Well, I think the only way to explain an anomalously large annihilation signal from dark matter is to assume that we actually just happen to be in an unusual region of the galaxy with a good amount of extra dark matter nearby. So I think the more likely explanation is that any such signals are just due to some poorly-understood astrophysical process.

Ok, that’s exactly the point, what about the bounds about collisionless DM versus the pamela results that require large cross sections (Somerfeld enhanced?). We are not in a particularly hot spot in the galaxy, so this is happening everywhere, how can it remain ‘colisionless’? can you throw out some numbers?

Must CDM be self-collisionless or could it simply be
reluctant to form small scale dark matter structures?

I’m pretty sure those are one and the same question. In order to collapse, it must emit energy in some way. And in order to do that, it has to collide with things. Obviously there are some limits as to just how collisionless it must be based upon our current observations of dark matter both in the early universe and today. For the possibility of a dark charge, you may want to check out Sean’s relatively recent blog post on dark photons.

A large pool of dark matter would not only slowly accrete
more dark matter, it would also sweep up normal matter,
which has probably been denser at small scales than DM
since BBNS. The denser material would sink into the pool.
If DM is allowed to become too dense, the normal matter
simply would sink into the centre of mass forming layers
of increasingly heavy atoms and nuclei. In the early days
of initial structure formation, that would mean isotopes
of H to Be and the products of huge pressures upon the
corewards layers (heavier elements, and heavy isotopes
that are stabilized by pressure).

Well, not quite. Big bang nucleosynthesis (BBNS) is pretty insensitive to the nature of dark matter as the universe just had hardly clumped at all back then. BBNS was driven almost entirely by the overall temperature and density, not the minuscule fluctuations thereof. What places the most stringent limits upon dark matter’s properties in the early universe is the CMB itself, which basically places an upper limit upon the temperature of dark matter (too warm, and you don’t get any structure, but it can be as cold as you like).

Could the annihilation question also be put as: if a dark
matter structure arose in a part of the universe with a
very small amount of visible matter, would it eventually
accrete sufficient dark mass to trigger ongoing
annihilations in the core of what by analogy could be
called a “dark matter star”?

Not sure. Depends upon the details of how quickly it loses energy to its environment. For example, if annihilations are sufficiently common that more energy is lost through annihilations than collisions, I suspect that the dark matter halo may never gain in density significantly except by the addition of new matter, such that it won’t ever achieve the densities required. If the collisions are common enough that they dump more energy into their environment than annihilations, then sure, I might imagine something like this happening, but it’d take a very long time.

Sure, but not at very high densities in wells like our
solar system, or surely we would have noticed it affecting
planetary orbits and the sun’s physics.

Right, the effect is pretty small, and you need very large amounts of normal matter for the overdensity of normal matter to count significantly. And yes, the converse is also true, that normal matter is more likely to be found in the more dense regions of dark matter. But the point is that the normal matter interacts much more strongly and tends to fall inward into the potential well, whereas the dark matter stays relatively stable.

Thanks. I have a couple dumb, unrigorously framed questions:

Right, the dark matter doesn’t collapse much
on its own, because it is almost collisionless. Once you
get a halo of dark matter, unless it interacts with
something external to it, it largely just stays as it is
for then after.

A DM halo obviously interacts with lots of things external
to it, at least gravitationally: the rest of the mass in
the structure the halo is part of, and mass external to
the structure.

Must CDM be self-collisionless or could it simply be
reluctant to form small scale dark matter structures?
Obviously it is much less dense than normal matter,
because it went unnoticed by classical astronomers like
Kepler and Newton (and pretty much everyone else until the
20th century). Perhaps there is a dark charge which works
against gravitational attraction analogously to electron
degeneracy pressure.

dark matter is still more likely to be found
in wells where normal matter has condensed into compact
objects

A large pool of dark matter would not only slowly accrete
more dark matter, it would also sweep up normal matter,
which has probably been denser at small scales than DM
since BBNS. The denser material would sink into the pool.
If DM is allowed to become too dense, the normal matter
simply would sink into the centre of mass forming layers
of increasingly heavy atoms and nuclei. In the early days
of initial structure formation, that would mean isotopes
of H to Be and the products of huge pressures upon the
corewards layers (heavier elements, and heavy isotopes
that are stabilized by pressure). Conversely, if DM is
forbidden from becoming dense enough, we have something
similar to the hot dark matter smearing problem of
attracting mass out of the centre of galaxy scale
structures. However at a “Goldilocks” temperature, the
outer layers of the structure would retain a lot of mass,
counteracting the infall of mass to the centre of the
overall structure, and would thereby enable the formation
of stable stars and star clusters.

This is why, for example, there have been some
suggestions that perhaps there’s some annihilation of dark
matter going on near the galactic core.

Could the annihilation question also be put as: if a dark
matter structure arose in a part of the universe with a
very small amount of visible matter, would it eventually
accrete sufficient dark mass to trigger ongoing
annihilations in the core of what by analogy could be
called a “dark matter star”?

dark matter is still more likely to be found
in wells where normal matter has condensed into compact
objects

Sure, but not at very high densities in wells like our
solar system, or surely we would have noticed it affecting
planetary orbits and the sun’s physics. There may be a
gentle density gradient within the solar system, but there
are tight limits on that from observations of
trans-Neptunian objects, right?

So, couldn’t one also say that hot, dense normal matter is
more likely to be found in wells where dark matter has
condensed into relatively dense structures?

Self gravitating globules could lead to DM particles in the centre of a sufficiently massive DM globule to be squashed into a short interaction range, analogous to degenerate phases in sufficiently dense collections of normal matter (as in degenerate dwarfs or neutron stars).

The problem is that the gravitational collapse of the DM globule should do work, but the DM should not radiate photons (and in CDM, it should not result in highly accelerated DM particles in a way that causes problems at the epoch of matter-radiation equality). Where does the energy go?

Right, the dark matter doesn’t collapse much on its own, because it is almost collisionless. Once you get a halo of dark matter, unless it interacts with something external to it, it largely just stays as it is for then after. Much of the annihilation, then, comes as a result of the normal matter, as though there’s much less of it around, it is far, far clumpier. The greater density of the normal matter, then, causes the dark matter to fall into the normal matter’s potential wells, producing more annihilations. This is why, for example, there have been some suggestions that perhaps there’s some annihilation of dark matter going on near the galactic core.

Bear in mind that even though there is no net work being done on the dark matter (to a very good approximation), the dark matter is still more likely to be found in wells where normal matter has condensed into compact objects.

Self gravitating globules could lead to DM particles in the centre of a sufficiently massive DM globule to be squashed into a short interaction range, analogous to degenerate phases in sufficiently dense collections of normal matter (as in degenerate dwarfs or neutron stars).

The problem is that the gravitational collapse of the DM globule should do work, but the DM should not radiate photons (and in CDM, it should not result in highly accelerated DM particles in a way that causes problems at the epoch of matter-radiation equality). Where does the energy go?

Alternatively, there could be interactions analogous to normal matter nuclear fusion, such as p-p fusion chains in stars, again with most of the energy dissipating through some mechanism other than photon radiation.

Either of those bug me a bit — wouldn’t we expect some ordinary visible matter to be swept up into the self-gravitating clump of DM? That matter would heat up and radiate photons, right? (And also absorb and reflect light from elsewhere).

Alternatively alternatively, these high energy gamma rays could be because of DM particle decay rather than annihilation; I think this would be allowed by WIMPy LCDM models.

Alternatively^3, this could be ordinary-matter pulsar physics, such as primary production of e+ within the pulsar or secondary production in nearby ordinary matter (in collsion with high energy nuclei ejected from the pulsar) with the pulsar’s magnetic field accelerating those secondary positrons into high energy jets.

I could of course be completely out to lunch, and welcome corrections from people who aren’t.

Their collisionless property is only an approximation. It is, however, exactly because this is a very good approximation that we expect there to still be lots of anti-particles remaining for dark matter (whereas the much stronger-interacting normal matter we have today is entirely devoid of anti-particles, except those created at later dates). So, they don’t interact much, but every once in a while you still get an annihilation when a dark matter particle manages to smack into its anti-particle. It’s not going to happen often, naturally, as they don’t experience much of any attraction (mostly just gravitational, and that’s extremely weak between individual particles). But there’s a lot of dark matter out there, so we expect the signal should be measurable.

* guarantee void if falsified.

This all looks really exciting. As the PAMELA results suggest DM at the TeV range it will be interesting if some channel production turns up in the LHC.

Lawrence B. Crowell