Phwwwwwhoopslsssssh !

That's the sound a galaxy would make as it falls into a cluster. Pressure from the hot cluster gas builds up until the galaxy's own gas disc can't take it any more and it streams out into a disintegrating wake, filling the cluster with hundreds of little gas clumps.

Explaining the origin and nature of certain little clumps is a long-running obsession of mine, particularly ones that look like they're rotating. Directly measuring rotation isn't possible, but we can compare what rotation and other motions look like with the velocity field of the object. Currently the alternative explanations aren't doing very well. If they are rotating, then the clouds would need a lot of dark matter to avoid going splat. That would make them genuine dark galaxies and would make a lot of people very upset.

This paper simulates the effect of ram pressure stripping on a set of galaxies falling into a cluster, and finds that they produce quite a lot of little clumps as the gas from the galaxy and cluster mix together. Their numbers are quite high, but very variable - typically 50-150. The tidal encounter simulations we were playing with were giving much lower numbers. It seems that ram pressure is a much more efficient way of producing clumps, which is very interesting.

Except, of course, we need to have a fair comparison...

The models in this paper are a major improvement over the simplified ones I was doing. The most important difference is that they have gas in both the galaxies and in the cluster, which is easy with grid-based hydrocodes but very computationally expensive with the particle models I was running. So our models just had gravitational interactions, whereas we think (with some confidence) ram pressure is the major mechanism of gas loss. We showed that tidal effects were important, but we were still missing the most important process.

What's also really neat is that they simulate a bunch of galaxies falling in all at the same time, whereas ours only had one galaxy at a time (again because of computational cost). I would wound small children for the chance to combine gravitational and hydrodynamic physics. Unfortunately, as the author pointed out to me when I emailed him, this isn't really the case yet. The density of galaxies in the simulation is just too low - none of them have close encounters with each other. So this is a good approximation to a set of independent examples of pure ram pressure.

What they find is that the clumps can sometimes be so dense that they form stars, but not enough to make up a significant fraction of the intracluster light. Even these few star-forming clumps are very rare, only 3% of the total. That's interesting in itself, and predicts lots of little clumps that could be detectable with future (gas) surveys. And in reality there'd be at least an order of magnitude more clumps, since they're only using a dozen or so galaxies. But what about those "rotating" clumps ?

Well, it's kindof neat and unexpected how many dense clumps ram pressure produces, so that's cool. But on the specific sorts of clumps I'm interested in, they can't say much yet. The ones they form are generally much less massive, by about a factor ten, though they do have a few which are as massive. They report slightly longer lifetimes (100-300 Myr) than in our simulations (< 100 Myr) due to mixing with the intracluster gas. But the key point is the velocity structure of the clouds - can you form something which looks like it's rotating (but isn't really) by this mechanism ? Alas, they don't have the resolution to tell. Also, while the title says "molecular", they can't really distinguish molecular from atomic gas in these simulations.

I find the result that large numbers of clumps can be produced by ram pressure very interesting. Potentially, given a realistic number of galaxies, that might, and I stress might, be enough to explain the number of massive clouds that are detected in real clusters. However, it's not yet clear if such clouds ought to be embedded in long streams (unlike the real ones), and the crucial measurement of their dynamics is still unknown.

The research continues.
https://arxiv.org/abs/1807.09771