Workshop on tests of gravity in Case Western – day 2: aether and modified gravity
ASTRO — By Dmitry Podolsky on May 29, 2009 at 9:54 pm
Save This Post as PDF
Let me finally briefly review the reminder of the second day of the workshop.
Justin Khoury (whom I knew from Perimeter years and who is in Penn now) gave the first talk afternoon – titled “observational hints of IR modified gravity”. His talk followed Nima’s, and the latter almost completely blew me away, so I was somewhat unfocused during Justin’s presentation. Yet, I was able to capture that as such observational hints he wanted to present local bulk flow of matter within 50 
MPc scale, excess power in Lyman 
(about 30%) and small scale CMB anomalies (which I would hardly call anomalies due to lack of statistics there). According to him, all this shows that gravity should be stronger at larger scales – and that’s exactly what many models of modified gravity predict.
Next, Stacy McGaugh has talked about MOND. I should really confess that my attitude towards MOND was always somewhat arrogant – I had no idea why people might be interested in such an ugly theory. This time I got it, I guess: in order to fit all galaxy rotation curves, you only need to pick one and fit the MOND parameter (acceleration 
) – and you’ll automatically fit all others, that is, an almost infinite data set turns out to be one-parametric. Well, I guess, choosing the density of cold dark matter 
will do the job as well.
Nima Arkani-Hamed has asked him about the bullet cluster and Stacy’s answer was like “I find it somewhat fishy that all proponents of cold dark matter model always use this example to defend their models”. Well, that’s the point – they use this argument, because it works.
The first talk after lunch (by Ted Jacobson) was devoted to the discussion of Einstein aether theory. What is the theory possessing such a scary name that would probably create an itchy feeling in guts of every 1 year physics student?
Consider a theory with the action
,
where
,
is a time-like unit vector (since 
is really lagrangian multiplier). The name of the theory involves the word “aether”, since a “4-velocity” vector (velocity of aether) introduces a preferred direction in spacetime (Lorentz transformations along and perpendicular to it are different). It therefore belongs to the class of exciting deeply flawed theories according to Nima’s classification, and not surprisingly the theory messes with BH thermodynamics – no unique Hawking temperature or BH entropy has been found so far and generalized second law seems to be violated (that is, you can construct perpetuum mobile in the universe described by this theory).
Since Ted Jacobson has mostly discussed classical properties of the theory, Nima Arkani-Hamed immediately pointed out that if one turns to the quantum level, the theory above features a non-perturbative mass scale – since it looks similar to non-linear sigma model.
My personal comment to him: actually, this is true that lagrangian multiplier acquires a non-zero VEV due to non–perturbative effects, however, its fluctuations are only suppressed if the vector has many components (namely, inifinite number of them), while in this case 
, and lagrangian multiplier fluctuates strongly.
Levon Pogosyan has talked about CMB and LSS observations – are we able to determine from them whether gravity is modified or not? As it turns out, our data are particularly sensitive to scale dependent modifications of gravity, and future surveys such as LSST might give us more information that just effective EOS for dark energy 
.
Finally, Mark Wyman has discussed N-body numerical simulations of DGP. The result he presented is well understandable: since DGP means stronger gravity at larger distances (and therefore earlier times), in DGP large scale structure is formed faster and its features are stronger than in GR (see the Fig. below)
9 Comments
Dear Dmitry,
are people discussing interesting alternatives to CDM, such as that
http://xxx.lanl.gov/abs/hep-th/0701169 in the conference?
Please, let us know!
Thanks.
Dear Maximo,
I am afraid, a blatant advertisement of your own paper (like the one above) is not really acceptable on NEQNET
On the other hand, a smarter advertisement – in the form of the question “Dear Dmitry, is it possible to write a guest post about … on NEQNET?” – is acceptable. The answer is not guaranteed to be positive, though.
Cheers,
Dmitry.
Dear Dmitry,
it is not Maximo here. Neither somebody doing “advertisement” for him. But, naturally, I think you could have checked it, since IP locations didn?t match, right?
I am just sending a question (and related paper) because, after having read some posts in your blog, from people “explaining” CDM using “exotics”, I personally believe readers would be awared about more conservative (yet succesfull) approaches!
Cheers,
Rafael.
Dear Rafael,
if the author of the comment wasn’t Maximo, I apologize
Still, you write “more conservative (yet successful)” – which sounds again as a blatant advertisement to me
In which respect, a theory featuring zeros of 
can be considered more conservative? And in which respect it is successful?
Cheers,
Dmitry.
Dear Dmitry,
Yes, we may disagree about my writting style
. However, I prefer to see that as an “advertisement” (If you wish) to the idea of understanding CDM by using GR and “Occam?s Razor”, i.e. without exotics.
Ok, back to your comments.
Well, by reading the aforementioned article one sees that the author uses Einstein?s GR and considers some additional (g->0) limits. That limit was first observed by Eddington (e.g. see http://xxx.lanl.gov/abs/0807.5088 and Refs), but, which was not really exploited in astrophysics.
So, I consider it “conservative (yet successful)” since:
(1) The due profiles for circular motions of galaxies are obtained (without dark matter).
(2) Resulting FRW equations contain an additional matter sector as in CDM models, due to g->0.
(3) The used formalism is “just” Palatini action for gravity (with a cosmological constant) plus well-known duality transformations (see Refs.).
So, to best of my knowledge, it seems a very simple “CDM without CDM” approach to experimental puzzles… isn?t it?
Cheers,
Rafael.
Dear Rafael,
Palatini formalism, tetrad formulation etc. etc. is only equivalent to Einstein-Hilbert GR if the metric is not degenerate. That’s however exactly the limit that Banados wants to discuss. In other words, I don’t quite see why you say this approach is conservative. He does use some kind of a theory featuring spin 2 field, but it’s not a GR in the sense we know it.
Second, I don’t quite see any fit for galaxy rotation curves in his papers. All he does is writing a spherically symmetric solution for his theory and showing that it has a constant asymptotics at r=0 which is determined by an integration constant. What defines the value of this integration constant? I am supposed to find a fit for any galaxy separately? If so, ugly MOND does a better job – you need just one parameter to fit all curves, for all spiral galaxies you see on the sky. In other words, I am yet to see why you call his approach “successful”.
But if you have an answer to these two questions, sure, let us discuss – the idea is so crazy that it seems interesting.
Cheers,
Dmitry.
Dear Dmitry,
1) By reading second reference I cited above — sorry by “advertisements” again
— we see (Eq.4) that the “kind of a theory featuring spin 2 field” is just GR + Eddington term. Well, (Eq.4) reduces to EH-action for gravity with renormalized G-constant (by dualities): (Eq.5 to 10), and comments therein.
2) “What defines the value of this integration constant? I am supposed to find a fit for any galaxy separately?”. Yes, sure, since it is just an integration constant (as \Lambda is!!). I see no great problem on that, since CDM models have to postulate a hole distribution of dark matter to fit galaxy curves that “constant”. On the other hand, “Ugly MOND” could not give you some sensible CDM for FRW-cosmology!
However, if one want to know how G-runs depending just on the observable matter distribution, I agree with you that a more “elaborated” theory is needed. This means, a coupling between EH and Eddington pieces of action is needed. That is not a great problem, since available options are constrained.
For instance, you can check new developments following (0807.5088) on arxiv that show so-called “EBI-GR” applied on cosmology (e.g. arXiv:0801.4103 or arXiv:0811.1272). But, this new “developments” are someway “less economical” (in Occam’s sense).
Cheers,
Rafael
Hi Rafael,
Actually, he acknowledges the fact that actions are not completely equivalent – see the end of the page 4, this is exactly what I said above. He tries to go around saying that he is only interested in non-degenerate limit. A paragraph later he is talking about phase transition between g_{\mu\nu}=0 phase and normal EH phase – so he apparently does need degenerate limit. Sorry, I am not buying equivalence you are talking about.
No,
is not an integration constant, it is cosmological constant 
(which he introduces in his model by hands by the way – since “dualities” do not work for 
case).
Well, if you want to consider solutions without spherical symmetry (unfortunately, you have to – spiral galaxies are not really spherical) in Banados’ model, then guess what happens?
By the way, how do you explain the physics of bullet cluster in this model?
Cheers,
Dmitry.
a brief comment about MOND vs CDM. Note that CDM does not explain
why the need for dark matter in all galaxies is felt for accelerations less than 10^-8 cm/sec^2.
Moreover CDM does not explain the Tully-Fisher relation (which basically states that the visible mass of our galaxy goes as fourth power of asymptotic rotation velocities of galaxies
If most of our galaxy is made up of dark matter, why should there be such as strong correlation between the baryonic component and
peak values of rotation velocities of spiral galaxies? For more
details on regularities in cosmic structure see this paper
by Sanders
(see in part. figs 2 and 4). It is going to be awfully hard for ordinary CDM to explain this.
Having said that however MOND does not mean that GR needs a modification. It could be due to backreaction as argued by David Wiltshire or it could be dark matter is made up of dipolar particles as argued by Luc Blanchet.
Trackback responses to this post