New research both proves and disproves ... wait, what?

[unixronin]

Here’s an interesting one for the books.  In the continuing debate over the widely-accepted cold-dark-matter theory, a new study of the dwarf galazies surrounding the Milky Way has found that their arrangement appears to simultaneously require and refute the presence of dark matter.

More exactly, the orbital dynamics of the dwarf galaxies are consistent with what would be predicted by the cold-dark-matter theory ... but their arrangement indicates that they were formed by collision with other galaxies, a mechanism that — according to dark matter theory’s predictions of the properties of dark matter — precludes them containing any dark matter.

“The fragments produced by such an event can form rotating dwarf galaxies,” Metz said.  But there is an interesting catch to this crash theory, “theoretical calculations tell us that the satellites created cannot contain any dark matter.”  This assumption, however, stands in contradiction to another observation.  “The stars in the satellites we have observed are moving much faster than predicted by the Gravitational Law.  If classical physics holds this can only be attributed to the presence of dark matter.”

Doubt has been shed on the cold dark matter theory before.  This is the first time, however, that an observation has been found for which assuming dark matter creates a paradox.  If the dark matter theory is assumed to be correct, then dark matter is the only explanation for their orbital dynamics, yet their arrangement precludes the possibility of dark matter being responsible for their orbital dynamics.

How to resolve this?

Or one must assume that some basic fundamental principles of physics have hitherto been incorrectly understood.  “The only solution would be to reject Newton’s classical theory of gravitation,” adds Kroupa.  “We probably live in a non-Newton universe.  If this is true, then our observations could be explained without dark matter.”  Such approaches are finding support amongst other research teams in Europe, too.

Mordechai Milgrom’s MOND — Modified Newtonian Dynamics — theory has long been proposed as an alternate explanation for the observations used as evidence for cold dark matter.  But there are hard scientific reasons why MOND cannot be correct if the rest of our understanding of the Universe is correct — specifically, MOND is not consistent, and cannot be made consistent, with relativity.  A greatly more complex approach to the problem is Jacob Bekenstein’s TeVeS, tensor-vector-scalar gravity.  TeVeS, unlike MOND, is consistent with relativity; but it has been argued that TeVeS cannot simultaneously account for both galactic dynamics and gravitational lensing.  Both TeVeS and dark-matter theory are also unable to completely explain the Bullet Cluster observations without postulating additional factors; TeVeS requires the presence of dark matter as well to fully explain the observations, while dark matter theorists have found it necessary to invoke not only both dark matter and dark energy but a possible fifth basic force in order to fully explain them.  A related theory, STVG (scalar-tensor-vector gravity), takes a slightly different approach from TeVeS, and is able to successfully explain galaxy rotation curves, galaxy cluster mass profiles, gravitational lensing, the Bullet Cluster observations, and the accelerating expansion of the Universe without requiring the presence of either dark matter or dark energy.  Another very recent theory (to which I’m unable to find a reference right now), based on M-theory, posits that gravitons are weakly bound to the brane and can drift off of it and diffuse away into the bulk, resulting in a net gravitational force that follows the Newtonian inverse-square law at “normal” distances, is slightly stronger at huge distances as required by MOND and TeVeS to produce the effects which CDM theory explains via dark matter, and on truly vast scales becomes weaker than inverse-square again, in order to explain universal expansion without dark energy.

We don’t know yet which, if any, of these theories may be correct, and new theories are being proposed all the time.  However, with this finding, particularly if it can be confirmed, the holes in cold dark matter theory are beginning to become larger and more evident.

Comments

[smandal.livejournal.com]

I've asked you before, but you haven't gotten back to me: why should we doubt CDM? It is consistent with all the observations it is meant to explain (galactic dynamics, structure formation, CMB). Dark energy is needed for largely independent reasons (large scale structure, supernovae redshifts; CMB to a lesser extent).

Specifically, for the Bullet Cluster it's a slam dunk: gravitational lensing shows a cold, virialized mass in the center, and X-rays show the hot-gas shockwave, and these two are dynamically consistent. If you know of a reference that argues otherwise, please link me!

The far-reaching goal is a unified theory, but these unified theories are generally a mess as you have described. In addition to being inconsistent with observations, they have theoretical problems like ghosts and fine-tuning. I am not familiar with STVG, but I'll be very impressed if it doesn't have the same problems. It's not surprising the Wikipedia page would only highlight the successes, since it's probably maintained by stakeholder (perhaps even Moffat himself).

As for the business of satellites, you'll notice that the actual paper is much more boring than the news article. The infall model doesn't work (I don't know the details), but they point to an alternative model that need not contradict CDM: tidal dwarf galaxies, formed from fragments due to collisions between larger galaxies.

[smandal.livejournal.com]

To be specific, in the news article they claim that tidal formation contradicts CDM, in the actual paper they don't.

Here's an article (http://www.jpl.nasa.gov/news/news.cfm?release=2009-024) from a (IMHO) more reliable news source, talking about dwarf galaxies formed by tidal stripping that don't have dark matter:

In a recent study, Thilker and his colleagues found the ultraviolet signature of young stars emanating from several clumps of gas within the Leo Ring. "We speculate that these young stellar complexes are dwarf galaxies, although, as previously shown by radio astronomers, the gaseous clumps forming ,b>these galaxies lack dark matter," he said. "Almost all other galaxies we know are dominated by dark matter, which acted as a seed for the collection of their luminous components--stars, gas and dust. What we see occurring in the Leo Ring is a new mode for the formation of dwarf galaxies in material remaining from the much earlier assembly of this galaxy group.

[unixronin.livejournal.com]

Um, I thought that was the point: that they don't have associated dark matter, but that a dark-matter explanation of their orbital parameters requires that they do.

[smandal.livejournal.com]

The disc-like distribution and the fact that they all orbit in the same direction suggest tidal formation. This is what the two (http://arxiv.org/abs/0901.1658) papers (http://arxiv.org/abs/0901.1658) discuss, and nothing else.

The press release (http://www.eurekalert.org/pub_releases/2009-05/uob-sps050509.php) argues that tidal formed dwarf satellites cannot have CDM, but the dynamical observations of the dwarfs show that they spin too fast not to have dark matter.

I prefer to believe something that others researchers will read and scrutinize, not a press release. The article I link above shows that dwarfs can indeed form without dark matter, so let's wait for a confirmation of this group's observations before getting too excited ...

Many experiments reports anomalous results only to be contradicted by later, more capable experiments. (Most recently, the Fermi satellite did not confirm the same excess of electrons and positrons reported by PAMELA. Many in the community were excited by the prospect that the excess could be caused by dark matter decays ...)

[unixronin.livejournal.com]

The article I link above shows that dwarfs can indeed form without dark matter, so let's wait for a confirmation of this group's observations before getting too excited ...

Oh, sure. It'll become a lot more interesting if confirmed. On the other hand, if no-one else can reproduce it, or if someone else points out an error in the math ...

I guess to me, the main thing that makes this particular result interesting — if correct, of course — is that it might possibly be the first finding that has a chance of giving us a definitive answer on CDM. It doesn't have the ability to prove CDM or any other theory, but if it were to confirmed that this observation cannot be explained via CDM without incurring paradox, then it would be a strong indication we need to look elsewhere. (It'd be nice if it had the potential to prove CDM as well, but you take what you can get.)

That's what really caught my eye about it: the chance to narrow the uncertainty. All the guessing and "If X", "If Y" is interesting speculation, but there's a frustrating lack of confirmation in either direction. We're making progress; Milgrom's MOND is pretty much dead, and TeVeS seems to have run into irreconcilable problems. But those themselves do not constitute proof that CDM is correct.

I guess a part of the influence on my position on CDM is that I'm bothered by the apparent attitude of some of the more outspoken CDM theorists that "They're wrong, therefore we're right." It doesn't work that way. But I shouldn't allow my reaction to that to color my perception of the entire field.

[smandal.livejournal.com]

Indeed, if there is a confirmed observation with which CDM is inconsistent, that's problematic for CDM.

Even though I work on CDM, I'd be pleased as punch if the entire parameters spaces of DM candidates were eliminated by null observations -- it means the story must be more wonderful than, say, the lightest supersymmetric particle. I also have a hope that the LHC will see something far more interesting than the Higgs ...

[unixronin.livejournal.com]

I certainly hope we get some significant new results from the LHC. It's going to be potentially a big blow to future science funding if, after all the money spent on the LHC, it doesn't make some major findings. It's hard enough to get basic science properly funded now.

bah. stupid 4300-character comment limit again. (Part 1)

[unixronin.livejournal.com]

Actually, yes, I have answered you before. But I'll answer it again.

I think it comes down to two main things:

First, CDM theorists deride non-Newtonian gravitational theories as being purely descriptive and not offering a mechanism for why gravity should behave so. Yet at the same time they handwave away the fact that CDM theory is also purely descriptive. Non-Newtonian gravity theories say, in essence, "These observations can all be explained if gravity behaves slightly differently than we understand over vast distances at very low accelerations." Cold dark matter theory turns around and says, "But you don't have a mechanism for why it should be so. However, these observations can all be explained if cold dark matter exists that has these specified properties." It's pot-kettle-black. CDM fails to offer any explanation of why dark matter should exist, or what it's made of, or why it has those properties, and gets around the problem of no such dark matter having ever been detected by saying that it doesn't interact except gravitationally, and thus is undetectable except by its effects — when its properties have been defined exactly so as to explain the effects that are then used to argue for its existence. The reasoning is circular. None of the proposed CDM candidates have so far been detected, so there is at this time no tangible evidence that the stuff actually exists at all. All the evidence used to support CDM is descriptive and circumstantial.

My second problem with it is that, well, first we had the anomalous galactic rotation curve observations. And CDM theory said, "Well, those can be explained if 75% to 90% of the universe is cold dark matter that we can't detect, which we define to behave like this, that interacts only gravitationally with anything else and concentrates around galaxies, but for some reason doesn't have any angular momentum and so forms spherical halos rather than discs." Then that left the Universe too massive to be expanding at the observed rate, so it became necessary to invent dark energy, a mysterious and unexplained repulsive force without any explanation or proposed mechanism except that the observations could be explained if such a mysterious force existed and behaved in this defined manner. What is it? Oh, we don't know. Why does it exist? Oh, we don't know that either. Do you have a theory about what it is and why it exists? Uh, no. But if it did, it would explain the observations. And then you get into some of the Bullet Cluster and the various comments I've read in which some research groups have said that even dark matter and dark energy together are unable to fully explain the observations; some groups have said that maybe some non-Newtonian gravity aspects are required as well, and at least one other group has speculated that it may be necessary to posit a hitherto unknown fifth basic force in the Universe ... and it all just smacks far too much of epicycles for my liking. It has too much complexity and requires too many unknowns about which the only thing we can say is "We stipulate that it is necessary for them to behave like this."

Re: bah. stupid 4300-character comment limit again. (Part 2)

[unixronin.livejournal.com]

Frankly, I'm also bothered by the idea that CDM theory requires 75% to 90% of the entire universe to be made of what amounts to "magic invisible stuff whose existence and nature we can't explain." It strains credulity, and smacks far to much of the Far Side cartoon where the middle of the blackboard reads "And then magic happens", with a group of scientists standing around and the caption "I think you need to be a bit more specific here."

I'll place a little more trust in CDM theory when the theorists comes up with some actual evidence for the existence of dark matter, or at least comes up with some plausible and logically consistent explanation of what dark matter and dark energy are, why they should exist, and most importantly, why they should behave as CDM requires that they do. But so far, they haven't. It's handwavium. Not that the non-Newtonian gravitational theories aren't also handwavium to varying extents, though TeVeS and STVG seem at least to be built on a sound mathematical footing; but the non-Newtonian gravity theories say "We have these anomalous observations, here's a possible theory to explain them," while CDM says "We have these anomalous observations, here's a theory that explains them, oh and by the way all those other theories are bunkum because they don't have a mechanism", skipping over the inconvenient fact that the last time I heard, CDM theory doesn't have one either. (And for that matter, the weakly-bound-graviton theory — to which I'd post a link if I could find it again — does propose a single, simple, plausible mechanism underlying the observed gravitational anomalies, and if correct would make both dark matter and dark energy unnecessary.)

I'm not positing that any of the non-Newtonian gravitational theories is necessarily correct either (though the weakly-bound-graviton theory does have an appealing elegance from the viewpoint of Ockham's razor). We simply don't know enough yet to be declaring a "winner". It bothers me the way CDM theory insists that it's the One True Way and all other theories are flawed, but then when flaws (or possible flaws) in CDM theory are pointed out, such as its own lack of any proposed causal explanation, the answer too often seems to amount to "Pay no attention to the man behind the curtain." It particularly bothers me when I have heard CDM theorists say that even considering any possible explanation other than CDM is a complete waste of time. It betrays a closed mind that I don't believe is conducive to an objective approach to the problem.

I'm not ruling out CDM. I simply feel that CDM has flaws that are too often being simply ignored or wallpapered over, and that it has thus far failed to convince me. I think it's too early in the race yet to be putting all our money on one horse.

Re: bah. stupid 4300-character comment limit again. (Part 1)

[smandal.livejournal.com]

Let me respond to you point-by-point:

1. CDM is circumstantial

That's right -- all theories are :) We have observations that can't be explained by current theories, so we cook up a new one and then test their novel predictions. The problem with theories like MOND is that they are not even consistent with the data; TeVeS has to have a lot of moving parts to be consistent, so in addition to being less predictive it has technical problems that spin the theory out of control.

2. CDM is not testable

This is not correct. CDM may interact weakly, i.e. through electroweak interaction, and so we could detect them in underground labs, in colliders or via their astrophysical decay/annihilation products. Even if CDM only interactions gravitationally, it may still be detectable by other strategies such as lensing.

3. CDM makes the universe too massive, so we invoke dark energy

This is not right. The universe is not too massive to keep expanding -- the cosmic microwave background (CMB) tells us that the energy density in the universe is very close to the critical value, where the Hubble parameter is constant. The question that remains is what are the relative fractions of normal matter/light, CDM and dark energy.

We have three unknowns and one equation, so we need two more relations to get the parameters of the theory, and yet one more to falsify it. Supernova observations gives us the ratio total matter-to-dark energy, CMB also gives baryon-to-CDM ratio, and finally galactic observations of total luminosity as well as their dynamics gives us the total matter density.

4. Bullet Cluster dynamics are not explained by dark matter alone

Is that so? Please send me a reference arguing this!

It is very unlikey that dark energy would be involved since the cluster is ~1 megaparsec in size, and the effects of dark energy are negligible below ~100 megaparsec.

Re: bah. stupid 4300-character comment limit again. (Part 1)

[unixronin.livejournal.com]

1 — yes, MOND (per se) and TeVeS both appear to be out of the running at this point. SVTG ... er ... STVG (too many acronyms!) looks as though it may have promise, but needs more evaluation yet. The weakly-bound-graviton theory is self-consistent and truly elegant (and I suspect it could have applicability to other aspects of M-throry as well), but it's very new, there hasn't been a lot of time to examine it yet, and it's going to be a bitch to test.

2 — I didn't say CDM wasn't testable, I just said that so far it has not been successfully tested and either verified or refuted, but I dislike the tendency to regard it as all but proven despite that. As mentioned,¹ if — say — the LHC detects the Higgs boson, and they measure the mass and say, "Well, dang, the math works out for CDM with these numbers", hey, I'm ready to shrug and say "I guess you guys have got something after all." But in the meantime, I consider it hypocritical of CDM theorists to dismiss non-Newtonian gravitational theories as a class as "purely descriptive" when, at this point in our knowledge, the same charge can with equal fairness be leveled at CDM.

3 — I'll freely admit I don't fully grasp the arguments behind dark energy and its role. It still bothers me that CDM has to invoke multiple new entities, required to behave in certain tightly specified ways, but for which we are given no explanation of what they actually are — or why they should exist — other than that this specific mathematical framework relies upon the assumption that they do. It's the "If we had some ham, we could have ham and eggs, if we had any eggs" problem.

4 — I know I've read papers (my memory says out of Germany) pointing out aspects of the Bullet observations which [the authors assert] cannot be fully explained by CDM without positing additional factors not currently contained within CDM, even if they considered the contribution of dark energy. This is one of several things I went looking for links to, and couldn't find. (I also tried — but failed — to find the paper that invoked the possibility of a fifth basic force in the Universe, and the weakly-bound-graviton theory.)


[1] Hmmm .... actually, in hindsight, I'm not certain I did explicitly mention that. But I intended to. Honest...

Re: bah. stupid 4300-character comment limit again. (Part 1)

[unixronin.livejournal.com]

Actually, a question there: the comment that the CMB gives us the baryon-to-dark-matter ratio. I dont think I've come across that relation before, or at least I don't recall it when stated in those terms. Could you please enlarge upon that a little?

Re: bah. stupid 4300-character comment limit again. (Part 1)

[smandal.livejournal.com]

2. Scientists, being human beings, do have prejudices and get into flamewars over their pet theoretical frameworks. STVG, if it's self-consistent and consistent with the data, is a perfectly valid hypothesis. Like I said, I'll be really impressed if STVG can solve all of TeVeS's theoretical problems without being hopeless complicated. A complicated theory is undesirable because it can be tweaked to get results, making it less predictive and therefore less useful as a hypothesis.

3. Any new theoretical construct is a placeholder for our ignorance. Take for example the neutrino: Fermi basically pulled it out of his ass in order to conserve momentum in neutron decay. Purely circumstantial. But it seems to have worked out ...

Just to be clear, the expansion history depends only on the ratio of total matter-to-dark energy, whereas CMB, galactic dynamics and structure formation depend on CDM but not dark energy. So it's not right to say that CDM and dark energy need each other ...

Re: bah. stupid 4300-character comment limit again. (Part 1)

[smandal.livejournal.com]

Read the section on "Higher Peaks" in the wonderful web tutorial (http://background.uchicago.edu/~whu/intermediate/intermediate.html) by Wayne Hu.

Basically, the location of the first peak tells you the geometric curvature of the universe; the ratio of the first to the second tells you the baryon abundance; and the ratio of the first and third tells you the dark matter abundance for that given baryon abundance.

(Technically, the third and higher peaks tell you the dark matter-to-radiation ratio for the given baryon abundance, but the total radiation is very tightly constrained by Big Bang nucleosynthesis (http://en.wikipedia.org/wiki/Big_Bang_nucleosynthesis).)

[ithildae.livejournal.com]

I think you are trying to reach too far, with too little information. We have only gathered universal observations from a single position, and a comparatively instantaneous sample at that.

I don't believe we are at the point of trying to unify our theories, but finding a model that will direct us to look someplace productive next. The models are useful only as far as they point us to new information. CDM seem somewhat useful in that regard.

For me, it doesn't matter what the "Fundamental Truth" really is, I doubt I would be able to understand it, even if I knew it. I just want enough, that I can grasp, so that I can learn more.

[unixronin.livejournal.com]

I think you are trying to reach too far, with too little information. We have only gathered universal observations from a single position, and a comparatively instantaneous sample at that.

Precisely. It's too soon to be "picking a winner", and ignoring inconsistencies in any theory only makes that worse.

As noted above, one of the problems I have with CDM is that it doesn't point us to new information. As CDM theorists are fond of saying about competing theories, CDM theory is purely descriptive. It adds no new information. It just says "These anomalous explanations could be explained if 90% of the universe were made of magic invisible stuff that behaves in this defined way. We can't tell you what it is or why it should behave like that, but we need it to do so. Oh, and we need this other magic repulsive force, too, but we have no explanation for why that should exist either, we just need it to exist and behave like this to make the math work." The public declarations I have seen from some CDM theorists that studying any other explanation is a complete waste of time lead me to feel that CDM theory is beginning to actively obstruct studying the questions. To touch on other recent discussions, it seems to be verging dangerously close to "A wizard did it."

[ithildae.livejournal.com]

I agree, it is far too soon to be picking a winner. As smandal states, A theory is only a placeholder for our ignorance. A useful theory is predictive.

I am not ready to be picking a loser either. Simplicity is a plus with any theory, even if it is not perfectly represented by all the data, as long as it increases our understanding and gives us new directions to look. I don't have anywhere near your specific knowledge of the subject. What I have read (and understand) implies that CDM theory is not the monster that it appears to you as. We are still wrapping our heads around the implications of several theories. I don't think we understand what our suppositions imply well enough to throw CDM out, yet.

[unixronin.livejournal.com]

I'm not seeking to throw it out, particularly; I just feel that it is not as flawless and immaculate as many of its advocates claim. It is not the known answer, yet; we shouldn't be acting as though it is and dismissing all other possibilities out of hand. Yet I've heard top-level CDM theorists assert that studying any competing theory is a complete waste of time, and shouldn't even be funded. That bothers me.

[ithildae.livejournal.com]

Science changes one funeral at a time.

There are also significant research dollars at stake. The declarations should bother you. But they should not be taken seriously. It is no more than self-interest talking. The science process is still intact.