A snip making fun of the concept of dark matter- why is this incorrect?

[Galteeth]

This is a snippet from a mock encyclopedia which i will not link to. While I am not endorsing this viewpoint, I want to understand, what is fundamentaly inaccurate about what is being said?


"
In reality, the highest level of education required to comprehend the theory of dark matter is just below first-grade elementary school curriculum. We called the subject "coloring books," remember? Basically, astronomers spend about 100 years looking at a crudely drawn map of the universe and paint dark spots on it with a Sharpie wherever their calculations for gravitational force are not adhering to their formulas--by their own admission, these formulas have failed in over 70% of the universe. Therefore, Dark Matter is to blame.

The bulk of the calculations for dark matter can be compared to the God of the Gaps mentality:

God of the Gaps: If explainable life - unexplainable life = x, then God made x. Dark Matter: If predictable observation - unpredictable observation = x, then dark matter = x.

In both instances, the existence of the explanatory variable is unfoundedly prestanding. God/dark matter must be true, so God/dark matter must equal x. Dark matter does not explain why modern gravitational theory is failing all over the cosmos, but rather is used to fill in the gaps. "

 

[Wallace]

It is wrong because the theory of dark matter does make many different and independantly verifiable predictions. If it was akin to 'god of the gaps' then it could simply give you whatever answer you wanted to any problem with gravity. This is not the case with dark matter.

We need dark matter in order to get the correct CMB anisotropy power spectrum, we also need dark matter in order to get the correct structure formation history (for instance the number of galaxy clusters per unit volume as a function of redshift), we also need dark matter to explain the discrepancy between the internal dynamics of galaxies and galaxy clusters (e.g. galaxy rotation curves) and the amount of visible matter.

Any one of these factors could alone point to the need for dark matter to exist, however the crucial thing is that they all require the same amount of dark matter to exist on average in the Universe. If it was simply a fudge factor, the fudge would be different every time, the amazing thing is that we can get a consistant picture by a minimal additional of collionless dark matter to the model.

Numerical simulations following the evolution of structure in dark matter as well as the galaxies that form in response to this match what we observe in the real universe. Simply adding dark matter to the theory doesn't automatically give you this agreement, the universe really is telling us that dark matter is there.

Of course it is still possible that dark matter in fact doesn't exist, or has some unexpected properties, but no better performing theory has yet been proposed. You can do as well as dark matter in say galaxy rotation curves, but then that theory gives you the wrong number of clusters or the wrong CMB anisotropy power spectrum. No currrent alternative explains all the data we have as well or as simply as dark matter.

 

[twofish-quant]

This is also wrong because people are thinking about alternatives to dark matter. Yes, it's quite possible that there is no dark matter or dark energy and it's all the result of some modified gravity or unknown force. But that doesn't get you off the hook, because if you argue that the effects of dark matter and dark energy are the result of some unknown force of some modification to gravity then you have to come up with a description of the force and a description of how to modify gravity to get the exact observations. It's not easy, people are trying to do it (there are *tons* of papers of people proposing changes to gravity).

Right now dark matter and dark energy are preferred because they are the simpliest explanation. But there is a ton of new data coming in, and I wouldn't be surprised if this time next year, everybody looks at each other and says well it looks like dark matter doesn't work as well as modified gravity.

 

[Chalnoth]

Right now dark matter and dark energy are preferred because they are the simpliest explanation. But there is a ton of new data coming in, and I wouldn't be surprised if this time next year, everybody looks at each other and says well it looks like dark matter doesn't work as well as modified gravity.

I would. The evidence for dark matter is quite varied and strong.

 

[Wallace]

I'd also be pretty surprised if dark matter got completely canned. Things like the Bullet Cluster result as well as plenty of other lensing data are pretty convincing. That's before you even consider the CMB which is very hard to fit without dark matter.

On the other hand, I think it's likely that dark matter will prove to be more complicated than simply collionless stuff interacting only via gravity. There are some persistent problems relating to the central regions of dark matter halos that may indicate some amount of coupling to baryons (or possibly dark energy??) or possibly some self interaction or non-zero dispersion velocity, so called 'tepid' dark matter (since it isn't cold but isn't the hot dark matter than was ruled out about a decade ago).

One of the real problems with modified gravity vs dark matter is that you end up needing a different law of gravity for different systems, which makes no sense. On the other hand you would expect this with dark matter, since each galaxy or cluster will be expected to have some unique variation in the baryon/dark matter ratio. When people say MOND fits this cluster or that galaxy when you look at the individual papers each 'fit' ends up implying a different and unique law of gravity for each system studied!

That being said, people keep working on this, so maybe there is more to it than I give credit for, but the case for modified gravity to remove dark matter seems weak to me at this point, and getting weaker with each new bit of data.

Modified gravity on larger scales, addressing the need for dark energy, is a different story though.

 

[twofish-quant]

I'd be very surprised if there were *no* dark matter, but I'd wouldn't be too surprised if it turned out that there were about five or six different types of dark matter, and none of them matter at cosmological scales. I was thinking mostly in terms of cosmological scales, and if I haven't misremembered, you need much more dark matter to get the expansion of the universe to work than you need to get galaxy rotation curves to match up.

At cosmological scales, it's not hard to imagine some effect that mimics dark matter. I don't think it's *that* hard to get the CMB to match up without dark matter. One reason it's easy is that you don't get the problem that you have with MOND. All of the impact of dark matter on cosmology come down to a single number that you plug into LCDM models, and so it's not too hard to think of something else that mimics the behavior of that one number.

What makes this interesting is that it's not either/or. You could have a mix of dark matter, dark energy, and some modifications in gravity. You could end up with six types of dark matter, five types of dark energy, and three or four modifications to gravity.

Or maybe not.

But in any case, it's not the situation that cosmologists have simply rejected non-dark matter theories. People have thought this through and where people favor dark matter, they have some pretty good reason.

 

[Chalnoth]

At cosmological scales, it's not hard to imagine some effect that mimics dark matter. I don't think it's *that* hard to get the CMB to match up without dark matter.

Yeah, it's fantastically difficult. Dark matter induces a difference in amplitude between the even and odd peaks of the power spectrum of the CMB. This comes about because baryonic matter has charge and thus interacts with photons. This causes it to bounce in the early universe, before the CMB is emitted. Dark matter does not. The lack of bouncing suppresses the even-numbered peaks of the CMB power spectrum.

And that's not something you can easily explain without dark matter (if you had only baryonic matter, then all of the peaks would be the same size, though with suppression of higher peaks due to how the CMB is emitted).

 

[Wallace]

Edit: ninjad by Chalnoth who gave a much clearer explanation of the CMB trouble without DM

I'd be very surprised if there were *no* dark matter, but I'd wouldn't be too surprised if it turned out that there were about five or six different types of dark matter, and none of them matter at cosmological scales. I was thinking mostly in terms of cosmological scales, and if I haven't misremembered, you need much more dark matter to get the expansion of the universe to work than you need to get galaxy rotation curves to match up.

(emphasis mine)

The dark matter/baryon ratio is consistant between 'cosmological' dark matter and 'astrophysical' dark matter. This doesn't mean that the ratio in galaxies or clusters is the same as the mean ration, but what we do know is that simulations and modelling of how astrophysical objects form give you galaxies and clusters that look like the ones we see in the real Universe. There is an agreement between the two, so in fact you do need 'the same amount of dark matter' to fit both data sets. The amount of dark matter that gives you the right expansion history also gives you the right amount of structure. Constraints on both are fairly tight and give a consistent picture.

At cosmological scales, it's not hard to imagine some effect that mimics dark matter. I don't think it's *that* hard to get the CMB to match up without dark matter. One reason it's easy is that you don't get the problem that you have with MOND. All of the impact of dark matter on cosmology come down to a single number that you plug into LCDM models, and so it's not too hard to think of something else that mimics the behavior of that one number.

You can't just say that because dark matter is described by a single number in the paramaterisation, then it has just a simple effect. This is not true at all. The physical implications of that one number have to be carefully modeled in relation to all observable effects. Have a look at Wayne Hu's site for the animations of what happens to the CMB peaks if you don't include dark matter, they look nothing like what we see. It is very hard for modified gravity to mimic this, indeed no one has managed to do it as of yet.

You might not think it's hard to get something to mimic 'one number' but its the physics behind that number that is important, and it is in fact very hard to get something to mimic it, so hard that it has not yet been succesfully done!

What makes this interesting is that it's not either/or. You could have a mix of dark matter, dark energy, and some modifications in gravity. You could end up with six types of dark matter, five types of dark energy, and three or four modifications to gravity.

Let's hope it's not that complicated, but you're right, it may well be! One of the real difficulties in ever really concluding the case for dark matter is that even if you detected a plausible candidate at say the LHC or a future even bigger detector, you would still have to be really sure you understood the production of this particle such that you could predict the abudance post inflation/BBN such that you were really sure that the particle is question not only could be dark matter, but that it really existed in the right quantity. That would actually be quite tricky to do I think.

 

[jnorman]

dark matter may indeed account for some of the observed anomalies. at least there is some potential explanation of what dark matter may comprise, in terms of WIMPS. it is dark energy which borders on sheer mythology, and unfortunately, DE is deemed as 70% of the overall effect. ihave yet to see any explanation of what DE may be, or how it might fit into the standard model, its source, or how it could be invisible to direct detection. please inform me if i am misinterpreting DE. thanks.

 

[Wallace]

Why does DE 'border on sheer mythology'? It isn't 'deemed' to be 70% of the current energy density, that what the data is telling us! The simplest explanation of DE is a non-zero vacuum energy, which would fit in with the standard model if some symmtery breaking mechanism was shown to cancel most of the vacuum energy you get if you do a naive cancellation, leaving a tiny amount which would explain the cosmological observations. No one has yet been able to do this though.

Remember that the late time acceleration of the Universe driven by dark energy is no weirder than the early time acceleration in the inflationary era, and indeed it's possible that two epochs are related (in terms of the physics of a single field responsible for both periods of acceleration).

The bottom line though is that on a theoretical level we really don't have that much of an idea about dark energy, but that doesn't mean that the wealth of data telling us that it exists (or that some other new thing we don't understand makes it look like it exists) qaulifies for the title of 'mythology'. Just beacuse we don't understand it doesn't mean that it doesn't exist.

 

[twofish-quant]

Agree with most of your points. Right now, dark matter/energy is the simplest explanation for what is going on because you get the most explanation with the smallest set of weird assumptions. The question is one of degree. Also one problem with arguments based on "minimum weirdness" is that something things bother people less than other things.

Also I don't think that the data has been around long enough for me to feel comfortable saying that there is no way to fit the data with a simple modified gravity scheme. The longer people try hard to do it without being able to, the more likely it is that it can't be done. We are moving in that direction, but I don't think we are quite there yet.

The bottom line though is that on a theoretical level we really don't have that much of an idea about dark energy, but that doesn't mean that the wealth of data telling us that it exists (or that some other new thing we don't understand makes it look like it exists) qaulifies for the title of 'mythology'. Just beacuse we don't understand it doesn't mean that it doesn't exist.

One thing is that it's clear that *something* exists. The question is then what's the list weird thing that you can assume that explains the most amount of data, and right now it's pretty clearly some form of dark matter/dark energy. One point is that people are trying to find the explanation that's the "least weird" and right now dark matter/dark energy is less weird than changing gravity.

 

[twofish-quant]

Just to clarify. I do agree that you need *something* to get the peaks to match up and a universe that is zero anything that resembles dark matter just won't work, and is pretty much excluded.

However, since dark matter just doesn't quite fit with the standard models of particle physics, I wouldn't be surprised if what is causes dark matter effects is strictly speaking not "matter" per se. I think I'm much bothered more by the fact that we have no idea what dark matter is than most cosmologists are, and much less pessimistic that you can figure out some weird matter-like effect that isn't strictly speaking matter, than most people.

We have something out there that looks like a duck, quacks like a duck, but since it turns out to be a really, really weird duck, I'm would be surprised if it was a goose or a loud swan rather than a duck.

 

[Chalnoth]

I wouldn't lump dark matter and dark energy together quite so strongly. Dark matter is more or less confirmed. We basically know now that it's made up of some sort of weakly-interacting massive particle. We don't know precisely what particle it is yet, or precisely what its (weak) interactions are. But we are pretty confident it's out there, and are extremely confident that modified gravity cannot explain our dark matter observations.

With dark energy, the field is much more wide open. We really don't know what's causing the observed accelerated expansion, though it seems that the picture should be much, much clearer within 5-10 years.

 

[Chalnoth]

However, since dark matter just doesn't quite fit with the standard models of particle physics, I wouldn't be surprised if what is causes dark matter effects is strictly speaking not "matter" per se.

How does it not quite fit?

 

[jnorman]

wallace - inre: "dark energy is no weirder than the early time acceleration in the inflationary era,..."

point taken, but guth never proposed any mechanism by which that could occur either, as i recall. 'reality: what a concept!'

 

[Chalnoth]

point taken, but guth never proposed any mechanism by which that could occur either, as i recall.

No. Others did.

 

[twofish-quant]

How does it not quite fit?

The properties of dark matter matches no known particle in the standard model, and other than stable higgs bosons, it doesn't match any particle that you can get from any simple non-speculative generalizations of the standard model. To get any particles that matches dark matter you have to get rather speculative (axions or supersymmetry).

Don't get me wrong. If I had to place a bet, I'd say that dark matter is probably matter, but I am bothered by the fact that it doesn't quite fit anything resembling a known particle.

 

[Chalnoth]

The properties of dark matter matches no known particle in the standard model, and other than stable higgs bosons, it doesn't match any particle that you can get from any simple non-speculative generalizations of the standard model. To get any particles that matches dark matter you have to get rather speculative (axions or supersymmetry).

Don't get me wrong. If I had to place a bet, I'd say that dark matter is probably matter, but I am bothered by the fact that it doesn't quite fit anything resembling a known particle.

Except the standard model is known to be incomplete, from purely internal arguments. So it shouldn't be any surprise at all that there are some particles beyond it. Furthermore, there is a particle which very much resembles the properties required for dark matter: the neutrino. It's just that the neutrino doesn't have enough mass for the job.

 

[twofish-quant]

Except the standard model is known to be incomplete, from purely internal arguments. So it shouldn't be any surprise at all that there are some particles beyond it.

Un/fortunately the incompleteness arguments tell you more or less what types of particles to expect, and none of them quite match.

Furthermore, there is a particle which very much resembles the properties required for dark matter: the neutrino. It's just that the neutrino doesn't have enough mass for the job.

It's somewhat more involved and amusing than that. Basically from nucleosynthesis arguments, we think know the number of neutrinos in universe. For every baryon there is a certain number of neutrinos, and the number of the three types of neutrinos are the same. This poses a problem. If neutrinos are heavy, then the universe would have collapsed eons ago, and the most stringent limits on neutrino masses come from this limit. Trouble is that if you limit the mass of the neutrinos to keep the universe from imploding, it turns out that they are fast moving and "hot" and so they smear out galactic structure. If you start tinkering with early neutrino physics then lots of things in big bang nucleosynthesis break. Also you could experiment with a massive neutrino that decays, but since we know the number of neutrinos, we can set a limit based on the decays that we don't see,

People spent about a decade trying to get neutrino as dark matter models to work, but I think people ended up figuring out that they wouldn't.

There's also a cool argument that says that they can't be galactic dark matter. If you imagine a galaxy as a atom with the neutrinos as "electrons" there is a maximum number of neutrinos you can can before all of the "atomic shells" are filled. It turns out that if you take the limits of the mass of the neutrinos, you don't have enough neutrinos to account for all of the mass before you fill up all of the "atomic shells" of the galaxy.

 

[Wallace]

I agree that Neutrinos are all but ruled out as dark matter, but I find the 'atomic shell' argument about galaxies completely implausible. The density of quantum states allowed for something the size of a galaxy would be enormous, way way into the continuum limit. You simply don't see quantum effects on that kind of scale! Do you have a reference for this? I'm happy to be wrong, since that would be a neat calculation if it's true, but it just doesn't sound right.

 

[twofish-quant]

I agree that Neutrinos are all but ruled out as dark matter, but I find the 'atomic shell' argument about galaxies completely implausible.

See the references here...

http://www.astro.princeton.edu/~dns/MAP/Bahcall/node6.html

Also there was a review paper about this limit in the Annual Reviews of Astronomy and Astrophysics on the cosmological limits on neutrino properties. The neat thing about this argument is that it's extremely robust and requires nothing more than first year quantum. You treat a galaxy as the nucleus of an atom, then you use your standard quantum results for a hydrogen atom. This gives you the number of states. Then you calculate the number of neutrinos you need to get the required mass, and you find that you can't have that many neutrinos without degeneracy problems.

The density of quantum states allowed for something the size of a galaxy would be enormous, way way into the continuum limit.

True. The number of possible quantum states in a galaxy is enormous, but it's finite. The trouble is that as the mass of neutrinos goes to zero, the number of neutrinos that you need to explain dark matter goes to infinity. Putting in known numbers, you find that the number of neutrinos you need are way, way, way more than they number you can fit into a galaxy.

You simply don't see quantum effects on that kind of scale!

Yes you do. :-) :-)

I should point out that the fact that you do see quantum mechanical effects at galactic scales is why I wouldn't be surprised if you have some quantum effect that acts like dark matter.

Do you have a reference for this? I'm happy to be wrong, since that would be a neat calculation if it's true, but it just doesn't sound right.

It is a neat calculation, and it would be an ideal problem set problem for first year quantum. I almost fell out of my chair the first time I read it.

 

[Wallace]

Ah I see what you mean now, because the allowed neutrino energy is so small you need so many. Yep, I can see how that would work. However, it is the constraints on Neutrino mass from elsewhere (primarily the observed power spectrum of the CMB and large scale structure) that give you the lower mass limit in the first place. You couldn't actually use this argument to demonstrate the neutrinos can't be dark matter, because you could simply increase the mass such you wouldn't need nearly so many of them. It's the other things that already rule out neutrinos as dark matter that give you the mass limit in the first place.

 

[twofish-quant]

You couldn't actually use this argument to demonstrate the neutrinos can't be dark matter, because you could simply increase the mass such you wouldn't need nearly so many of them.

Can't do that. It turns out that the standard models (both big bang and particle physics) sets very strong limits on the number of neutrinos. If you increase the mass and decrease the number of neutrinos, you end up huge nucleosynthesis and standard model problems. If you assume that heavy neutrinos are unstable and decay, there are so many neutrinos that you would have seen decay traces.

Also the most stringent limits on the mass of the neutrino are also cosmological. Since we think we know how many neutrinos there are in the universe (and it's a huge number), it doesn't take much mass before you cause the universe to collapse. To its a "name your poison" situation.

neutrinos are massive -> universe collapses
neutrinos are light -> large scale structure gets smeared out
neutrinos are unstable -> you should see background radiation
neutrinos are uncommon -> big bang nucleosynthesis breaks down or you have to massively redo standard particle physics model

Now people *are* trying to get around this problem by doing things to big-bang models to try to get the neutrino production rate down, but they haven't come up with anything really convincing. The basic problem is that we aren't talking about a small "fudge factor" that can be fixed with a slight model tweaking. The differences are huge orders of magnitude requiring you to take a sledgehammer to get things to work.

One piece of "good" news is that big bang nucleosynthesis is pretty consistent with their being three types of neutrinos. That makes this more difficult since you can't postulate a new neutrino.

The reason I got into neutrino physics would be is that I was running type II supernova simulations and since non-electron neutrinos cause an energy loss, it would be really convenient for me if tau and mu neutrinos where heavy. However, it turns out that if tau and mu neutrinos where heavy enough to be important for supernova simulations (i.e. MeV energies), then the universe would have collapsed.

 

[Wallace]

Right, exactly. Because of a bazillion reasons there are tight limits on the mass of neutrinos, meaning that they can't be the dark matter that appears to exist in galaxies (or more accurately that galaxies appear to exist in). Now, having got those limits you can also make an argument about quantum states in a galaxy, but that only tells you anything about dark matter because of the prior limits on the neutrino mass.

All I'm saying is that considering galaxies alone, you can't rule neutrinos out because you can simply increase the mass until everything works. Of course you then violate all those other constraint on the mass, so you can't do that. Therefore it is those other constraints, not some quantum mechanical argument about galaxies that is actually giving you any information.

 

[Ich]

you can't postulate a new neutrino.

I think you misunderstood Chalnoth. He was not implying that neutrinos make up Dark Matter, he said that extensions to the Standard Model natually come with a suitable particle (neutralino), which indeed very much resembles a neutrino with larger mass. Baryogenesis arguments do not apply here.

 

[Chalnoth]

I think you misunderstood Chalnoth. He was not implying that neutrinos make up Dark Matter, he said that extensions to the Standard Model natually come with a suitable particle (neutralino), which indeed very much resembles a neutrino with larger mass. Baryogenesis arguments do not apply here.

Well, actually, the neutralino, despite its similar name, isn't much related to the neutrino. What I was stating is that there is a particle that has basically all of the requirements of a dark matter particle, except mass: the neutrino. It doesn't fit the bill, of course, but the lightest superpartner is a very compelling candidate (the neutralino is one of the possibilities here...I believe it is a neutral-charged combination of the Higgs and W and Z bosons, but I am not certain).

The superpartner of the neutrino is the sneutrino.

 

[Ich]

Well, actually, the neutralino, despite its similar name, isn't much related to the neutrino.

A massive, spin 1/2, color- and electroneutral particle, thus only interacting by the weak force and gravitation, I think (I'm not a specialist). I know it's not related to the neutrino, the point is that it's quite similar to a massive neutrino in these properties.

 

[Chalnoth]

A massive, spin 1/2, color- and electroneutral particle, thus only interacting by the weak force and gravitation, I think (I'm not a specialist). I know it's not related to the neutrino, the point is that it's quite similar to a massive neutrino in these properties.

Fair enough.

 

[twofish-quant]

The trouble with these extra particles is that they all come from supersymmetry, and there is not the slightly shred of experimental evidence in favor for supersymmetry, and the theoretical reasons for supersymmetry aren't IMHO that compelling. Replacing one unknown particle with a bizilion unknown particles doesn't see to me to be that much of an improvement.

 

[hamster143]

Does anyone know of a good recent review (on arxiv or elsewhere) that summarizes state-of-the-art observational constraints on dark matter, from CMB, big bang nucleosynthesis & such?

 

[Chalnoth]

The trouble with these extra particles is that they all come from supersymmetry, and there is not the slightly shred of experimental evidence in favor for supersymmetry, and the theoretical reasons for supersymmetry aren't IMHO that compelling. Replacing one unknown particle with a bizilion unknown particles doesn't see to me to be that much of an improvement.

Even without supersymmetry for specific suggestions, I don't see how extra particles are in any way unexpected regardless.