Andromeda: born out of a massive collision?

By Phil Plait | November 23, 2010 12:27 pm

Was the Andromeda Galaxy, the largest and most massive galaxy in our local neighborhood, shaped into its current structure due to a monstrous collision over 6 billion years ago? According to a new study by some French astronomers, the answer is oui.

They created a lovely animation based on the model. It shows the collision of the two galaxies and how they interact:

Wow! You can see how the galaxies get disrupted, and perhaps get something of a feel for just how violent and incredible an event on this scale can be.

Using a sophisticated computer code that models the gravitational and fluid (pedantic: hydrodynamical) interaction between stars, gas, dust, and dark matter, they found that an ancient and massive collision between a galaxy a bit bigger than our Milky Way, and a smaller one about 1/3 the mass, reproduces a large amount of the structure we see in Andromeda today. That includes "…the large thin disk including its giant ring of gas and dust, the massive central bulge, the gigantic thick disk, the giant stream of old stars, as well as many other stellar streams discovered in the galaxy halo" according to the press release (the paper itself is in French).

m31_collision_modelNot only that, but it may solve another big mystery: what the heck are the Magellanic Clouds? These are two small, irregular galaxies that apparently orbit the Milky Way; they are the closest galaxies to us at a distance of about 180,000 light years each. Recent studies have hinted that these galaxies may not actually be orbiting us as always thought, but might be passing us by at nearly a million kilometers per hour. This new model of Andromeda shows that during the collision, a long streamer of gas would’ve been drawn out and ejected. This is called a tidal tail, and is common in large collisions. In this case, the model indicates the tail may have formed the two Magellanic galaxies and sent them heading our way!

The most basic aspect of a good scientific model is that it should solve some issue you see. For example, why does the Andromeda galaxy have a ring of gas circling its center? If your model explains that, then great! But if it also explains lots of other structures, then that means the model has a much better chance of being right, or at least being mostly right. This new model seems to do just that.

WISE_andromedaSo this new study is pretty cool. We know that there have been lots of little collisions over the eons; we think all big galaxies like the Milky Way and Andromeda built up to their current size by eating and absorbing smaller galaxies. But those are usually dinky dwarf galaxies; this new study may show that Andromeda also had at least one really, really big collision long ago.

And regular readers know as well that there’s another one coming: in a billion years or two, the Milky Way and Andromeda may themselves collide. Both galaxies are among the largest in the near Universe, so when this happens it’ll be a spectacular and amazing event. I wonder if alien astronomers hundreds of million of light years away, and billions of years hence, will get images of the ensuing merger and gasp (assuming they breathe) in awe as I do when I see such a magnificent tableau?

Image and video credit: GEPI, Observatoire de Paris / NAOC

Related posts:

Spiraling tentacles of galactic doom
A Swift view of Andromeda
The first spectacular views of the sky from WISE!
Awesome Antennae!

CATEGORIZED UNDER: Astronomy, Pretty pictures

Comments (37)

  1. Oli

    One-third of the Milky Way… Is that about the size of the Triangulum? How come the MW or dwarf galaxies in our group (which, obviously, is the best group) haven’t already collided with either of those galaxies?

    I suggest we fly 6 billion lightyears away and then look back: we’d see the collision happening. Should be easy, right?

    That’s quite a number of question marks there.

  2. Chris

    If I remember right, Andromeda has two big black holes in its core.

  3. Joseph G

    Awesome stuff, Phil! Both in the classical and colloquial definitions of the word :)

  4. Joseph G

    I always wonder what it’s like for those few stars that get ejected into intergalactic space, alone. What a bummer it must be if any intelligent life were to evolve on any of those systems. They’d have spectacular views of the nearest galaxy, sure, but imagine the nearest star being a couple hundred thousand light-years distant… They’d probably wonder, upon first developing astronomy, why they were quarantined from all the other distant stars, out there cozily packed together – either that, or they might assume that a galaxy is simply far too dangerous a place for life to evolve.

    Probably much more common, it looks like, are clusters of stars. It’s amazing to think of galaxies swapping star clusters as they collide and groups of stars are flung about. Between the post a few days ago about stars from another galaxy in our own milky way, and this one, I wonder how many galaxies certain long-lived red/orange stars may actually see in their lifetimes?

  5. Kullat Nunu

    Not only that, but it may solve another big mystery: what the heck are the Magellanic Clouds? These are two small, irregular galaxies that apparently orbit the Milky Way; they are the closest galaxies to us at a distance of about 180,000 light years each.

    Actually, both Canis Major Dwarf and Sagittarius Dwarf Elliptical galaxies are closer, both are in the process of being absorbed by our galaxy. There are probably more of such galaxies to be found.

  6. Kullat Nunu

    This is not the first time that Andromeda has been suspected of being a result of a galactic merger. According to an old study, the birth of Andromeda could have involved galaxies in the nearby Maffei galaxy group in an event they call “Little Bang”. AFAIK later, more accurate distance measurements have shown that the group is too far away to have participated in the event. So sad we know so little about the group, as it includes some of the closest large galaxies, including the closest large elliptical galaxy.

  7. Michael Swanson

    Spectacular, but I’d love to see the same model slowed down enough that one can appreciate what’s happening.

  8. Leon
  9. Morgan

    I always wonder what it’s like for those few stars that get ejected into intergalactic space, alone. What a bummer it must be if any intelligent life were to evolve on any of those systems.

    One of Iain M Banks’ books, Against A Dark Background, is set in such a system. The fact that the species there has nowhere else to go has a big influence on their history, and sets it apart from his other SF which is generally set in a populous Milky Way. (As you might guess from the title, that history isn’t a happy one.)

  10. Joseph G

    @#9 Morgan: Thanks, sounds like one to add to my reading list :)

  11. Gary Ansorge

    9. Morgan

    Haven’t read the book but my contention is that it would take several hundred thousand years to really develop all the resources of a single solar system, so not being able to get to a nearby star would be no big deal,,,for a long while. IF it turns out to be possible to travel faster than light, a few hundred kiloYears should be more than enough time to develop such tech. If it’s NOT possible, it wouldn’t have particularly devastating effects. I doubt our descendants 100 KiloYears from now will have anything like our sensibilities but they may still survive quite well, despite being confined to a single solar system.

    Gary 7

  12. Levi in NY

    To give credit where credit is due, the press release you linked to says researchers at the Chinese Academy of Sciences contributed to the study as well. The answer is not just oui, but 对 (duì) as well. :)

  13. Cusp

    This model doesn’t really reproduce the observed substructure – especially with regards to detailed streams, and metallicity.

  14. Nick

    I wish I had a massive central bulge…

  15. johno

    xscreensaver has had a galaxy collision simulator since the early 90’s. You need a linux or unix system to use it but it’s very pretty, and a huge time sink.


  16. So my question is, how much disruption does this actually do to planetary systems within both galaxies? I know, it’s a slow process and the stars simply pass by one another light years apart, but we’re also talking about stars adapting to a new gravitational source to change their orbits around a new galactic center, too. Does the change in orbit affect everything equally, so the systems simply realign to a different galactic orbit gently and without any “internal” changes, or does the star get tugged and the planets lag behind a bit? In other words, how often does a planetary system get radically changed by the merging process? I’m fuzzy on the whole relative gravitational pull thing.

  17. Daniel J. Andrews

    My favourite merging spiral galaxy animation is this one:

    There’s a high definition video on the original site which you can download (or could when it first came out). It shows two merging spiral galaxies using individual stars as well as gas temperatures. Educational video and I haven’t tired of watching the full-screen HD version. Music is nice too.

    @Michael–this one is slowed down so you can more easily appreciate what is happening.

  18. John Paradox

    15. Nick Says:

    I wish I had a massive central bulge…

    I wish I didn’t.


  19. Dragonchild

    Only way to peer review this simulation is to pore over the source code. Is that usually how it works?

  20. Gary Ansorge

    I like re-writing other writers work, just to see if there’s any poetry implied.

    Phil? How’s this for a title?


    Of massive collision


    (That was fun)

    Gary 7

  21. 15. Nick Says: “I wish I had a massive central bulge…”

    Funny, most guys are trying to get rid of theirs!

    – Jack

    Post edit: Aaack! I see J/P=? beat me to it!

  22. The BA says: “Using a sophisticated computer code that models the gravitational and fluid (pedantic: hydrodynamical) interaction…”

    How is this fluid/hydrodynamic interaction? The stars generally stay light years apart. The only direct collisions are between dust particles, for the most part.

    At these scales of distance and object size, does gravity take the place of Vanderwaals forces so that whole stars can be assumed to act like molecules in a Newtonian fluid?

    – Jack

  23. Messier Tidy Upper

    Fascinating news even though it seems like we’ve missed the best bits by a few (hundred?) million years. 😉

    Excellent animation there – but I wish it was a bit longer and perhaps a bit, just a fraction, slower.

  24. Messier Tidy Upper

    21. Gary Ansorge Says:

    I like re-writing other writers work, just to see if there’s any poetry implied. Phil? How’s this for a title?
    Of massive collision
    (That was fun)Gary 7

    That’s good, fun indeed, & makes a nice set of opening lines but needs more I reckon.

    There’s usually poetry even in the strangest places.

    Hope you don’t mind if I continue the poem from there with my own bit of doggrel? :

    Messier Thirty One

    Of massive collision born
    Spiral shattered
    Then reformed

    Nebula we called you first
    Till 1885
    A super supernovae seen
    The first we recognised
    Our island universal clue
    On cosmic size and what you mean

    So distant we did find you then
    Yet close in cosmic terms
    Our partner in the Local Group
    Our killer in long term

    We’ll fuse one day in billion years*
    We’ll intermix and merge
    Another collision yet unseen
    Except in mental terms.

    For now you glow beneath** the Square
    Like dust by Pegasus kicked
    We learn of you more all the time
    But haven’t got you licked!

    – Stevo Raine, first draft, just now.

    Everyone please feel free to quote with attribution and also to come up with better versions too! :-)


    * Actually, about 4.5 billion years so I’m using ‘billion’ in the plural sense despite the singular form! Poetic license okay? 😉

    ** That’s “beneath” as seen from the southern hemisphere. Northern hemispheres are welcome to substitite the word ‘above’ there! 😉

  25. Messier Tidy Upper

    PS. That 1885 Andromedan Supernova I mentioned was S Andromedae (SN 1885A) and I’m pretty sure it was one of the one’s that helped prove the existence of other galaxies resolving the dispute over whether the “Great Andromeda Nebula” (M31) was a nearby or distant object; a still forming stellar system or a distant island universe, thus revealing how vast our cosmos really is hence that reference. For more see :

    @ 2. Chris Says:

    If I remember right, Andromeda has two big black holes in its core.

    That was what I once thought too, but it turns out that’s not quite the case although the structure of M31’s nucleus is certainly interesting. See via wikipedia :

    One last link for y’all here :

    gives a photographic finder chart via James Kaler’s Stars website.

    Note that Alpheratz or Sirrah as Alpha Andromedae is also known is also one of the 4 main stars making up the “Great Square of Pegasus” asterism formerly being known as Delta Pegasi as well – being one of a couple of “shared stars” along with El Nath which was called Gamma Aurigae as well as Beta Tauri. Sirrah / Alpheratz marks the start of the flying horse’s hind leg(s) which is how I tend to see the whole constellation of Andromedae! :-)

  26. Cusp

    > Only way to peer review this simulation is to pore over the source code. Is that usually how it works?

    The source code is public and can be downloaded and used by anyone.


    It’s a standard astrophysics resource used my many astrophysicists.

  27. TRL

    @Messier… S And was used to prove that M31 was NOT a large external galaxy. Those who favored the hypothesis that “spiral nebula” were not similar systems to the Milky Way cited the conclusion that S And would have to be vastly more luminous than any other stellar variable known at the time if M31 were at a large distance. What was not known to them was that S And was a Type Ia Supernova, thus it WAS much more intrinsically luminous than anything that had been seen before. There was no context in the 1920s that would have allowed SN to be understood. M31 was shown to be external when Edwin Hubble was able to identify Cepheid variables in its disk.

    I’m afraid your poem is in need of revision…

  28. Yeebok Shu'in

    Gah, been looking for 5 minutes and I can’t find it. To those mentioning a galaxy merger simulator, I recall being pointed to one from here when Phil had mention of a solar system simulator. This might help you recall (twitpic :

    Here we go ..

    Somewhere from there I wound up looking at Universe Sandbox (or the other way around) which was way cool too.

    The whole reason for me posting though was that another item I somehow found during that night was a galactic collision simulator named Galaxy 3D. Anyway as KurtMac (comment 54) said, Galaxy 3d is a tool that does mergers.

    When I initially got it, that version only used one CPU core. Seeing my PC sit on 25% usage (it’s a quad) and the app lagging I wrote to the author. Long story short they were working on a CUDA version which used the graphics processor rather than the CPU. The difference was astounding. Sadly I cannot find it anywhere and I am loathe to share the beta version I have/had (I don’t delete stuff but ..) without explicit permission, and it appears the new file is not linked to on the site but is at the same location the author gave me to get the test copy. That said the version on the site, is still good.

    So excuse the ramble about old unrelated posts and dead links. I remember the night Phil posted the original story I linked to above, leading me on quite the journey of discovery. I’m reposting the link in the hope of someone else having a cool night like I did.

    However when you see stuff like this : (twitpic again) whizzing around your screen at 50FPS, it can make the little nerd in you all gleeful. I know it did that to my inner nerd.

  29. Messier Tidy Upper

    @ ^ TRL : Oh. D’oh! :-(

    Thanks I guess.

    I didn’t think that was the case. I’ll have to research that further.

  30. Jon Hanford

    Concerning a possible interaction between M 31 and M 33, this 2009 paper makes the case for just such a interaction using new, deep panoramic imaging with the CFHT and the INT as part of the PAndAS survey:

    Besides offering some fantastic images of the pair, extensive computer modeling of the interaction was performed and can be seen here:

    While no unique solutions for the proposed orbital geometry were found, the idea is nonetheless backed up to some extent by observations of substructures in the M 31 extended disk , and are compared in the video and accompanying paper. A couple of other papers came out recently with support for this idea, one based on neutral hydrogen observations of the two galaxies. Was any mention made of the possibility of anM31-M33 interaction in the work presented here by Phil? BTW, the PAndAS survey is scheduled to be completed in 2011, hopefully with more details on this possible interaction. Some amazing video and imagery here though.

  31. Messier Tidy Upper

    @ ^ 32. Jon Hanford : Thanks for that info and those links. :-)

    The M31 – M33 Andromeda – Triangulum galaxies one was particularly good. I wonder if someone has done something similar with the interactions between our Milky Way Galaxy and the two Magellanic Clouds?

    (PS. Did a quick Youtube search but, alas, can’t find anything on that.)

  32. swok

    just too cool…

    These galactic collisions are obviously important drivers for star formation. When this happens the galaxies are much more efficient at sqeezing stars out of their gas clouds.

    Since star formation is one of the terms of the Drake equation, this may be as important to our existence as the slime we evolved from.

  33. Nick

    I actually have a legit question this time. What would two super massive black holes colliding look like? I imagine that it would be more energetic than a GRB. Towards the very center of these two galaxies would you actually visibly see stars moving as they’re being flung around the black holes? In that simulation, given that one second is probably 20 million years, those two black holes snap together very quickly. When they are within a few thousand miles of each other, they might be accelerating to enormous speeds. Would the singularities break apart in any way and form normal matter again, and then re-accrete? Is there any evidence of exotic particles or nebulae near the center of andromeda to hint at this type of collision?


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!


See More

Collapse bottom bar