Opening the lid on Pandora's Cluster

By Phil Plait | June 24, 2011 6:29 am

The largest structures in the Universe are superclusters: not just clusters of galaxies, but clusters of clusters. They can stretch for millions of light years and be composed of thousands of galaxies.

Abell 2744, at a distance from Earth of about 3.5 billion light years, is one such megastructure (if you want to sound fancy, astronomers call it "large-scale structure"). Astronomers have been studying Abell 2744 with an arsenal of telescopes, and have discovered that it’s actually the result of the ongoing collision of four galaxies clusters. If you’ve ever wondered what 400 trillion solar masses of material slamming into each other looks like, well, it’s more than a bit of a mess:

[Click to enclusternate.]

Yeah, like I said, it’s a mess.

First off, this picture is a combination of observations from Hubble (in visible light, colored blue, green, and red), the Very Large Telescope (also blue, green, and red), and the Chandra X-Ray Observatory (X-rays, colored pinkish). In visible light you can see literally hundreds of galaxies, probably more, dotting the supercluster. The pink glow is from very hot gas between galaxies; it started its life as gas inside of galaxies that got stripped off and heated to millions of degrees as the galaxies plow through the space around them (I like to think of it as opening a car window to let a noxious smell out — the wind from the car’s motion pushes the air inside the car out the windows).

The blue glow is perhaps the most interesting bit here: it’s a map of the location of dark matter. This type of exotic matter neither emits nor reflects light — hence the name — but it has mass, and that means it has gravity. As I described when this method was used to trace dark matter in the Bullet Cluster, gravity bends space, and light follows that curve. Galaxies farther away get their light distorted by the gravity from dark matter, and that distortion can be measured and used to trace the location of dark matter. The blue glow in the image above maps that.

The thing about dark matter is that it doesn’t interact with normal matter (electron, protons, you, me, lip balm, oranges, whatever). But all that gas between galaxies shown in pink is normal matter, so when one galaxy cluster slams into another at a few thousand kilometers per second that gas gets compressed, mixed-up, and heated. But dark matter just blows right on through. So by comparing the location of the galaxies, the dark matter, and the hot gas, a lot of the cluster’s history can be unraveled.

For example, one of the four clusters, located on the far right in the image, has very little hot gas associated with it. Apparently, it rammed through one of the clusters near the top of this image and had all its gas stripped away. For another, there’s a slight offset of the dark matter and hot gas distribution, indicating lots of secondary mergers that messed things up. Bizarrely, they found that some of the hot gas had moved ahead of the dark matter. That’s really weird, since the hot gas should’ve slowed down in the collision. It would be like seeing the tortoise way ahead of the hare in a footrace.

What the astronomers involved with this study think happened is this: about 150 million years ago, several clusters were headed for each other and began to merge, as shown in the diagram here. Most likely, there was one big one and several smaller ones. The smaller ones merged with the bigger one at pretty much the same time. This scenario apparently explains the structure seen, including that weird hot matter/dark matter swap: the gas underwent a gravitational slingshot, whipping around a local clump which flung it into a relatively empty region of space.

This is all very complicated, obviously. The scenario spun by the astronomers depends on a lot of complex and advanced physics and math, and painstaking analysis of a lot of data.

But if the details get difficult to understand, then let me at least leave you with this thought instead, something you can mull over and wonder about: by taking the data from a fleet of telescopes on and above the Earth, telescopes that see across the electromagnetic spectrum well beyond what the eye can perceive, we can piece together a history of an object with hundreds of trillions of stars spanning quintillions of kilometers of space and hundreds of millions of years in time.

And that, my friends, is what scientists do. And that’s pretty cool.

Image credit: NASA, ESA, ESO, CXC, and D. Coe (STScI)/J. Merten (Heidelberg/Bologna)

Related posts:

- Most massive cosmic pileup ever seen
- AAS report #2: Dark matter and large scale structure

Comments (28)

  1. Pete Jackson

    Presumably the gravitational bending of radiation by these large blobs of dark matter should cause detectable effects on the details of the cosmic microwave background (CMB) radiation in that direction. If the present data are not accurate enough to show it, perhaps future studies of the CMB will.


    Phil Plait:

    I like to think of it as opening a car window to let a noxious smell out…

    Is that because the mother-in-law, in the back seat, has farted again?

  3. gski

    What examples are there, of galaxies or clusters, that are flying apart due to having been separated from their dark matter?

  4. J. Earley

    #3, gski, asks some good questions. If a galaxy has been stripped of its dark matter, some long time ago, then its rotation curve should show a considerable difference from the Milky Way type galaxy rotation curve that is attributed to the presence of dark matter. For his second question, many (most?) galaxies in clusters show intra cluster velocities great enough to cause the cluster to fly apart. Have we found evidence of clusters that have lost their dark matter component and are now in the process of losing member galaxies?

  5. BJN

    At 3.5 billion light years away, I presume that the “150 million years” ago event was actually 3.65 billion years ago. That’s during the Paleo-archaean era of the Precambrian. And from what I understand, galaxies “slam” together a little like smoke rings mixing into each other.

  6. it’s a map of the location of dark matter If it were a map of “dark matter,” then there would be a gamma-ray emission map conformal to it from dark matter-dark matter annihalation. There isn’t, so it isn’t.
    Excess in Diffuse Gamma Rays Interpreted as a Dark Matter Annihilation Signal
    Cosmic gamma-ray background from dark matter annihilation
    Gamma rays from dark matter annihilation

    One problem with defective theory is that its contingent failures propagate. Adding parameters (curve fitting in physics; heteroskedasticity in economics) does not make it work for the next empirical falsification.

  7. david

    Not that I understand that paper in any detail of course, but high collision velocities and a complex merging scenario don’t exactly fill me with great confidence in the lambda-CDM view of this object. I’ll be interested to see what the modified gravity people make of it.

  8. As a lay-person, let me say this: It is simply awe-inspiring to look at images from Hubble, or, even better, my own telescope, and witness the grandeur of this vast Universe with my own eyes, realizing that we are all, in a way connected – we are all made from stardust.

  9. doctor_doctor_doctor

    What’s amazing is that each one of those lights is even bigger than the moon, in some cases.

  10. davidlpf

    I think the pandorica is loacted in the center and the doctor is trying to cry for help.

  11. Kim

    If we subtract the galaxies that are NOT in the clusters from the image, would it get easier to spot them? I assume we can do that based on their speeds.

    I see some stars photobombed the picture. Is the pink-haloed galaxy near the bottom left a photobomb too?

  12. CB

    @ BJN

    At 3.5 billion light years away, I presume that the “150 million years” ago event was actually 3.65 billion years ago.

    It means “150 million years ago from our vantage”, Which of course means that when you account for the distance and light travel time, it’s actually much longer ago (from the super-cluster’s perspective). But you can’t just add the two numbers, because that implies a false precision. There’s probably more than a 50 million light year error bar on the 3.5 billion light year distance.

    That’s why we usually just talk about how long ago things appear to have happened from our reference frame. it’s why when a super nova appeared in another galaxy in 2001, we say it happened 10 years ago, instead of something like 100,000,010 years ago.

    After all, it was impossible for those events to be visible to us, or for that matter to have had any impact on our portion of the universe at all, until the light from those events reached us. As far as our interaction with these distant galaxies is concerned, what we’re seeing now, is now.

  13. CB

    The thing about dark matter is that it doesn’t interact with normal matter (electron, protons, you, me, lip balm, oranges, whatever).

    That’s only in standard cosmology.

    In my Cosmetic Cosmology, dark matter does interact with lip balm, but only very weakly. Scarlet Passion colored lipstick, on the other hand, interacts powerfully with dark matter.

    But can I convince any scientists to put some lipstick on their detectors? No! I even stopped suggesting they put lips on their detectors for the lipstick, even though it helps. Still they scoff. They’re just afraid of the truth — that it’s no accident that the magazine “Cosmopolitan” is often abbreviated “Cosmo”, and it contains the secrets of the universe!

  14. Robin Byron

    What do the ‘clear’ lanes wandering through the image tell us, if anything? I’m wondering if they have anything to do with dark energy.

  15. David Taylor

    Does the description of the collisions imply that the relative space velocities of these clusters exceeds the cosmologic expansion associated with their mutual separations by “a few thousand kilometers per second”?

    Is that usual for the peculiar motion of galaxy clusters? Didn’t know they scooted around that much.

  16. DrFlimmer

    @ #7 Uncle Al

    1) The pictures above do not show a gamma-ray measurement, neither does the text say something about it. This means that either measurements in gamma-rays were not done, or the authors did not use any measurements, because they were irrelevant for their work.

    2) The gamma-ray signal of dark matter (if it really exists, see number 3) is very weak. I guess, with current observatories such a signal is invisible on such a long distance. Even the most powerful gamma-ray emitters (blazars or GRBs) have very low counts. Statistics are sometimes obtained from only a few dozens of photons. So, no detection would not necessarily imply that it doesn’t exist. Our detectors could just be too weak.

    3) These headlines “Dark matter signal bla bla” always give a feeling of uneasiness in these days. Many are so eager to find such a signal that they tend to interpret as much as they want in such data. For instance, a positron excess has already been interpreted as a dark matter signal, but it can also be explained by a much simpler and more “natural” approach (Klein-Nishina-scattering, just to give a name ;) ).

    So, all this implies that dark matter does not necessarily emit gamma-rays, those gamma-rays would be very weak, and almost undetectable. Therefore, your comment that dark matter does not exist due to your reasons, is not correct.
    Dark matter has only been detected indirectly, so far. That’s also, how it was done in this research. And it’s quite convincing: We see the effect of gravity (lenses, and only gravity can do it), but no “ordinary” matter to curve space – therefore, that gravity must come from something invisible. Science has given it the name/placeholder “dark matter”.

  17. CB

    @ DrFlimmer:

    I do believe that #1 was the “tell” of a kindly old-fashioned troll. A mechanism by which the troll lets the savvy reader know what they’re up to with an over-the-top logical fallacy: There was no gamma ray observation, therefore no gamma rays, therefore no gamma ray source. You know, like how I disproved Black Body theory by looking around without any infrared googles, saw no infrared radiation, and concluded there wasn’t any.

    Usually the flamebait comes first, and the tell last, but still, I think this qualifies. It’s almost quaint.

  18. DrFlimmer

    @ #18 CB

    Sure. This is possible, and I even considered it. Nonetheless I wrote my answer, maybe even just for the heck of it.
    I also gave some points of view, and maybe some points for others to think about. Or maybe not. What the hell…. ;)

  19. Gary Ansorge

    19. DrFlimmer

    Al is hardly a kindly, old fashioned troll. This from HIS site,,,

    Uncle Al says, “If god loves the poor, crippled, and stupid, then god can (fraking) well pay for them or take them back. My wallet is broken.”

    He has an agenda, often shows up on Physics web sites to spout his non sense and appears to be a far right winger.

    Other than those few personalities quirks, he’s at least a classic example of the human species,ie, xenophobic and about as compassionate as a brick.

    Gary 7

  20. Keith Bowden


    I remember Jor-el’s reference to “the 28 known galaxies” in Superman (1978) and somehow that colored my perception of the number of galaxies for a long time. I realized there were more, I just thought that’s all they (and we) could see. Of course, they’re just the galaxies that Kryptonians had explored/observed in detail. But compared to how many there actually are… it’s stunning.

    While I’m talking movies, I always took Han Solo’s Kessel Run bragadoccio in Star Wars as a reference to finding a sort of short cut through a hazardous area of space or something. I suppose that still doesn’t really make any sense, but at least it reflects the definition of “parsec”. (And either way, it doesn’t seem to be a blooper – just look at Alec Guinness’ brilliant expression in response!)

  21. Jason Hosler

    What is the likelyhood that blackholes could just be conglomerations of dark matter. Since, to my understanding, dark matter is an attractive force (as opposed to dark energy), could some blackholes be causer simply due to an accumulation of dark matter in a given region that then collapses upon itself?

  22. Doug

    @Pete Jackson (#1): Gravitational lensing of the CMB is quite hard to detect, as gravitational lensing preserves surface brightness. If you had a uniform intensity background behind the cluster, you would have a uniform intensity background after the lensing as well. The bumps in the CMB in the maps are on angular scales larger than the clusters, so the CMB background is close to uniform over the region you would normally hope to detect such an effect. You also have the problem that the Sunyaev-Zeldovich effect from the X-ray emitting plasma in the cluster alters the CMB much more than the lensing would.

    One thing which is expected to be seen is that the polarization signal from the CMB should be changed by lenses like this cluster. In a few years the Planck satellite should have enough data to detect this, although it will be a statistical detection rather than being able to look at a given area and see how any given cluster is altering the signal.

    @gski (#3): Dark matter and stars/galaxies should react similarly in these collisions, as they’re both effectively collisionless particles (due to the large separation between stars in a galaxy compared to their size). Removing the dark matter from a galaxy would require a much larger self-interaction cross-section amongst dark matter particles than most models predict (or a 5th, dark matter only force). There have been some observations of galaxies in clusters using a galaxy-galaxy lensing technique which hint at being able to see the outskirts of the galaxies’ dark matter halos having been tidally stripped, similar to how the Milky Way tidally strips the outskirts of the dwarf galaxies surrounding us.

  23. gski

    @Doug #3
    From the OP “…normal matter, so when one galaxy cluster slams into another at a few thousand kilometers per second that gas gets compressed, mixed-up, and heated. But dark matter just blows right on through.”

    I’ve also read on another blog that the dark matter passes through unimpeded unlike the normal matter which is slowed by interactions and so the two are separated from each other.

    I would think that at a minimum some distortion of the galaxy must occur. I’ve not read of this being observed.

  24. Dan Zee

    I’m not a scientist, but this doesn’t exclude dark matter being the gravitational effects that we’re seeing from a nearby dimension. We know that gravity is suppose to be much stronger than it is. Perhaps in the prime dimension, gravity is 10x and in our dimension only 1x. In a massive collision like this, the matter in the prime dimension is torn from it’s connection from the matter in our dimension, and because the matter is “heavier” it tears through “stuff” that is stopped (because it’s “lighter”) in our dimension. I know a few astronomers have offered this explanation. It seems more realistic than saying that dark matter is composed of particles we’ve never seen.


Discover's Newsletter

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


See More

Collapse bottom bar

Login to your Account

E-mail address:
Remember me
Forgot your password?
No problem. Click here to have it e-mailed to you.

Not Registered Yet?

Register now for FREE. Registration only takes a few minutes to complete. Register now »