Earlier questions (e.g. How does Dark Matter form a halo?) referred to the distribution of dark matter in the galaxy. My question is different (not a duplicate). I wonder why dark matter would form a halo around a galaxy at all.

Dark matter does not interact electromagnetically and therefore does not lose its kinetic energy, at least not nearly as much as normal matter. Therefore, a flying by particle of dark matter that is not on an orbit around a galaxy may change its direction, but will still fly away. In other words, dark matter that is not in the halo of a galaxy already will not be in this halo, unless somehow slowed down.

So, how does a halo form in the first place? Does a forming galaxy simply attract the dark matter that happens to be local there? Or is dark matter a decay product of normal matter in the galaxy?

Image credit: Wikipedia

enter image description here


There are two aspects to this, one straightforward and one rather more speculative.

The straightforward one is that dark matter is weakly interacting but not non-interacting. In particular dark matter particles will undergo dynamical friction just like every other gravitationally interacting material. In effect a dark matter cloud cools by evaporation. When particles interact gravitationally a small proportion will get a high momentum just by chance, and those particles will escape the cloud carrying away excess momentum with them.

The rather more speculative bit is how primordial density differences were created in the first place, and this is normally attributed to fluctuations during inflation. The fluctuations measurable in the CMB are consist with this mechanism, though the whole question of inflation seems to be getting a bit vexed these days.


Having rushed into print I now see there is a closely related question How can dark matter collapse without collisions or radiation? that is possibly a duplicate.

| cite | improve this answer | |
  • $\begingroup$ Can you provide a reference on inflation being disputed? I wonder what the arguments are. Thanks! $\endgroup$ – safesphere Nov 4 '17 at 6:12
  • $\begingroup$ @safesphere: this is a good place to start though note that Hossenfelder is on the sceptical side so apply the usual pinch of salt. Follow the links in her blog post for a broader perspective. This is one of those arguments where the issue is conveniently free of any experimental data that might spoil the fun :-) $\endgroup$ – John Rennie Nov 4 '17 at 6:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.