Got Gassy Galaxies ? Get Gasisgone !

What happens if your galaxy is feeling a little... bloated ? No problem ! Simply pop it into a galaxy cluster and let it dissolve slowly for a few billion years.

Say what you like about AI, it has its uses.

Okay, time for the public-outreach version of my latest paper. I'm gonna do my very best to keep this one to a readable length.

Introduction : How To Do Body Shaming For Galaxies

Back in 2015, astronomers got a nasty shock. Most galaxies, we thought, all had roughly the same surface brightness level, having about the same number of stars per unit area. Oh, sure, they varied a bit, of course, and there were a few oddball hipster galaxies that just insisted on doing things differently. But most of them were basically the same.

Then along came shiny new telescopes which were much better at finding fainter galaxies. Lo and behold, it turned out that there were plenty of such objects to be found. Now if they'd been really small tiddlers, I don't think anyone would be surprised. Dwarf galaxies do all kinds of crazy stuff anyways : we already know we don't fully understand how they work.

Like this guy, caught in the middle of two giants and causing a right mess. In some cases, interacting giants can rip off so much material from each other that whole new dwarf galaxies form. 

But the new galaxies weren't dwarfs : they were, by some measures, as large as giants. The stock phrase – and it's a decent one – is that they're the same size as our own Milky Way but a hundred to a thousand times fainter. And this is really weird, because everyone was quite happy we at least understood the basics of the biggest galaxies. The implication that maybe we didn't made a lot of people quite worried. Well, anyone would be if they suddenly realised they were surrounded by hundreds of invisible giants.

The immediate question was : ah well, yes, they're very big, but are they heavy ? Our models are mainly based on mass, not size. So maybe they're just like Zeppelins : enormous, but they might not weigh very much. And just like Zeppelins, they'd be spectacular but largely useless (or at least irrelevant for our models), and we could write off the whole thing as a Hindenburg-like tragedy except without all the horribly fiery death.

Too soon ?

Then things took a very confusing left turn. More than that, the world of extragalactic astronomy went down blind alleys, into ditches, and got thoroughly lost. We still don't know where exactly we are or where we're going, much less how to get there, but as best as I can tell, the current state of affairs is something like this : 

Most UDGs probably are those over-inflated dwarfs. They don't have much mass, but they spread it around a lot, taking up two seats on a plane not because of poor lifestyle choices or unfortunate physiology, but because they can't stop flailing their arms and legs everywhere. All their stars are very spread out, but their total mass – mainly dark matter, which we measure by seeing how fast their gas and stars are moving – is actually quite modest.

Manspreading is also surely even worse when you're a spindly fellow, and it certainly doesn't help matters if you're a contemptible fuckwit either.

That sounds not too bad, and indeed it isn't. Just like real life, body shaming the UDGs has come back to bite us, because they're (mostly) nowhere near as heavy as we might have guessed. Why, then, are we still confused ? That's because there are two major headaches nobody has yet solved. 

First, at least some UDGs are still plausibly very massive galaxies after all, which we don't know how to explain. Second... some of them seem to have far too little mass. Absolutely nobody expected this, and what's worse is that we see this even in isolation. Some of the first, most widely-reported examples of under-massive UDGs were found in groups, and these we think we can explain quite well as being the result of interactions with other galaxies (though even this isn't 100% proven). But in isolation this simply doesn't work at all, and nobody predicted galaxies lacking dark matter would ever be much of a thing. That's why we're still bloody confused.

So, too massive ? Nothing special ? Not massive enough ? The reality is likely a mixture of all three, and we're still trying to find our feet. 

If you want a more detailed introduction, I did two much longer write-ups on this. This post looked at the initial discovery while this one looked more at the first results of measuring the masses of UDGs. We still don't have much data on that. And there'll be a bunch more links throughout for those interested in more up-to-date results.

One other major issue is that we know about lots of UDGs in clusters but relatively few in the field (a catch-all term that basically means "not clusters"). So it's possible that the situation isn't so bad. Maybe most of the (more numerous) cluster-UDGs are indeed just normal dwarfs after all, but inflated by interactions with the other galaxies so they look bigger. The field objects without much dark matter could just be crazy weirdos, exotica that are extremely interesting but not actually that important in the grand scheme of things. After all, a few lunatics don't tell you much about psychology more generally, thankfully.

Essentially, we know of a few crazy galaxies outside clusters, but just a few. We have no idea of the far more numerous objects within clusters, but if they're as weird as the ones outside... well then things get freaky.

What we need, then, is to find a UDG just entering a cluster which still has its gas so we can measure its dynamics. That would help tell us if cluster UDGs are basically typical or typically weird.

We Need More Data

Which is where the current paper comes in. To be fair, there have been a few cases of estimating the masses of cluster UDGs before*, but very few indeed with data from their gas – quite possibly at the level of low single-figure numbers**. The rest have had to be done almost entirely by less direct methods, making clever inferences from globular cluster numbers and suchlike. It's very clever, but also unsatisfying, a bit like guessing someone's bank balance by the size of their house. It's nobody's preferred option. 

* One of which seems to be overly-massive for its size, but far short of being a true giant. UDGs really do seem to probe the full parameter space of weirdness.
** The paper has more on this in the Discussion section, but basically I only know of two, and even these weren't set out explicitly as UDGs.

This lack of data isn't surprising. Measuring how fast the stars are moving is extremely difficult when there's hardly any stars to measure, and as galaxies enter clusters they can lose their gas very quickly. Even if UDGs have relatively normal masses, it's not that strange that so few have been detected with gas while in clusters. 

So at last, in this paper we present a candidate for a UDG entering a cluster and still retaining most of its gas. And it too seems to have a depleted dark matter content.

Here's the data !

Credit for this discovery goes to my PhD student, who found it at the southern edge of our Widefield Arecibo Virgo Environment Survey (WAVES) data set. This is just north of the Arecibo Galaxy Environment Survey (AGES) data that I studied for my PhD, and there's a little bit of overlap – in fact it's found in both. Now characterising a galaxy as a UDGs is quite difficult since it needs such precise measurements, and while we're working on doing this for ourselves, we began by looking at a pre-existing catalogue of UDGs called SMUDGES. This one, VCC 1964*, was the only one found in our HI (atomic hydrogen) data sets.

* The name is from a catalogue from 1985. Some UDGs were found well before the big kick-off in 2015, which is when people began to recognise them as a distinct class of object.

Well then, here it is :

The contours show the gas, coloured according to how fast it's moving. The galaxy is relatively small, but you can see it more clearly in the inset image on the left. The inset image on the right shows the same galaxy as seen in ultra-violet, which usually corresponds to emission from young stars. Finally, the white arrows show the directions to two of the most massive galaxies in the cluster, and the green circle just shows the resolution of the atomic hydrogen (HI) data.

What we see here is immediately quite interesting. Normally the gas is centred almost exactly on the position of the stars (white X), but here there's a clear offset. Moreover, this same offset is seen in both AGES and WAVES data sets, so it's definitely real. The orientation of the gas and the stars is a bit weird here : we'd expect the gas to be further away from the cluster centre than the stars, but it's actually doing the opposite. That strongly suggests the galaxy has a weird orbit.

Much more fun, though, is the mass of the galaxy. The gas tells us how fast stuff in this galaxy is moving around, and the answer is... not fast enough. For this we use my favourite way to plot galaxies : the Tully-Fisher Relation.

Black circles (filled and open) show normal galaxies. The orange points show VCC 1964 as measured in a couple of different ways, while the grey points show the famous ALFALFA detections. The dashed and dotted lines show the observed scatter, with the uppermost dotted line showing five times the standard deviation – the usual criteria for judging something to be significant.

You might remember my extremely detailed run-through of how this is plotted, but no need to go through all that again. Basically it just shows the total mass of gas and stars (vertical axis) as a function of how fast the galaxy is rotating (horizontal axis). This object, much like some other UDGs with gas measurements, is rotating much more slowly than normal galaxies – suggesting it's got significantly less dark matter.

You might also notice that the mass of this galaxy appears to be much less than the other UDGs plotted. In fact it turns out to have the lowest mass of gas detected in a UDG to date, with the next highest being something like a factor five times more massive (and most much more than this). But can we then trust these measurements ? With so little gas, how can we be sure we're really measuring its rotation speed accurately ? Especially since the gas appears to have been pushed out. As someone said during the analysis phase, you can claim the gas has been removed, or you can claim an offset from the Tully-Fisher, but can you really claim both as important at the same time... ?

That is definitely the most uncertain bit of our analysis. We've found previously that galaxies do deviate from the Tully-Fisher simply when they've lost enough gas, with no implications for their dark matter at all. That can happen because the outermost gas is both the most easily removed and rotating the fastest. Strip this away and the measured rotation speed will shrink, but this doesn't tell you anything about the dark matter content at all.

A simple simulation of ram pressure stripping. Dark matter (not shown) is completely unaffected. But since we infer dark matter by the rotation of the gas, if the fastest-moving gas is removed, it can look as though the galaxy has less dark matter than usual.

It's certainly possible this is the case here, but there are mitigating factors. We found that large deviations only happened when galaxies had lost a lot of gas, but in this case the gas loss only seems to be very modest : actually what we're seeing is more gas displacement than loss. All of the gas appears to have been shifted in bulk away from its happy place; effectively all of the gas should count as the green "disturbed" component in the above diagram. 

This makes the interpretation tricky. We can say quite confidently that the rotation hasn't been reduced just due to sheer loss of the fastest-moving gas. The problem is that we have almost no clue what happens when you get a wholesale bulk movement of the gas like this – it might disturb the gas in other ways we haven't accounted for. There just aren't nearly enough other examples like this for us to make a comparison.

And we also tried another plot of the TFR, this time not using total mass but just optical brightness. This is subject to less corrections that have to be applied to the data, and we found an even stronger deviation : at least six times the scatter in the normal galaxies, and possibly even more than this. So two different plots by two different methods gave us the same result, which is pretty neat. And in a few other cases, it's been shown that the velocity dispersion actually increases when gas is stripped, not decreases. 

What's Going On ?

VCC 1964 has a great deal of collective weirdness :

• Its gas is displaced from its stars, but little is actually missing : it has the lowest mass of gas in any UDG detected to date, but that's consistent with it being a right little tiddler (not because a lot of gas has gone missing).
• The offset between the gas and stars suggest a weird orbit, with most of the motion across the sky rather than along our line of sight.
• The apparent rotation speed is quite a lot lower than expected. This can't be accounted for by gas loss or measurement errors, but we don't know if gas displacement by itself could also cause this effect.
• And what I haven't mentioned is that this galaxy is both smooth and blue. This combination is itself very unusual : more often, blue galaxies have lots of structure and star formation. It might be related to the galaxy just having lost all of its gas, with enough time for the structures to smooth out but not enough for the colour to change. Maybe. It's also strange that removing the gas, which is probably about as massive as the stars, doesn't appear to have altered the stellar structure at all.

What if it's all just a horrible coincidence ? Could the gas actually have nothing to do with the stars at all, with both at different distances ?

This has actually been proposed to be the case for another possible example of a UDG losing gas in the Virgo cluster. Personally I'm a bit skeptical about this, but it's possible in that earlier example. But to find two examples of UDGs in Virgo where the gas is just coincidentally aligned with some unrelated stars... nah, I simply don't believe it. And we looked very carefully to see if there are any other possible sources of the gas nearby and there just aren't. I tend to rule this idea out.

What we know with some confidence is that this is a UDG caught in the act of losing gas. That's already extremely unusual, and we must be seeing it at a very precise moment indeed for it to be both blue and smooth. What's much more uncertain is whether shoving the gas out of the galaxy would cause the illusion of a lower rotation speed : in other words, does it really not have much dark matter, or is this just because it just got kicked in the backside ? Having little dark matter would be consistent with other UDGs, and – maybe – also help explain how come the gas can been removed in bulk, wholesale, rather than dragged into a long tail as is more normal.

And even more speculatively... what does this mean for the screaming hordes of UDGs in clusters more generally ? Are they just giant collections of stars rather than more typical, dark matter-dominated galaxies ? We just don't know, and intuitively, it's hard to see how they could survive very long in a cluster... though it seems they do tend to avoid the densest parts of clusters, so maybe they simply don't. Maybe they just get torn to shreds. 

All we really know is that everything is very confusing.

The final question is whether objects like this are connected to the so-called "blue blobs". In my opinion these are some of the most interesting objects found lately, though I haven't talked about them much. They're... umm, well, they're fuzzy blue things which don't quite resemble ordinary galaxies, and there's been some very interesting suggestions that they might form in the gas stripped out of larger galaxies by ram pressure. They too would lack dark matter. So is there a connection ? Is VCC 1964 actually a very large blue blob ? Or will it even spawn a blue blob from its own stripped gas ? How do all of these relate to the dark galaxies I more usually go on about ?

Well, good news ! Our paper on using the VLA to get higher resolution on those dark clouds is damned close to acceptance. And we have more observations accepted with the VLA for VCC 1964 itself, which should give us a much clearer picture of what its gas looks like than the big blob we see in the Arecibo data. More research is needed... but this time, more research is actually happening.