards' is fundamental to its purpose, it's not something which is undesirable in a cosmological theory.
I think the main point here is, that at present there is no seriously competing theory which does predict CMB and Nucleosynthesis as BBT does.

 

[Chronos]

The goal of science is to identify best fit models - but, perfect is impossible [re: see uncertainty principle and statistical mechanics]. BBT and LCDM are our current best fit models in the opinion of an overwhelming majority of scentists. Can the majority be wrong? Of course, but, something more substantial than finger pointing is necessary to be pursuasive on PF. On the other hand, that has never dissuaded other people from preferring logic, ghosts, aliens or traditional lore over good science.

 

[Jonathan Scott]

I've also heard it alleged that the most popular interpolating function doesn't make sense relativistically but it's not my field, I wouldn't claim to know what argument they were making. If anyone more knowledgeable has any idea please feel free to enlighten me.

The interpolating function is a clearly unphysical arbitrary rule which has the sole purpose of explaining why MOND effects aren't seen in laboratory experiments and solar system results. In standard MOND, this provides a "cut-off" switch based on an unspecified function of some sort of "absolute" acceleration, which is way beyond unsatisfactory for multiple reasons.

The MOND fit to experimental data for galactic rotation curves does not require any interpolating function; one can just add the MOND acceleration to the Newtonian acceleration. In this area, MOND provides a ridiculously good predictive formula, and I had the impression (from studies a few years ago) that similar values of the acceleration parameter worked for the full range of galaxy types.

I'm not up to date on attempts to apply MOND to globular clusters and similar within the Milky Way galaxy (I need to look at those referenced papers). Unlike Newtonian theory, MOND effects are horribly non-linear and don't add up like vectors, so I wouldn't be surprised if the only case that works really well is for galaxies, where there is a single central concentration of mass being considered.

It seems to me that MOND, like the Tully-Fisher relation, should be treated for now as an empirical formula that seems to provide a surprisingly good fit for a specific set of experimental results, suggesting that there may be some sort of physical cause behind it. However, so far I'm not aware of any satisfactory theory (whether involving changes to gravity theory or dark matter) that explains it.

 

[RUTA]

However, so far I'm not aware of any satisfactory theory (whether involving changes to gravity theory or dark matter) that explains it.

There is a corresponding relativistic action for MOND produced by Bekenstein. Sean Carroll describes it at the 35:13 mark of

 

[Jonathan Scott]

There is a corresponding relativistic action for MOND produced by Bekenstein. Sean Carroll describes it at the 35:13 mark

Thanks for the reference. Carroll's presentation says "(Lorentz invariance is spontaneously violated, but okay)." To my way of thinking, that disqualifies TeVeS from being "satisfactory", although it is clearly much better than the original MOND "theory". Also, I haven't looked at TeVeS recently, but I think it had a set of parameters to be adjusted, and it was proving difficult to find a set which weren't already ruled out by experiment.

 

[RUTA]

I finally delved into the dark matter problem this summer and I would say modified gravity approaches are as reasonable as looking for non-baryonic dark matter candidates. Observations implying dark matter were first made over 80 years ago and we still don't have any confirmed non-baryonic candidates. LCDM works well for cosmology by simply assuming something exists (no one knows what), but it doesn't get the mass distribution at the galactic level correct (e.g., the core-cusp problem). The dark matter problem is very difficult, e.g., after listing its attributes, i.e., dark, cold, abundant, stable, and dissipationless, Sean Carroll concludes, “So should we be surprised that we live in a universe full of dark matter? I’m going to say: yes.” Here is a more general view of MOND recently published in PRD. Brownstein showed MOND beat metric-skew-tensor-gravity (MSTG) for galaxy rotation curves while MSTG beat MOND for the mass profiles of X-ray clusters. I like the way MSTG handles the famous Bullet Cluster, originally touted as "direct empirical proof of the existence of dark matter,'' without non-baryonic dark matter. Anyway, I think both approaches (non-baryonic dark matter and modified gravity) are reasonable at this point.

 

[RUTA]

Thanks for the reference. Carroll's presentation says "(Lorentz invariance is spontaneously violated, but okay)." To my way of thinking, that disqualifies TeVeS from being "satisfactory", although it is clearly much better than the original MOND "theory". Also, I haven't looked at TeVeS recently, but I think it had a set of parameters to be adjusted, and it was proving difficult to find a set which weren't already ruled out by experiment.

I prefer MSTG, where the gravitational "constant" varies in strength, to MOND, but I'd rather see an alternative to both that doesn't have any "gaps." By that I mean MOND has to invoke too much baryonic dark matter at the scale of galactic clusters where MSTG works well, while MSTG fits for galactic rotation curves aren't as good as MOND. Both have been applied at the cosmological level (MOND and MSTG), but I haven't seen them fit the CMB temperature power spectrum. I'm not sure they have to do that per se if they can account for Fig 1 here. In that case, the rest of the fit should follow.