Wednesday, March 14, 2018

More Trouble For Dark Matter

An analysis of ambient radio wave activity at particular frequencies provides a window into the nature and temperature of the universe very early in its existence. What does it show?
The profile is largely consistent with expectations for the 21-centimetre signal induced by early stars; however, the best-fitting amplitude of the profile is more than a factor of two greater than the largest predictions. 
This discrepancy suggests that either the primordial gas was much colder than expected or the background radiation temperature was hotter than expected. Astrophysical phenomena (such as radiation from stars and stellar remnants) are unlikely to account for this discrepancy; of the proposed extensions to the standard model of cosmology and particle physics, only cooling of the gas as a result of interactions between dark matter and baryons seems to explain the observed amplitude. The low-frequency edge of the observed profile indicates that stars existed and had produced a background of Lyman-α photons by 180 million years after the Big Bang. The high-frequency edge indicates that the gas was heated to above the radiation temperature less than 100 million years later.
Of course, the (literally) defining property of dark matter in the standard model of cosmology (called ΛCDM for a cosmological constant with "cold dark matter") is that dark matter is completely or almost collisionless with respect to ordinary baryonic matter . . . . dots that this abstract refrains from connecting.

What correctly predicts what is observed?

A universe with no dark matter where the matter in the universe is entirely made up of baryonic matter (i.e. a modified gravity hypothesis), as explained in a readable two page (exclusive of references) paper supplementing the paper in the journal Nature linked above, that accepts virtually all of its data, theoretical framework and equations. It merely connects the dots that the paper in Nature didn't. They money language:
Equation 1 then predicts a maximum absorption of  
T21, max = −0.24 K for fb = 0.16; (2)  
T21, max = −0.60 K for fb = 1. (3)  
The observed value is T21 ≈ −0.5 K (Bowman et al. 2018), and appears from their Fig. 2 to be closer to −0.55 K. Such a strong signal is expected in a baryonic universe, but is unobtainable in ΛCDM. 
Figure 1 of the supplemental paper shows that a naive, unmodified ΛCDM prediction (as opposed to the maximal possible variation of it) is T21 ≈ −0.05 K, about 9% of the observed value, as opposed to 92% of the value predicted in a purely baryonic scenario (the exact margin of error isn't stated).

Once again, modified gravity theorists have correctly predicted the results of future experimental observations and dark matter theories made the wrong prediction, on a matter independent of previous predictions that have been confirmed.

McGaugh's linked supplement paper also makes some further predictions that new data can test:
1. Strong 21cm absorption will also be observed during the dark ages (z > 30).  
2. The 21cm power spectrum will show pronounced baryonic features.  
3. Large galaxies and the cosmic web emerge earlier than anticipated in ΛCDM (Sanders 1998; McGaugh 2015).  
The first two predictions stem simply from a universe made of baryons. Only the third prediction is model-specific; some hints of large early structures already exist (e.g., Steinhardt et al. 2016; Franck & McGaugh 2016). This implies that structure grows nonlinearly, erasing baryonic features at late times through mode mixing (McGaugh 1999).

UPDATE March 16, 2018: A completely separate paper constrains dark matter models via metal enrichment in very old galaxies.

UPDATE: March 20, 2018: Virtually all of the scatter in the radial acceleration relation observed in galaxies can be explained by measurement error - the residual scatter after estimating the effects of various forms of measurement error is about 3%.

These tweets from Federico Lelli explain the paper's significance:
In and we show that baryons and dynamics are tightly linked in galaxies at a *local* level: the observed acceleration (from the gas rotation curve) correlates with that expected from the distribution of baryons (using Newton's law).
At high accelerations gobs=gbar because baryons fully dominate the dynamics. These data come from the central parts of massive galaxies where there is no need for dark matter. At low accelerations, the data deviates from the 1:1 line. That's the classic indication of dark matter! 
We were expecting something like this! @DudeDarkmatter did a similar work back in 2004: … What really shocked us was the tightness of the relation! The deviations from the mean relation (aka "scatter") are of the order of 30%! 
This scatter is absurdly small for the (poor) standards of galactic astronomy. So we did a simple math exercise: let's propagate the errors on the measurements and compare the resulting expected scatter from errors (black) with the observed scatter (red). The two are consistent!
This means that the relation is consistent with NO intrinsic scatter. The width of the relation can be fully explained by observational errors. In general this happens only when you deal with Laws of Nature, like Kepler's laws. This is a big deal, so we further looked into it. 
The problem is that errors in Astronomy are a mess. We don't have complete control on them as Physicists can have with their experiments in their labs. Our lab is the Universe and the Universe doesn't care about what we wanna do. And here comes Pengfei's paper. 
There are three main sources of error on gobs and gbar: the galaxy distance, the galaxy inclination, and the conversion from stellar light to stellar mass. If you got them a bit wrong, the galaxy will shift perpendicularly or horizontally from the mean relation, adding scatter. 
The fact that the observed scatter is only 30% means that we did a pretty good job in estimating these quantities. But now we want to know whether this 30% can completely disappear when the errors are taken into account. 
To do this, Pengfei used Bayesian statistics and a technique called Markov Chain Monte Carlo (MCMC). Essentially you can fit the mean relation to individual galaxies, allowing the light-to-mass conversion, distance, and inclination to vary a little, within the observed errors. 
This is an example of how good the fits can be. They are not all this good. But if we measure the mean deviation from the fits for all galaxies, the resulting number is ridiculously small: 13%. We have a recipe that explains rotation curves better than 13% using only baryons!
When we optimize distance, inclination, and mass-to-light conversion, the radial acceleration relation becomes extremely tight. This didn't necessarily have to happen because we are varying *global* quantities while looking at a *local* relation, connecting different galaxy radii. 
If you are still following me, you may wonder what drives the remaining 13% scatter. Well, part of it could be intrinsic, but most of it must surely come from errors on the rotation velocities (due to non-circular motions and other nuisances) which are indeed of the order of 10%. 
To summarize: we have a relation which is very tight for astronomical standards. The deviations from the mean can all be explained by measurement errors. The relation is consistent with no intrinsic scatter. So is this a Law of Nature? Or just another galaxy scaling relation? 
Ops! I forgot to add the key plot. Here you see the radial acceleration relation using best-fit distance, inclination, and mass-to-light ratio. The tiny residuals need two tiny Gaussians, but we think we understand why this happens. It's likely how velocity errors are estimated.

The paper's abstract and citation are as follows:
Galaxies follow a tight radial acceleration relation (RAR): the acceleration observed at every radius correlates with that expected from the distribution of baryons. We use the Markov Chain Monte Carlo method to fit the mean RAR to 175 individual galaxies in the SPARC database, marginalizing over stellar mass-to-light ratio (Υ⋆), galaxy distance, and disk inclination. Acceptable fits with astrophysically reasonable parameters are found for the vast majority of galaxies. 
The residuals around these fits have an rms scatter of only 0.057 dex (∼13%). This is in agreement with the predictions of modified Newtonian dynamics (MOND). 
We further consider a generalized version of the RAR that, unlike MOND, permits galaxy-to-galaxy variation in the critical acceleration scale. The fits are not improved with this additional freedom: there is no credible indication of variation in the critical acceleration scale. The data are consistent with the action of a single effective force law. The apparent universality of the acceleration scale and the small residual scatter are key to understanding galaxies.
Pengfei Li, Federico Lelli, Stacy McGaugh, James Schormbert, "Fitting the Radial Acceleration Relation to Individual SPARC Galaxies" arXiv (February 28, 2018).

This successfully replicates and improves upon a finding from a December 14, 2015 paper.

And, don't forget the stunning case of the Crater II galaxy where MOND accurately predicted an outlier result that could not be obtained with a dark matter model.

Crater II isn't completely independent of the low scatter result but tests the same thing in very different domains of applicability in different ways. And the other empirical datasets discussed above are completely independent of these results.

A 2016 paper with overlapping authorship with this one shows that observation is not consistent with dark matter that does not interact with baryonic matter is not consistent with observation, a finding that another independent study has replicated.
Cosmological N-body simulations predict dark matter (DM) haloes with steep central cusps (e.g. NFW, Navarro et al. 1996). This contradicts observations of gas kinematics in low-mass galaxies that imply the existence of shallow DM cores. 
Baryonic processes such as adiabatic contraction and gas outflows can, in principle, alter the initial DM density profile, yet their relative contributions to the halo transformation remain uncertain. 
Recent high resolution, cosmological hydrodynamic simulations (Di Cintio et al. 2014, DC14) predict that inner density profiles depend systematically on the ratio of stellar to DM mass (M∗/Mhalo). Using a Markov Chain Monte Carlo approach, we test the NFW and the M∗/Mhalo-dependent DC14 halo models against a sample of 147 galaxy rotation curves from the new Spitzer Photometry and Accurate Rotation Curves (SPARC) data set. These galaxies all have extended H{\small I} rotation curves from radio interferometry as well as accurate stellar mass density profiles from near-infrared photometry. 
The DC14 halo profile provides markedly better fits to the data compared to the NFW profile. Unlike NFW, the DC14 halo parameters found in our rotation curve fits naturally fall within two standard deviations of the mass-concentration relation predicted by ΛCDM and the stellar mass-halo mass relation inferred from abundance matching with few outliers. 
Halo profiles modified by baryonic processes are therefore more consistent with expectations from Λ cold dark matter (ΛCDM) cosmology and provide better fits to galaxy rotation curves across a wide range of galaxy properties than do halo models that neglect baryonic physics. Our results offer a solution to the decade long cusp-core discrepancy.


neo said...

are you familiar with Lee Smolin's paper that MOND is a QG effect?

andrew said...

Sounds vaguely familiar. I think that it probably is a QG effect in any case. I'll have to look for it.

neo said...

I had in mind



what about other evidence for DM, for sean carroll on youtube CMB BAO, weak gravitational lensing large scale structure clinches case for DM, which modified gravity cannot reproduce

i suspect its both DM + MOND

andrew said...

Relevant: Big black holes are present too early.

The notion that you can't get CMB BAO and large scale structure from modified gravity is bullshit. It's been done a couple of different ways at least. Any theory that mimics DM in galaxies is also going to be pretty similar to DM from a cosmology perspective.

RE Smolin The notion that MOND could be a renormalization effect that is prominent in the limit of low acceleration rather than high acceleration is interesting, although I am somewhat skeptical of some of the axioms chosen. But, this passage, which reflects the same strategy of Deur in methodology is particularly striking: "The observation is that all the cases we have studied situations where variances in time can be neglected, which hence involve static configurations such as uniformly accelerated observers or circular motion. In these situations there appear applications of equilibrium thermodynamics, at
the classical and semiclassical level. These applications give rise to the Einstein equations, as was proposed in [27]. But what if we extend our analysis to describe strongly time dependent situations? Then we will have to extend our use of thermodynamics to nonequilibrium thermodynamics."

Re 1704.00780 The identification of DM v. MOND is standard and basically right. The focus on the horizon distance has been noted before, although I'm a bit skeptical of it. Footnote 2 notes that lots of people are looking at this: "The idea that MOND is an expression of quantum gravity has been considered earlier from various points of view by Milgrom[21], van Putten[22], Verlinde[23], Hossenfelder[24], Woodard[25], Modesto and Randano[26], Minic et al[27], Hendi and Sheykhi[28], Mike McCulloch[29] and others[30], including the
author[31, 32]. The argument below is, I believe, on the whole, novel, but in places it overlaps some of
these discussions" I would add Moffat and Deur and even Lubos. I'm not a fan of the inertial v. gravitational mass formulation.

Overall, right or wrong, Smolin is headed in the right direction as are the fellow investigators he identifies and that is very good and gives us hope that the sociological problems of the field may be overcome.

andrew said...


"i suspect its both DM + MOND"

There may be "dim matter", i.e. low luminosity ordinary baryonic matter and black holes and neutrinos that nibble a bit at the edges especially MACHOs, intermediate sized black holes, interstellar gas, dust, and neutrinos. But, I don't think that their contribution is ultimately going to be important and I don't think that there is some new kind of non-baryonic matter out there to be discovered. We don't need it to explain what we see, so it probably doesn't exist.

andrew said...

The radial acceleration relation still has too little scatter to make sense in a CDM world.

andrew said...

The paper that is based on is linked here:

and keep in mind that a fair amount of the scatter is attributable to experimental measurement uncertainty.

andrew said...

Who knew? I did block the first Smolin paper you reference.

neo said...

Sean Carroll and Ethan Siegal are both astrophysicists and are clear that modified gravity might get the first and second peak, but not the third peak in BAO, and that only dark matter 5 times the amount of baryonic matter can explain it, and not any version of modified gravity. i can provide references if you want

there's also weak gravitational lensing and Modified gravity doesn't produce large scale structure formation.

I do think its MOND and dark matter with dark matter being black holes.

cold dark matter might be dead but there's plenty of other dark matter theories.
fuzzy dark matter is still allowed by RAR

andrew said...

Modified gravity theories can produce structure faster than DM and that is what new observations are showing. Carroll and Siegal are looking at straw man arguments. Weak gravitational lensing doesn't obviously distinguish the two categories either, I can't imagine why they would have to.

neo said...

Weak gravitational lensing but no visible matter implies dark matter causing it.

and BAO cannot be reproduced by modified gravity.

other than that I'm on the MOND train ;) but even Stacy McGaugh admits MOND doesn't explain certain observations like galaxy clusters


Only Dark Matter (And Not Modified Gravity) Can Explain The Universe

andrew said...

"Weak gravitational lensing but no visible matter implies dark matter causing it and BAO cannot be reproduced by modified gravity."

Modified gravity can definitely give rise to both of these phenomena. And, there are modified gravity theories other than the toy model of a theory called "MOND" (e.g. MOG, TeVeS, Deur's theory, etc.). I can't follow the links due to the . . . . without a hyperlink in the comment.

Ask Ethan 94 commits the sin of comparing a non-relativistic toy model to GR rather than TeVeS which is Bekenstein's relativistic generalization of MOND or MOG by Moffat that do basically everything that GR does in the strong field limit. MOG is better with clusters which Deur's work also manages well.

Starts with a bang is flat wrong about the Bullet Cluster (which MOG and other modified gravity theories do explain) because the Bullet Cluster is actually inconsistent with dark matter.

Ethan Siegel is very emphatic, but he is simply wrong. And, DM has been pretty much ruled out entirely as a theory.

neo said...

doesn't neutron gravitational wave observation rule out MOG TEVS etc. i can provide links if u want. it is only compatible with standard GR

Ethan Siegel is an astrophysicist

Sean Caroll also makes the same points

andrew said...

First, for reference: Ruling out dark matter cooling in EDGES experiment almost completely:

I'm aware of the papers you reference and probably have them linked in prior posts somewhere at this blog. They definitely don't rule out a Deur's style mechanism where the self-interaction of a massless graviton provide the same outcome and the strong field, spherically symmetric gravitational interactions are indistinguishable (at least with anything remotely close to current instrumentation) because the self-interaction effect is much smaller than direct interaction except in extremely weak fields that are also not approximately spherically symmetric.

The claim that it rules out MOG and TEVES is that both of those theories have a scalar, vector and tensor component of a gravitational field, and if those propagate at different rates, then it is incompatible with the neutrino wave observations. In particular, they argue that the scalar component should be slower in propagation than the tensor component. It is not only compatible with Standard GR however and the papers in question don't say so. They say that it is incompatible with scalar-tensor and scalar-vector-tensor fields, but not, for example, a purely scalar or a purely tensor graviton.

I am not entirely convinced that the scalar, vector and tensor fields have to propagate at different rates, or to put it differently, that each of the three gravitational bosons associated with the respective fields in those theories can't all be massless. But, I haven't had time to look more closely into that claim.

Basically, the neutron star merger data proves that (1) all gravitation transmitting bosons must have equal mass and (2) that they must not have mass significantly greater than neutrinos (which travel as a speed so close to the speed of light that you can't distinguish them from massless particles with current instrumentation even in this kind of extreme scale situation, per distant supernova observations).