Or are they ?
Today, two papers on my favourite science topic of all : dark galaxies. In the past there have been a multitude of candidate detections but spread out very thinly. You get, I'd guestimate, of order one or two such claims per year on average, with the total number now being somewhere in the low to mid tens. And not a single one is entirely convincing. Every single object is essentially unique, with its own particular considerations that make it more and/or less likely to be a dark galaxy.
Both of these papers claim to have alleviated the problem by finding a whole bunch more candidates. The first uses new data from the ASKAP telescope and comes up with 55 potential objects, while the second uses archival Arecibo data and finds 142. Impressive stuff – but are any of them plausible, or have the previous problems just reappeared in a larger sample ?
There are many difficulties with identifying a dark galaxy candidate. The resolution of radio telescopes that can detect their gas content is often much lower than optical instruments, which means you see a big blurry smudge on the sky. That makes pinpointing the exact position of the gas difficult, so it's hard to say whether it has an optical counterpart or not. It also makes estimating its total mass tricky : for this you need a precise measure of its size, so without it you can't really say how much dark matter it really has. And even if you do have good resolution, you need good optical data as well to say if it's really dark or just very dim (though when you get to sufficiently dim objects the difference is arguably not that important).
An even bigger problem happens when you manage to overcome all this. Even if you have an isolated gas blob with the signatures of stable rotation that would need lots of dark matter to hold it together, and even if you're darn sure it's so optically faint that it might as well be dark... it's damn hard to say if the thing really is stable. You could just be seeing a bit of fluff leftover from some interaction or other, which can sometimes mimic the appearance of a dark galaxy. Nevertheless, there have been a few cases where "dark galaxy" at least looks like a very plausible explanation, if never any where we can be certain that's really what's been found.
Both of the papers attempt to do much the same thing though in slightly different ways. Starting with large HI samples (30,000 for ALFALFA and 2,000 for WALLABY) they combine this with optical data sets and trim them down in various ways : quality of the HI signal, confidence in the lack of optical counterpart, isolation, etc. ALFALFA (the Arecibo data) has an enormous area of coverage and huge sample size on its size, while WALLABY (from the ASKAP telescope) has higher sensitivity and resolution.
Since even the final candidate catalogues are, by the standards of dark galaxy research, really quite large, I'd be reluctant to say, "yep, this is definitely the solution, hurrah chaps, we've found them !". But nor would I at all dismiss them out of hand. Rather I would look at both of these papers as being potentially the foundation of interesting research, but it's too soon for any definitive results yet. These are both very solid starts, but we need to examine each and every object here in more detail, or at least a subsample. We need higher resolution data in all cases, deeper optical data... and most importantly, detailed studies of the local environment. We need to find the quintessential case of an isolated object with no plausible other origins, preferably rotating nice and quickly (which would mean fast dissipation if it wasn't bound by dark matter).
All that requires very careful, detailed work. Which of course we can now do, so kudos to them for that. But scientifically I'm neither excited nor dismayed. I am... intrigued.
The first paper finds its dark galaxies pretty much everywhere throughout its fields. There's not really any distance bias, so they occur at all masses – a few at really quite respectable standards even when compared with optically bright galaxies. Line widths look to be typically around 100 km/s, which is where we'd naively expect rotation – and hence a dark matter component – to be needed for stability. Sadly the resolution isn't good enough for them to attempt dynamic mass estimates, though this seems to me a bit strange – they have the upper size limit from the HI, so they could at least put a broad constraint on it.
The other oddity is that they model the optical light profile of all their sources, where detected. This is ideal for quantifying whether any are Ultra Diffuse Galaxies (which are possibly closely related to truly dark galaxies) but they don't seem to do this. Maybe that's for a future paper.
The second paper attempts a lot more science. I have to say it's both strange and refreshing to see a member of the ALFALFA team being at least a little more enthusiastic about dark galaxy candidates; normally they insist on calling them 'almost darks' – including the quotes – which gets very annoying. None of that here ! I should stress, though, that both papers absolutely treat everything with the caution it deserves, so don't mistake the brevity of my summary as evidence that they leap to conclusions. Neither group does that – I'm omitting the caveats just to get to the point.
Which for this second paper is as follows. As per the first, their candidates are everywhere, spanning a wide range of masses and line widths, but generally found in less dense environments than bright galaxies. They have higher gas fractions (relative to their inferred dark matter masses*) than optically bright galaxies of similar masses. And these properties are qualitatively similar to what's found in numerical simulations of galaxy formation that produce dark galaxies.
* Being a bit more gung-ho than the first group, they assume a size of the galaxy based on the scaling relation with respect to HI mass, hence they get a dark mass estimate.
All this is very matter-of-fact, commendably so. It's a huge sign of much how things have changed in the last couple of decades : when I went to my first conference, back in 2007, dark galaxies were viewed by many as... not exactly fringe, but not really mainstream either. Most people agreed that they could at least exist, but were skeptical of their whole raison d'être – that they would be numerous enough to explain why cosmological models were massively overpredicting how many galaxies we would see. Indeed, for the next few years if often felt as if hardly anyone really believed in the standard models of galaxy formation, even if nobody had any better ideas to replace it. Quite frankly, if anyone had suggested they'd found a hundred or more dark galaxy candidates, no matter how cautiously, they'd have been laughed at. It wouldn't have been a career-ending move but it wouldn't have won them any friends either.
All that seems to have largely faded. The original models of galaxy formation, where gas falls into dark matter halos and a bunch of complicated stuff happens, now seem very much more popular, and so dark galaxies no longer seem like an almost dirty subject. What's happened is that we've got a lot better at doing all that complicated stuff and many of the problems which looked horrendous now look, if hardly definitely solved, then at least an awful lot more solvable.
So, good work people. It's going to be extremely interesting to see how this pans out over the next few years. Watch this space.
