Infographic : Galaxy Size Comparison Chart

EDIT : Several typos were found my myself and others over the last few months. To prevent links from breaking, the old charts are still online. For the latest, correct versions, see my website.

Type in "asteroid sizes" into Google and you'll quickly find a bunch of  images comparing various asteroids, putting them all next to each at the same scale. The same goes for planets and stars. Yet the results for galaxies are useless. Not only do you not get any size comparisons, but scroll down even just a page and you get images of smartphones, for crying out loud.

Well, it's time to correct this disgusting  nay, bigoted oversight with the following infographics (which is what I gather is now the cool term for "posters").  These were prompted partly because no similar images exist that I can find, and also by recent claims that the largest spiral galaxy has been discovered.

The images I used were selected purely on an ad-hoc basis. Obviously, the Milky Way had to be there. Since really giant galaxies are many times larger than the Milky Way, and those were the ones I particularly wanted to show, that basically ruled out showing any dwarf galaxies (like the LMC and SMC for example). I tried to get a nice selection of well-known, interesting objects. I was also a little limited in that I needed high-resolution images which completely mapped the full extent of each object (often, because of a small field of view, only the central regions are mapped).

Still, I think the final selection has a decent mix, and I reckon it was a productive use of a Saturday.

Zoomable version here.

As will be evident from the poster, "my galaxy is bigger than your galaxy" claims should be treated with caution. The latest hoo-ha is about NGC 6872 (very bottom of the poster), which, though indeed enormous, has been stretched by an interaction with another object. Is it really fair to claim it's the largest if it's been stretched ?

Even defining the edge of ordinary galaxies can be tricky. Especially since their various components (gas, stars, dust, dark matter) extend to different distances from their centers. In some cases, truly enormous radio jets extend many times further out than the stars. Should they be included as part of the galaxy ?

To my mind the gigantic (but very faint) Malin 1 has a better claim to the throne than NGC 6872, as its disc hasn't been temporarily stretched by some interaction. How such a large disc formed is a bit of a mystery, but it is at least a true spiral disc, even if it's very faint and not remotely photogenic (it's barely visible with Hubble, for heaven's sake). 

But to some extent, all this is just semantics, and it really doesn't matter which galaxy is the largest, any more than it matters who landed on the Moon second or if Pluto is a planet or not. In any case, spirals are puny. The hands-down largest galaxy of any kind is IC 1101. And it is truly, in every way, monstrous.

Zoomable version here 

I don't just mean it's monstrous in that it's staggeringly vast, although that's part of it. I mean monstrous because it probably got so large by eating its neighbours. cD galaxies like this are found at the center of rich galaxy clusters, where there are plenty of smaller galaxies falling in that can be absorbed. That makes the galaxy heavier, which means their gravity can pull in more and more galaxies. It's like a cannibalistic orgy on steroids*. 

* So just like every episode of True Blood then.

Perhaps surprisingly, there aren't many pretty pictures of IC 1101. The galaxy's morbid obesity is offset by its great distance from us, although there are a few nice ones with Hubble (see below).  So for the chart, I took the image of M87 (another giant elliptical) and scaled it up. This isn't such a cheap way to do it, because both galaxies are pretty smooth structures, and in any case no image exists that's large enough to display at the massive resolution required.

IC 1101, as you will have guessed, is the big bright one. More images can be found through the Hubble archive and this website.

Another point is that while colliding galaxies might be initially spectacular, eventually they run out of gas and stop forming new stars (during the collision, the gas gets compressed, triggering star formation). Eventually, there's nothing left but a huge ball of old, red stars, the blue (short-lived) ones having died off aeons ago. With no gas and no new stars being formed, over time the random motions of the long-lived red stars make the galaxy nothing more than a titanic stellar swarm. And that's why our cannibalistic juggernaut isn't going to win any beauty contests.

ACHIEVEMENT UNLOCKED

GOOGLEWHACK !

Binge Drinking vs Binge Coding

Binge drinking is a bit like wrestling a shark, it might be fun, but it's not necessarily a good idea to video your activities and post them on YouTube. Seriously, read the links.

Binge coding - that is, doing nothing but writing computer code day in, day out - can also have deleterious effects. Currently, moments after waking (usually around 6:30, which doesn't help) and overcoming the customary momentary existential crisis - a sort of mental flailing of the arms - the first thing I find myself thinking is something like, "Hmm, I wonder how I can translate the bounding box values in a postscript file into the world coordinate system of an HI data cube ?"

And yes, you're right, that is a little worrying.

I continue to be unwaveringly preoccupied with such esotericisms (no, spellchecker, I did not mean eroticisms) until very late at night. At which point, while shaking slightly and murmuring things like, "renzograms...  matplotlib...  mesh geometry all wrong..." I reluctantly surrender consciousness. And that's on a Saturday.

Alright, I'll confess. My ulterior motive for this post is that I'm really hoping to generate a Googlewhack*. It turns out that there are currently zero results returned in Google for "renzogram matplotlib." Matplotlib is a popular graph-plotting package for Python. Renzograms are a curiously unpopular (though hardly unknown) type of plot, of which more in a subsequent post, perhaps (they can be really quite pretty, especially in 3D).

* With luck, this will be the least viewed post ever !

They also look quite a lot like the time vortex in Dr Who. This may well be the first ever renzogram generated by matplotlib posted on the internet. GO ME !

Anyway, the effects of binge drinking and binge coding can be very similar. For one thing, both cause anti-social behaviour, though of radically different natures. An angry drunk may tend to hit people. A happy drunk may tend to hit on people, or possibly just sing loud obnoxious songs. Whereas the binge coder will simply refuse natural light, shun all human contact and occasionally wonder why no-one's calling any more.

The effects are also similarly dissimilar on health. In the short term, neither does very much harm except to cause an unpleasant sense of confusion and a headache. In the longer term, full-blown alcoholism leads to all kinds of problems, like death. I can't imagine that it's physically possible to code oneself to death -

if self.alive == true:
   reality.unlink(self)

- probably doesn't work, for instance, but it's certainly not going to convey much in the way of health benefits to be shut away in a darkened room gorging on all manner of snack food with nothing to drink except tea and root beer. Well.... not if this continues for a whole year, at any rate.

However, binge drinking can be (except in extreme cases) self-limiting. Eventually, you run out of money, or throw up, or pass out, or possibly all three (but not necessarily in that order). Unfortunately, there's no such in-built safety net with binge coding, In fact, binge coding will more often than not leave you with more money that when you started.

Case in point.

Granted (to pick an example ENTIRELY AT RANDOM) generating, oh, I don't know, renzograms with matplotlib is unlikely to lead to the next Angry Birds, but it's still earning a salary rather than burning one. While I suppose sheer exhaustion might cause you to pass out eventually, it's probably not possible to code oneself into puking :

if self.coding >= excessive:
   for x in self.stomach:
      remove.x

No, the only way to beat binge coding is to see it through to the bitter end. To fix ALL of the bugs so that generating renzograms with matplotlib becomes foolproof. And then eventually you find that all is done and you can resume science and write more coherent blog posts.

SPIKE

Thanks, reddit ! You rock !

Welcome, new followers ! May your visit be educational and your hovercraft full of eels !

Wait, what ? NEW followers ? NEW followers ??? Oh crap oh crap oh crap oh crap oh crap.... this means I have to make more content.... umm.... wasn't really prepared for this... bugger. Err, in the meantime, here are some popular articles you might enjoy, or might enjoy posting on reddit, or might not.

Project Orion : How to Nuke a Spaceship Without Killing Anyone

Deep Space Force

Galaxies are Pretty

Beginner's Guide to the Universe

Excuse me while I scramble to find something else to write about.... err.... back in a bit !

VY Canis Majoris - Or, Be Afraid. Be Very Afraid.

Recently I came across the following image :

Supposedly it depicts VY Canis Majoris, the largest known star. Not the most massive or the brightest, but the one with the largest diameter - a truly terrifying 1420 times bigger than our own Sun. It's a beautiful and inspirational image by the talented artist Facundo Diaz. But is it accurate ?

The image grabbed my attention, but the properly-cited physical characteristics on Wikipedia are what's really stunning. Other images have tried to put into perspective just how frickin' gargantuan this behemoth is, but I want to tackle this in a slightly different way. The most natural approach to me is to consider what would happen if, somehow, our Sun was magically and instantaneous swapped with VY Canis Majoris.

The result, in a word, is death. Complete and total death. Mercury, Venus, Earth, Mars, the million tiny worlds of the main asteroid belt, and even Jupiter - all gone. Jupiter, in fact, would find itself about 70 million miles below the surface of the star. It's a thought experiment to make Schrodinger's Cat look like the kindest act of animal welfare.

"Surface" in this case is a tricky concept. Although the star is staggeringly vast, it doesn't have all that much mass, considering. It may have a volume 10 billion times greater than the Sun, but its mass is only 17 times greater. That gives it an average density of next to nothing - a thousand times thinner than ordinary air (of course, the interior regions where nuclear fusion is occurring will be very much denser).

Not only is the outer part of the star a pretty good approximation to a vacuum, but the gravity would also be only about a quarter as much as on the surface of the Earth (on the surface of our own Sun, gravity is about 28 times higher than on Earth). And that means that the nice, well-defined spheres in the above images are just wrong. It might be better to think of it as a searingly hot, blinding nebula than a star. It might not even be spherical.

What it definitely would be is utterly spectacular. From the orbit of Saturn, the star would fill half the sky. It would appear less like a star and more like a stupendous flaming wall of fire.

I cannot hope to match the beauty of Diaz's render, but I shall give it a damn good try.

Actually, what it would look like to the naked eye is not much of anything, because it would cause instant blindness. The surface brightness of the star (i.e. how much energy it's putting out per square meter) is about one-sixth that of the Sun, which for all intents and purposes makes no difference. Imagine the Sun filling half the sky, and you get the idea.

In fact it's somewhat worse that that, because the amount of energy received by a luckless "observer" would be about five million Watts per square meter at Saturn - roughly 5,000 times what you get from the Sun on Earth. So there'd be no need to worry about the blindness because there wouldn't be time. The rings of Saturn - if not actually the planet itself - wouldn't last long either.

This means that whether Diaz's wonderful image is accurate or not is a moot point. It would be a sight, quite literally, of such awesome power that no human eye could ever see it.

Moving further out, things get better - slowly - but only because they can't very well get any worse. At the orbit of Uranus, the Star (it deserves a capital letter, and a better name) still absolutely dominates the view.

If we move along a planet, we come to Neptune. Now we can start to see that the Star is embedded in a nebula. This isn't surprising, given the insane amount of energy output, the low surface gravity and density - effectively, it's evaporating, wrapping itself in its own death shroud, if you will.

I cheated a little bit here. The "nebula" is actually an image of the Large Magellanic Cloud - a nearby galaxy - as seen in the infra-red.

At Pluto, the Star still appears much larger in view than the Sun does from Earth. Just in case I haven't hammered the point home yet, it's absolutely ginagorous.

The most distant man-made object is currently the Voyager 1 spacecraft. One hundred and twenty times further away from the Sun than Earth, it's somewhere that can only be described as cold, dark and incredibly lonely. Not so if our puny Sun makes way for YV Canis Majoris. Astonishingly, at this huge distance it would still span 6 degrees. The Sun as seen from Earth spans only 0.5 degrees.

Voyager model from NASA.

Oh, and I almost forgot. One day, this star is going to explode. Isn't science fun ?