Escaping the dark

We now return to those dark galaxies without dark matter. No, not those ones, the other ones. The originals.

This is a topic so controversial that even those claiming to have explained them say that the original discoveries were wrong, which is very confusing. But the best bet seems to be a tidal encounter. Since dark matter, like gas, is generally much more extended than the stars, it's easier to remove. Indeed, unlike gas, dark matter is collisionless, so if it happens to be in the centre of a galaxy on part of its orbit, there's nothing stopping it from moving out to greater distances a little later. Whereas the innermost gas gets trapped in the office forever, dark matter prefers a hybrid model : sometimes it's in the city centre, but sometimes it's in the vulnerable suburbs.

Topical metaphors aside, today's paper is the most convincing argument yet for a tidal origin. The authors re-examined an existing simulation that predates the whole hoo-hah, so this absolutely counts as a prediction. True, it would have been even better to predict these things ahead of the discovery itself, but there was no reason for anyone to look for them.

They find seven objects in their 21 Mpc-on-a-side simulation that lack dark matter. Most of them live in groups (three are all in the same group), and all are associated with a massive parent - more massive than the Milky Way. Now the properties of all but one are not all that similar to the notorious NGC 1052 DF2/DF4 (they have a few times as much stellar mass and, correspondingly, rather higher velocity diserpsions*), but this would probably be asking too much of a simulation. The main point about lacking dark matter is clearly satisfied, and they're not orders of magnitude different to the observational sample.

* These two parameters are intimately related though. The higher the stellar mass, the higher the velocity dispersion must be for the systems to be in stable equilibrium.

What I find most convincing about this is that they explain their objects as a rare but not exceptional process. Since there are at least two objects known in one group, and more dark matter free objects elsewhere, any explanation that relies on a freak encounter might have problems. It's a very tricky balance to find a mechanism that is routine enough that you actually expect to stand a reasonable chance of detecting the end results, but not so common that they should have been spotted everywhere and known about for decades.

The author's answer to this is that the satellites need to be very small in comparison to their parent galaxy (0.1% of its stellar mass) and come extraordinarily close - within the stellar body. And this seems to fit the bill quite nicely. One would expect most such encounters to result in the satellite merging with the cannibalistic parent or being utterly destroyed by it, but every so often one could just manage to survive. And another satisfying point is that while the after-effects of the encounter are still visible in some of their simulated objects, some show no visible signs of disruption at all : the main body of the galaxies survive long after the tidal tails have dispersed.

If anything, I worry that this explanation might overpredict the number of dark matter deficient satellites : around 30% of the most massive galaxies ought to have at least one such companion, they say. But such massive galaxies are themselves quite rare, and the smallest satellites are the hardest to measure, so this is not necessarily and outlandish fraction. Importantly, it's testable. We probably shouldn't get too hung up on the exact numbers given the limitations of observations and simulations alike - so long as it turns out these systems are not unique to NGC 1052, that's probably sufficient.

So this is very good stuff. What remains now is to look for analogues to those other galaxies without dark matter : the isolated, gas-dominated Ultra Diffuse Galaxies. Those, they say, might exist in their simulations as well... but that's another story.