Follow the reluctant adventures in the life of a Welsh astrophysicist sent around the world for some reason, wherein I photograph potatoes and destroy galaxies in the name of science. And don't forget about my website, www.rhysy.net



Saturday, 10 August 2019

Tracking The Gassy Stripper Parent Of An Orphaned Rhino Through WAVES

There's a question in Cards Against Humanity, "What's that smell ?" to which the answer is always something hugely offensive. And in any large group of people, at some point everyone starts wondering not what the smell is but who it's coming from.

Radio astronomy has a similar problem, which is basically the same except in all the important points, and we aren't content to let such hugely important questions remain unanswered. We WILL find the source of that gas, dammit ! But sometimes, as in the latest paper on which I'm co-author, this can be surprisingly difficult...

Galaxies are not just a collection of stars swirling about in space like gigantic fidget spinners. In fact, the only two similarities they have to fidget spinners are that they're flat and spin by rotating around a central point. But galaxies don't rotate like solid objects, where each part takes the same amount of time to complete one revolution. Instead, stars at different distances take different amounts of time to complete one orbit. By measuring distances and speeds, we can work out how much mass is in the galaxy in order for it to hold itself together :


The answer to this generally turns out to be, "a buttload". The movie on the right shows a galaxy with about the same total mass as all the visible stars, with the one on the left showing one with a bunch of unseen "dark matter". And that's what observations show : they rotate way more quickly than if they were just made of stars (neither looks anything like solid body rotation either).

Of course, galaxies aren't just made of stars, that'd be silly. One of the other key differences between galaxies and fidget spinners is that fidget spinners don't tend to look different if you take a long exposure photograph - pointing your camera at it for a whole month won't make it look any bigger*. With galaxies it's different. The longer the exposure, the more stars you can see. This is difficult, but in general, deeper images make galaxies look a lot more interesting. One of the best known examples is NGC 5907, which has these spectacular double loops, likely produced by an interacting dwarf galaxy.

* That's what she said.

Though the existence of some of the features visible here is now somewhat controversial.
What about gas ? Here too things get more complicated. Gas comes in three different delicious flavours :
  1. Atomic. Just a bunch of atoms of different elements flying about.
  2. Ionised. Gas which has had its electrons stripped off, usually because of hot young stars who are willing to do anything for money which blast high energy photons through the gas.
  3. Molecular. When the atoms are nice and cool, they huddle together for warmth combine to form molecules.
Gas can be made of many different elements. One of the easiest to observe, partly because it's the most common, is atomic hydrogen. Observing this needs a radio telescope. Whereas the wavelength of visible light is measured in billionths of a metre, atomic hydrogen radiates at 21 cm. Catching waves that big takes a suitably gargantuan telescope and a lot of exposure time. Just like with optical images, not only do radio telescopes make maps (and not "listen" to the data cos that'd be stupid), but we also see that galaxies look different the deeper we go.

The nearby galaxy M33. The animation starts with an optical image and then overlays the gas at increasing sensitivity levels. Just like with stars, since what we see depends strongly on our precise observing methods, this makes data visualisation very important in how we interpret the results. That's why we're gonna see lots of different maps of the same object later.

How far does this go ? We don't actually know. Getting the above image took five years with the world's largest radio telescope, and it still doesn't seem that we've found the edge of the gas disc. All the really dense gas is in the middle, where the stars are, but while the density decreases with distance, it never seems to reach a definite edge.

In the middle stage of the animation, you can see there's a loop of gas extending north - evidence that this galaxy has probably interacted with... something... at some point. We don't know for sure what's causing it, though it could well be the Andromeda Galaxy M31. But other such loops are much, much bigger and more impressive. Here's a map of the Virgo cluster - the nearest rich cluster of galaxies to us, with over a thousand crammed into a volume not much bigger than our own Local Group :

Blue points show spiral and irregular galaxies while red ones show ellipticals (the faint green ones show especially faint galaxies). Arrows show known hydrogen streams. Considering how many galaxies are present, the number of streams isn't that large, but that's a topic of another paper. Just north of the grey shaded area you can see a feature labelled as "Kent 2009", which is one of the most unusual features known.

This giant gas cloud was discovered by a certain Brian Kent using the ALFALFA survey at Arecibo. He found that there's a collection of possibly linked clouds, spanning about 150 kpc (just shy of half a million light years) with around 500 million times the mass of the Sun in gas, with no obvious connection to any galaxies. The Kent complex* is big, but isn't the biggest or most massive known gas cloud - but that it's so completely detached is very unusual. The other gas clouds have their own mysteries, but at least we can usually identify the culprit. But for the Kent complex, it's very much a case of, "who let that one out ?"

* We originally called it this in the paper, but Brian doesn't want it named after him so we changed it. 

What we need is more sensitive observations. That's where the Widefield Arecibo Virgo Extragalactic Survey comes in, which is about four times deeper than ALFALFA. Just for the record, I suggested a bunch of other names but they were all rejected :
  • BESTEVER : Big Exciting Survey To Explore the Virgo Extragalactic Region
  • WIVES : Wide-field Interesting Virgo Extragalactic Search
  • ADHESIVE : Arecibo Deep Hydrogen Extragalactic Survey In Virgo Environment
  • VADER : Virgo Arecibo Deep Extragalactic Research
  • THIEVES : The HI Extragalactic Virgo Exploration Survey
  • WHATEVER (my favourite) : Wide-field Hydogen Arecibo Telescope Exploration of the Virgo Extragalactic Region
  • RIVETED : Region In Virgo Explored To Extreme Depth
  • DISHWATER : Deeply Interesting Survey of Hydrogen in a Wide-field Arecbio Telescope Exploratory Region
  • MOVES : Mother Of all Virgo Extragalactic Surveys
Oh well. Anyway, more sensitivity = more chance of finding fainter gas = more chance of finding the culprit. This works equally well if you go for Robert Minchin's analogy of gas streams as being the bloody entrails of galaxy interactions and want to hunt down the murderer, but farts are less disgusting.

It didn't work though. We did find more gas - quite a lot more, actually, bringing the total to about 1.5 billion solar masses. And we confirmed that the clouds are indeed all embedded in a common envelope of gas - they're not just a bunch of clouds that happen to be having a party. Rather, what Kent saw were the brightest parts of a single structure. Here's a comparison :

Left : Kent's original map. Right : our shiny new WAVES map.

Actually, in our data set the complex looks for all the world like a collection of galaxies, with this fainter emission being evidence of an interaction. But that's highly unlikely : there's no evidence of any optical counterpart at all, and Kent also did higher resolution observations which showed that no parts of the structure are rotating. So it's not a complex of interacting dark galaxies. At this point I'd usually show a movie, but nowadays I can go one better. Here's an interactive 3D model !


You can read more about these models here. I wanted to show more intensity levels than the three here, but isosurfaces in Blender involve a lot of manual work and I had problems with layers intersecting each other and it looked horrible.

Can't run it ? No problem, here's a really pretty renzogram from the paper. Each colour shows the hydrogen detected at a slightly different frequency, which is equivalent to velocity - and not the same as distance. So don't go thinking the 3D model shows the true shape of the cloud along the third axis. It gives very interesting information about how the cloud is moving, but not about its 3D shape.

So what is this bloody entrail/giant space fart ? Kent suggested it could have been stripped out of some of the nearest galaxies in this region. We didn't find any evidence of a direct connection, but it still seems that this is the most likely explanation. But it's not at all easy to identify who's responsible. Here's a map of the surrounding region :
Each galaxy has been coloured according to its average velocity. Galaxies we didn't detect are shown in ellipses. The thin grey lines show the edge of our survey region.

And note that in this particular rendering, it looks quite a lot like a space rhino :

Given the rainbow colour scheme in the renzogram, this is clearly a giant gassy space unicorn. Real unicorns have curves and are gassy. #Diversity
All of the other known streams are connected to their parent galaxies in a head-tail structure, with the galaxy found at one end. So the galaxy supplying the gas to the space rhino ought to be found somewhere near either end of the stream and roughly aligned with its longest axis. Kent's suggested parents don't do that, they're practically at right-angles to the long axis, and that's weird. It's hard to imagine the kind of orbit that would produce a feature like this. One of the proposed galaxies even has a tail pointing in exactly the opposite direction, which basically rules it out completely.

And there's so much gas in the stream that raises a further problem. Just as galaxies don't consist of just stars, neither do clusters consist only of galaxies. They're also filled with their own gas : much hotter and thinner than that in galaxies, but with an enormous total mass. Galaxies moving through this material at a high enough speed can have their own gas pushed out (ram pressure stripping), where it can slowly disperse and evaporate. So for most galaxies with streams, only a few percent of their gas mass is found in their streams : once their gas leaves the safety of their stellar disc, it's vulnerable to being torn apart and heated to the point of ionisation.

But there's so much mass in the Kent complex that this doesn't seem to have happened. Nearly all of the gas that the proposed donor galaxies would have had is still in the stream, and doesn't seem to have evaporated at all. Oddly, it's the lack of missing gas that makes the Kent complex even more unusual - it's not at all obvious how the gas has survived. Alternatively, some fraction of the gas could have evaporated, but in order to explain the enormous amount of gas that survives, that would have made the parent galaxies exceptionally gassy before stripping. And no-one, no-one at all, ever wants to be gassy before stripping.

Actually, there is one galaxy which almost fits the bill as a parent : NGC 4522. This is in the right place and even has a tail pointing towards the Kent complex. But if this is the parent, then it's an errant one that's trying to escape real fast : it's moving at 1,800 km/s relative to the complex. And why it's disconnected is unclear. Maybe it just wasn't ready to deal with parenthood, but 1,800 km/s is so fast it means they're probably not associated at all.  And there are plenty of other galaxies with tails in this region, so the alignment could just be a coincidence.


Conclusions

At this point we're fast running out of options. We tried to track down the parent, but none of them are claiming responsibility. Some are in the wrong place, some are too small, some too far away, some moving too quickly, some are moving in the wrong direction. None of them are at all satisfactory.

We do at least have a least-worst option though : NGC 4445, one of Kent's original suggestions. It's in the wrong place and would have had an awful lot of gas (or there would have to be unknown processes at work preventing the gas from evaporating), so it's not a great candidate. Not at all. In fact it's just plain lousy, but all the others are even worse.

A more radical suggestion is that the galaxy stripped in an unconventional way, losing its gas in a rapid event which then dispersed, so the long axis of the stream doesn't point to its parent. That's something we're still trying to model. It would open up more possibilities but it doesn't seem terribly likely either.

About the only good news is that the complicated velocity structure of the complex is something that we can explain. In simulations, those sorts of features are pretty common : once gas is removed, other galaxies can harass it into all kinds of complicated shapes. That doesn't help us with any of the other problems, but at least it's not a complete mystery.

So our best guess for this orphaned, harassed, gassy space rhino is that its errant parent was a stripper who refuses to claim responsibility for their offspring. Astronomy : it's every bit as entertaining as the Jerry Springer show.

Yeah, but are you a stripper who's failing to claim parental responsibility for their gassy space unicorn ? I think not.

No comments:

Post a Comment