Question about dark matter...

[SRH]

How do scientists know that dark matter is not visible?

Astronomers keep discovering moons within our own solar system. Trying to spot a small moon, billions of light years away from us, would be much more difficult.

How do we know that the gravitational effects seen in distant galaxies are not due to planets and moons that our telescopes cannot detect?

Why do we attribute this mass and its gravitational affects to invisible dark matter and eliminate the possibility that it might come from visible matter that we just can't detect?

There has to be a reason, but I have no idea what it is. Thanks!

[ngc3314]

If that much additional ordinary (baryonic) matter is present, the level of agreement between abundances of light elements (in the least-altered environments, like intergalactic gas) with prediction of nucleosynthesis in the early Universe (which are functions of the baryon density a few minutes into cosmic history) goes away, and then requires some other explanation which hasn't been found. Other recent threads have beaten hard on the point that there is a size range where cold objects far from stars are currently almost impossible to detect (we might occasionally get very lucky with gravitational microlensing, for example). A key point is that the required distribution of mass in galaxies where we have the most radially-extensive data (from neutral hydrogen disks well beyond detected starlight, or companion galaxies) has to be quite different from that of the material we can detect, so these putative compact objects would need to have been formed by very different processes than the stars we can see. This applies to the total mass as well - a few loose planets per star falls into the error of global mass estimates for galaxies, but if you have low-mass objects floating around formed quite independently of stars, 1000 Jupiters or a million Earth-mass things per star, that's a different game.

[SRH]

I apologize but I am a novice.

What I think you are saying is:
1. Given the theory of nucleosynthesis in the early universe, there should be a much greater abundance of light elements observed in galaxies. (Or are you saying there should be much less abundance of light elements observed in galaxies??)
2. I don't understand what this means..."(from neutral hydrogen disks well beyond detected starlight, or companion galaxies)"...I think I may be confused by the word "beyond"...are you referring to distance, or time, or the observed red-shift of the emitted light's spectrum, or something else?
3. The distribution of the undetectable matter in the galaxy (which is a spherical halo??) is not well explained by current star formation theory.

To summarize... there needs to be many, many planets and moons to account for the needed mass to match the observation, AND there shouldn't be that much mass/matter if the big bang theory is correct, AND if star formation theory is correct. Is this right?

[kzb]

The problem with planets and moons as galactic dark matter is that the numbers required are enormous, and those numbers are said to be ruled out by microlensing studies. Recently we also had the WISE results, which rule out brown dwarfs as a significant contributor to dark matter.

[Cougar]

Your summary is pretty good, though I'd leave out the star formation part. But your numbered points need a little work.

1. Given the theory of nucleosynthesis in the early universe, there should be a much greater abundance of light elements observed in galaxies. (Or are you saying there should be much less abundance of light elements observed in galaxies??)

The observed abundance of hydrogen, helium, deuterium, etc. (baryons) appears to match up well with the theory of big bang nucleosynthesis. But observed gravitational effects suggest there must be 5-10 times more matter in galaxies than the baryons we can detect. Put together, these observations suggest the "extra gravity" is the result of some dark matter that is non-baryonic.

2. I don't understand what this means..."(from neutral hydrogen disks well beyond detected starlight, or companion galaxies)"...I think I may be confused by the word "beyond"...are you referring to distance, or time, or the observed red-shift of the emitted light's spectrum, or something else?

I believe ngc3314 is just talking about the observed movement of hydrogen gas (which we can detect) that is well beyond the visible galactic disk - it's moving as if there was 5-10 times more matter affecting it gravitationally than the matter we can detect. Same with small companion galaxies.

3. The distribution of the undetectable matter in the galaxy (which is a spherical halo??) is not well explained by current star formation theory.

I think he's just saying that the distribution of dark matter is different than the distribution of stars within a galaxy, "so these putative compact objects would need to have been formed by very different processes than the stars we can see." Just another "suggestion" that the dark matter is non-baryonic.

[Ivan Viehoff]

The simplest way to look it is as follows. When there are a lot of small things, they become a cloud, and we can detect clouds, even if we can't detect the individual items in the cloud. Certainly gas clouds and dust clouds are easily detectable. Even if they were very diffuse, there is simply so much matter required to be in them, we would be able to detect them, because we would be looking through such great depths of cloud. Clouds made up of bigger objects, what has been referred to as compact halo objects, which are the moons and planets you talk about, are harder to detect, which is why it has taken longer to rule those out.

[Cougar]

Clouds made up of bigger objects, what has been referred to as compact halo objects, which are the moons and planets you talk about, are harder to detect, which is why it has taken longer to rule those out.

Correct. But in order for difficult-to-detect "planets and moons" to fill the shoes of dark matter, there would have to be WAY too many of them. We need at least 7 times the mass of stars and gas and dust that are visible in the galaxy to account for the velocity of stellar/gas orbits due to gravity. There would have to be something like 5,000 planets and moons per star.* That is unlikely.

______________________
* Wiki says the mass of our solar system is 1.0014 solar masses, so apparently all the planets and moons and asteroids together make up just 0.14% of the easy-to-detect Sun's mass. Notice this calculation is very back-of-the-envelope, neglecting gas and dust and considering the Sun as average. But hey, I did do the math. : )

[SRH]

Are there any theories that use magnetism in addition to gravity to account for the velocity of stellar/gas orbits?

[Jeff Root]

Magnetism is a strong force at short distances, but very
weak at long distances. And it is very weak on electrically
neutral matter, which stars and planets are.

-- Jeff, in Minneapolis

[ctcoker]

The observed abundance of hydrogen, helium, deuterium, etc. (baryons) appears to match up well with the theory of big bang nucleosynthesis. But observed gravitational effects suggest there must be 5-10 times more matter in galaxies than the baryons we can detect. Put together, these observations suggest the "extra gravity" is the result of some dark matter that is non-baryonic.

BBN speaks to the relative abundances of elements, not the absolute quantities. What really suggests that dark matter is non-baryonic is that if there was a large enough swarm of undetected rocky bodies, there should be correspondingly more hydrogen and helium. We don't see any of this gas, which should be easily traced via 21-cm and CO emission, so therefore dark matter is almost certainly non-baryonic.

[SRH]

BBN speaks to the relative abundances of elements, not the absolute quantities. What really suggests that dark matter is non-baryonic is that if there was a large enough swarm of undetected rocky bodies, there should be correspondingly more hydrogen and helium. We don't see any of this gas, which should be easily traced via 21-cm and CO emission, so therefore dark matter is almost certainly non-baryonic.

But mars and the moon don't have corresponding hydrogen and helium.
Why cant you have an abundance of rocky bodies without these gases?

[AGN Fuel]

But mars and the moon don't have corresponding hydrogen and helium.
Why cant you have an abundance of rocky bodies without these gases?

Objects such as Mars and the Moon are formed from elements that were created by nuclear processes inside earlier generations of stars. Those stars in turn are formed from hydrogen and helium.

To have sufficient numbers of compact objects to account for the "missing mass", there would need to have been vastly greater number of stars having existed and died previously than can be accounted for by our current measurements of the gas distribution we see.

[Shaula]

But mars and the moon don't have corresponding hydrogen and helium.
Why cant you have an abundance of rocky bodies without these gases?

The abundances are global, not local. Look at the solar system - 99.9% of it is Sun. A big chunk of the rest is gas giant. Rocky bodies are a tiny amount of the total amount of matter. That is what is being said - that in order for there to be billions of billions of rocky bodies there would have to be untold amounts of hydrogen and helium. This is simply due to the fact that we have a fairly good idea on how much hydrogen there is globally for each atom of any other element.

[TooMany]

The problem with planets and moons as galactic dark matter is that the numbers required are enormous, and those numbers are said to be ruled out by microlensing studies. Recently we also had the WISE results, which rule out brown dwarfs as a significant contributor to dark matter.

Microlensing has ruled out a large contribution from Jupiter mass and larger objects in the direction of the galactic center and above the galactic plain toward the Magellanic Clouds. There is still a lot to be learned from further microlensing studies. Similarly, WISE found fewer brown dwarfs within it's detection ability than expected. I think it's going a little too far to say that compact objects comprising a substantial mass are ruled out or even that molecular hydrogen is ruled out as significant. Estimates of molecular hydrogen are largely determined using tracers like CO and HI whose emissions are much easier to detect. Often estimates of H mass are based on HI emissions with the assumption of little or no corresponding H2.

Cougar stated: "But observed gravitational effects suggest there must be 5-10 times more matter in galaxies than the baryons we can detect."

I'm not sure that this is correct statement without various assumptions, can you site a reference? Some dwarf galaxies have far more "dark matter" than that (e.g. 100 times visible), and some large galaxies require relatively little additional mass to explain their rotation curves (or velocity dispersion). Typically these estimates are based on the assumption that the dark matter must be distributed in a roughly spherical halo with a power law density. In that case more mass is needed than if the dark matter is distributed in the plane of the disk.

[SRH]

Why is the mass distribution generally thought of as a spherical halo?

I would think that the missing mass would need to be at the edges of the disk if we are trying to account for the speed of the outer spiral arms.

If the missing mass was in the center of the disk, for example, the predicted outer arms would still rotate slower than the inner spiral arms, no?

[Shaula]

Why is the mass distribution generally thought of as a spherical halo?

Dark matter is thought to be warm (i.e. the particles have a fair bit of Kinetic energy) and unable to cool very easily (as it only really interacts via the weak and gravitational forces).

I would think that the missing mass would need to be at the edges of the disk if we are trying to account for the speed of the outer spiral arms.

If the missing mass was in the center of the disk, for example, the predicted outer arms would still rotate slower than the inner spiral arms, no?

It does not work that way. Mass outside a shell does not affect the matter inside it gravitationally, when you do the maths the gravity inside a hollow sphere is zero everytwhere.

The trick is that the halo is huge. The rotation curves look flat simply because they are embedded in a large ball of matter we cannot see which has a density that drops off far more slowly than that of visible matter. This flattens the curve.

Also not that the outer objects do rotate slower than the inner ones. The curve does drop off, just not as fast or with the form we expect from the gravitational effects of visible matter.

[Cougar]

Cougar stated: "But observed gravitational effects suggest there must be 5-10 times more matter in galaxies than the baryons we can detect."

I'm not sure that this is correct statement without various assumptions, can you site a reference? Some dwarf galaxies have far more "dark matter" than that (e.g. 100 times visible), and some large galaxies require relatively little additional mass to explain their rotation curves (or velocity dispersion).

Sure, mine was a general statement, and there are exceptions. I've heard in the past that the apparent dynamics of large elliptical galaxies don't seem to require as much dark matter as, for example, spirals. However, large ellipticals are formed by mergers, and mergers are going to perturb existing near-circular orbits into more elongated orbits. Objects in elongated orbits are going slower than their average speed when they're far from the center of the galaxy. Dekel, et al. simulated the situation and found that these elongated orbits, with their slower outer trajectories, only make it look like there's less dark matter than we would expect:

Using numerical simulations of disk-galaxy mergers, we find that the stellar orbits in the outer regions of the resulting ellipticals are very elongated. These stars were torn by tidal forces from their original galaxies during the first close passage and put on outgoing trajectories. The elongated orbits, combined with the steeply falling density profile of the observed tracers, explain the observed low velocities even in the presence of large amounts of dark matter.

As I've always said, a single paper does not firmly establish a proposition. But this one makes sense to me.

[Grey]

Why is the mass distribution generally thought of as a spherical halo?

In addition to Shaula's point from a theoretical perspective, we also have observations of the motions of globular clusters, which tend to be in a roughly spherical distribution. Their motion is more consistent with a spherical distribution of mass than with mass concentrated in a flat disk like the visible matter.

I would think that the missing mass would need to be at the edges of the disk if we are trying to account for the speed of the outer spiral arms.

If the missing mass was in the center of the disk, for example, the predicted outer arms would still rotate slower than the inner spiral arms, no?

You've got the right general idea, that the distribution of mass is just as important as the amount, and that a large mass concentrated at the center wouldn't solve the issue. You can work out the distribution required to account for the roughly flat rotation curves seen in galaxies, and it turns out to be pretty much spherical, with a mass distribution that falls off like the inverse square of the distance from the center. Conveniently, it also turns out that if you do some basic kinematic analysis on a collection of particles that are gravitationally bound, but don't otherwise interact much, an inverse square mass distribution is exactly what you end up with. So that means that if there are some hypothetical particles that interact only through gravity (and maybe the weak force), it is reasonable for them to behave in a manner that would account for what we see.

[TooMany]

Using numerical simulations of disk-galaxy mergers, we find that the stellar orbits in the outer regions of the resulting ellipticals are very elongated. These stars were torn by tidal forces from their original galaxies during the first close passage and put on outgoing trajectories. The elongated orbits, combined with the steeply falling density profile of the observed tracers, explain the observed low velocities even in the presence of large amounts of dark matter.

So are they claiming that the stellar motions in ellipticals are not virialized because the mergers are too recent?

[TooMany]

You can work out the distribution required to account for the roughly flat rotation curves seen in galaxies, and it turns out to be pretty much spherical, with a mass distribution that falls off like the inverse square of the distance from the center.

I believe that this statement is false. While it is true that a spherical distribution can produce a flat rotation curve, there is no unique distribution required to produce a flat rotation curve. There are disk distributions which will also result in a flat rotation curve.

[Jeff Root]

Using numerical simulations of disk-galaxy mergers, we find
that the stellar orbits in the outer regions of the resulting
ellipticals are very elongated. These stars were torn by tidal
forces from their original galaxies during the first close passage
and put on outgoing trajectories. The elongated orbits,
combined with the steeply falling density profile of the
observed tracers, explain the observed low velocities
even in the presence of large amounts of dark matter.

So are they claiming that the stellar motions in ellipticals
are not virialized because the mergers are too recent?

It reads to me that they're saying the stellar motions
*are* virialized.

-- Jeff, in Minneapolis

[Grey]

I believe that this statement is false. While it is true that a spherical distribution can produce a flat rotation curve, there is no unique distribution required to produce a flat rotation curve. There are disk distributions which will also result in a flat rotation curve.

I don't think that will work for cases where we've got measurements of globular cluster motion. They tend to be distributed more or less spherically, and so provide an effective method of investigating the distribution of dark matter out of the disk. While it's certainly true that there may be a variety of distributions that will work (and different galaxies may have different distribution profiles, of course), I believe that in cases where we've been able to specifically look, a spherical distribution fits better than a disky distribution.

[TooMany]

I don't think that will work for cases where we've got measurements of globular cluster motion. They tend to be distributed more or less spherically, and so provide an effective method of investigating the distribution of dark matter out of the disk.

I'm not sure what your point is here. In LCDM theory, DM halos must be roughly spherical because the particles do not interact significantly except through gravity.

Apparently globular clusters have little or no dark matter:

Palomar 13 is the only known globular cluster with possible evidence for dark matter, based on a Keck/HIRES 21 star velocity dispersion of σ = 2.2±0.4 km s−1.
We reproduce this measurement, but demonstrate that it is inflated by unresolved binary stars.
...
We conclude that, while there is some evidence for tidal stripping at large radius, the dynamical mass of Palomar 13 is consistent with its stellar
mass and neither significant dark matter, nor extreme tidal heating, is required to explain the cluster dynamics.

While it's certainly true that there may be a variety of distributions that will work (and different galaxies may have different distribution profiles, of course), I believe that in cases where we've been able to specifically look, a spherical distribution fits better than a disky distribution.

I'd like to see the evidence for that claim. Most papers do not even consider a disk distribution of dark matter. Disks distributions are substantially more complex to analyze because they lack spherical symmetry. Also, the mass might be distributed within a disk in various ways, adding to the difficulty of concluding that a "spherical distribution fits better".

In fact there are known serious problems with the idea of interactionless dark matter halos; one is called the "cuspy halo problem" and another is the Baryonic Tulley Fisher relationship (or disk/halo conspiracy). Recently DM theorist have been trying to solve these problems by arguing that they are side effects of supernovas.

[SRH]

It does not work that way. Mass outside a shell does not affect the matter inside it gravitationally, when you do the maths the gravity inside a hollow sphere is zero everytwhere.
.

if that were true, then if you lived in hollow spherical house on the Earths surface, you would float. no?

[Shaula]

if that were true, then if you lived in hollow spherical house on the Earths surface, you would float. no?

The gravity from the house would be zero. Maybe I didn't make it clear - spheres do not have magic gravity shielding powers. What I was saying is that inside a hollow sphere the gravity from that sphere of matter is zero.

[SRH]

oh - i get it now

just to be clear, regarding the distribution of DM mass, that means that the DM "spherical halo" is actually a shell and not a uniform-density sphere. correct?

[Shaula]

Sorry, I was sloppy in my explanation. I should have been clearer and given a link that did a better job than I did. See here

[Grey]

I'm not sure what your point is here. In LCDM theory, DM halos must be roughly spherical because the particles do not interact significantly except through gravity.

Apparently globular clusters have little or no dark matter

You misunderstand. Not looking for dark matter in the globular cluster itself. Looking at the motion of globular clusters around a host galaxy. Since they're not confined to the disk, that gives you another tool to probe the distribution of dark matter of that host galaxy. In particular, while you can construct a disky mass distribution profile that would have the same net effect as a spherical distribution for the stars in the disk, it won't generally also have the same effect on objects located far from the disk.

I'd like to see the evidence for that claim. Most papers do not even consider a disk distribution of dark matter. Disks distributions are substantially more complex to analyze because they lack spherical symmetry. Also, the mass might be distributed within a disk in various ways, adding to the difficulty of concluding that a "spherical distribution fits better".

I'll see what I can dig up. I'm remembering this from a course on galactic structure some years ago, so it might take some time to track down again. Bear with me.:)

In fact there are known serious problems with the idea of interactionless dark matter halos; one is called the "cuspy halo problem" and another is the Baryonic Tulley Fisher relationship (or disk/halo conspiracy). Recently DM theorist have been trying to solve these problems by arguing that they are side effects of supernovas.

Sure, there are plenty of discussions about the details. The current view is still that CDM is the best explanation for the observations.

[TooMany]

Sorry, I did misunderstand what you were getting at. I don't really know how much we can conclude from the motion of globular clusters. I think if they provide truly convincing evidence of a spherical distribution I should have encountered that by now.

The current view is still that CDM is the best explanation for the observations.

It may be the best explanation, but that doesn't prevent it from being wrong. (Once upon a time the best explanation for day and night was that the sun revolves around the earth.)

[StupendousMan]

I don't really know how much we can conclude from the motion of globular clusters. I think if they provide truly convincing evidence of a spherical distribution I should have encountered that by now.

I don't really know ...

Perhaps you might go to ADS and search for papers which use the motions of globular clusters and satellite galaxies to determine the mass distribution in galaxies, then read those papers.