Funhouse galaxy

By Phil Plait | February 8, 2012 7:00 am

Sometimes, I like to think of a photon of light as a car on a road. As the road dips and curves, a car has to follow that path, dipping and curving as well. It might be weird to think of space as curving, but it does. Gravity from massive objects warps space, and a beam of light moving through that curved space curves along with it.

This is the principle behind what’s called gravitational lensing. A beam of light passing by an object — a big galaxy, say, or a cluster of galaxies — bends one way. A beam headed in a slightly different direction bends a slightly different way. This can really mess with what we see… which I can prove! Check this out: a Hubble image of the galaxy RCSGA 032727-13260.

What a mess! All those arcs and blue smudges are images of that one galaxy. The light from that galaxy traveled nearly 10 billion light years to get here! But when it was halfway here, that light passed by the big cluster of galaxies — the red fuzzballs — in the middle of the image. As it did, the curvature of space distorted and warped the light from the galaxy, and by the time it reached us here at Earth the image looks like this. The outstretched, smeared-out arc is amazing; I’ve never seen one that long and well-defined before.

Not only that, but the image gets broken up into several separate images. There are no fewer than four different repetitions of the background galaxy in the big image. To show that, I put three of them together here. It’s goofed up, to be sure, but you can kinda sorta see they are the same galaxy, flipped over and/or smudged out.

The cool thing about this is we can learn about the more distant galaxy by examining these images. While you might think this would be a pain for astronomers to sort out what’s going on — and it is — nature provides a little help: lenses like this can magnify and make the galaxy look a lot brighter than it normally would, so we actually can see galaxies farther away than we normally could. And c’mon, ten billion light years is a long, long way off!

So what happens when astronomers disentangle all that weird smeary stuff and reconstruct what the galaxy looks like? Well, here it is. It’s still a mess, but you can see it’s a wide-armed spiral, and it’s very blue. That means there’s lots of star formation going on — young massive stars are blue, and very bright — which isn’t too surprising; young galaxies are known to be stellar nurseries. But this one has very bright knots of star birth, even brighter than our own Milky Way. That gives astronomers precious insight into conditions in the distant Universe. And since it took 10 billion years for that light to get here, that means we’re seeing the Universe when it was young! So we’re seeing this galaxy as it appeared just about 3 or so billion years after the Big Bang.

All that, because space is warped. Think about that next time you see yourself in a funhouse mirror.

Image credit: NASA, ESA, J. Rigby (NASA Goddard Space Flight Center), K. Sharon (Kavli Institute for Cosmological Physics, University of Chicago), and M. Gladders and E. Wuyts (University of Chicago)

Related posts:

Galaxies swarm and light bends under dark matter’s sway
The galaxy may swarm with billions of wandering planets
Obese, gluttonous, and cannibalistic is no way to go through life, son
Hubble spots 67 gravitational lenses

CATEGORIZED UNDER: Astronomy, Cool stuff, Pretty pictures

Comments (25)

  1. For some reason, your opening line reminded me of Sam the Photon. :)

    Although, I don’t think Sam’s journey was nearly as exciting as this one.

  2. Mike G.

    This might all be very silly – I’m a programmer, not an astronomer :). But wouldn’t that lensed galaxy be 3x better as an observational target for things like supernovae than any other galaxy?

    I’m guessing that the different lensing paths are different lengths, so we can watch that galaxy over 3 or more different time windows for the price of one observation.

    Of course, the downside is the distortion.

    Actually, couldn’t radio telescopes time some pulsars in the different reflections and figure out the length differences of the lensing paths from the pulsar timing differences?

  3. Chris A.

    Seeing this image makes me chuckle a bit, remembering my days at Hubble (late 80s-early 90s) when there was still a pretty spirited debate going on about whether these “galaxy arcs” were due to gravitational lensing, or whether they were tidal tails from galaxy collisions. Not much doubt any more, thanks to images like this one.

  4. Scott P.

    I seem to see a second distorted galaxy — just left of center, about a third of the way up. Anyone else see it?

  5. Nigel Depledge

    I’m not sure your analogy is that useful. Photons don’t have steering wheels.

    As a road dips and turns, you have to make the car turn to follow it.

    Perhaps a bicycle in a velodrome is a better analogy. Or a motorbike on a Wall of Death.

  6. Pete Jackson

    The separate images will all have slightly different time delays, so that if you see a supernova go off in one image, then, after months or years, you’ll get a replay of the supernova in the other images!

    Fantastic picture, BA!

  7. Arthur Maruyama

    @ Pete Jackson

    True (and thanks–your idea didn’t occur to me), but considering the space-time scale involved the delay between some of the gravity-lensed images is probably more in terms of thousands or even (tens of?) millions of years.

  8. Acronym Jim

    Ah, so maybe the main “easter egg” in Holbein’s painting, “The Ambassadors” wasn’t an anamorph. It was just warped by the gravity produced by the apparently massive egos of the titular subjects.

  9. Chip

    Of course if one wanted a nice speculative image of what this distant galaxy looked like, we could have an artist’s rendition – (which is fine since space artists need work) – and it would show a spiral galaxy with lots of new star development.

    But on a scientific level, I wonder if there could be such a thing as a sophisticated, lens distortion deciphering program, based on a considerable back catalog of optical knowledge. Rather than stitch together fragmentary portions of the image, could such a computer program reconstruct images by reversing the distortions of the actual visual data?

  10. thaneb

    Does gravitational lensing exhibit dispersion of the constituent wavelengths/particles akin to a prism or diffraction grating? If so, is it used to study physical/chemical characteristics of the source?

  11. Pete Jackson

    @6Arthur Maruyama: I think I recall many years ago that a quasar had multiple images around some galaxy and that the variations in its radio emission were repeated in the various images after a time delay of only months.

    In any event, lensed galaxies like these should definitely be put on a supernova watch, if only because of the magnification of the light that will make detailed spectral studies of the progress of the supernova much easier to make.

  12. bouch

    Can someone tell me where that galaxy ACTUALLY is in that photo? If it appears to be in 3 places, I would think it would be somewhere in the middle, but I’ve got no clue if that’s actually the case…

  13. Sili

    How much mass in the lens?

  14. DrFlimmer

    @ #12 bouch

    You are right. The galaxy is (more or less) exactly behind the cluster. However, the cluster is much brighter than the background galaxy (its twice as far away as the cluster) and therefore invisible. The arcs are amplified a lot, as Phil explained, by the lens.

  15. Messier Tidy Upper

    Great photo – love it. :-)

    @13. Sili : “How much mass in the lens?”

    Could be mistaken but I’d say the mass of the cluster – galaxy cluster RCS2 032727-132623 – however much that is.

  16. Frank Snively

    Answer to Thaneb’s question:

    With very few exceptions, there will be no dispersion. The light follows a (null) geodesic through space-time, and since that is an inherent property of space, any light must follow the same path. Of course, intervening matter along the path (not the matter that is bending the light – that is somewhere else, in the cluster of galaxies) would affect the propagation a little bit; it is safe to assume there isn’t much matter along the geodesics.

  17. Messier Tidy Upper

    Continued @ 13. Sili : I’ve tried to find that galaxy cluster’s mass but had no luck. My google-fu has failed me, alas. :-(

    Anyone else know?

    @8. Acronym Jim :

    Ah, so maybe the main “easter egg” in Holbein’s painting, “The Ambassadors” wasn’t an anamorph. It was just warped by the gravity produced by the apparently massive egos of the titular subjects.

    LOL. Interesting, neat, painting. Cheers. :-)

    @4. Scott P. :

    I seem to see a second distorted galaxy — just left of center, about a third of the way up. Anyone else see it?

    Not sure. I see a few blue streaks+ which may or may not be part of the same untangled lensed galaxy we’re already discussing and what looks like a distorted and very blue barred spiral directly in line above the brightest elliptical of the cluster. Is that the one you were meaning?


    + Blue streaks NOT meaning the eponymous historic world land speed record setting rocket car! 😉

  18. don gisselbeck

    I’d love to see some AGW denier explain precisely why I should believe Phil on gravitational lensing but not on global warming.

  19. Paul A.

    I wasn’t quite convinced that gravity could bend space and gravitational lensing existed until I saw this picture.

  20. JB of Brisbane

    @Messier #17 – wasn’t that the Blue Flame (Gary Gabelich, 1973)? Blue Streak was the abortive British ICBM project.

  21. Puzzled Al

    If a galaxy is lensed then the light from that galaxy must travel farther than if had made a beeline to us. Would that have an effect on the perceived wavelengths of light? Is there a lot of lensing between us and billions of light years away? If wavelengths shift, could it be that this accounts for the red shift of distant entities? If that were true, could this have implications about an expanding Universe, the Big Bang, etc? Are these really stupid questions?

  22. Matt

    The bent path of the ligh IS the shortest path to us. Spacee is distorted by the gravity of the galexy cluster. Its just that our limited ability and our constraints about thinking in Euclidean three dimensions is hard to shed, so we think of the shortest distance as what we would think of as “line of sight” in non-distorted space.

    Anyway, Mother Nature is full of wonderfull things. This is pretty cool!

  23. @20. JB of Brisbane & #21. Neil Haggath :

    @Messier #17 – wasn’t that the Blue Flame (Gary Gabelich, 1973)? Blue Streak was the abortive British ICBM project.

    Oops. Of course! Yes, that’s the one I meant. Sorry. (Blushes.) :-(

    (Wiki-page now linked to my name here.)

    @22. Puzzled Al : Not stupid questions at all. Matt (#23) is correct though I think. There is a gravitational redshift as well as the distance-light travel time one but if my possibly unreliable (example above) memory serves, it ‘s only noticeable for supermassive objects, eg. Galactic Black Holes.

    @18. don gisselbeck : Indeed although I think the challenge is too much for them – ditto on why they’d listen to him about exoplanets, meteor showers & astrophysics more generally. Science works people. Everywhere. 8)

  24. Matt B.

    Correction for MTU (@24) re Puzzled Al (@22): While gravitational redshift is a phenomenon, it occurs when you observe something that’s at a lower gravitational potential (i.e. deeper in a gravity well) than you. Since the light started at mere stars and is received by us near a mere star, there’s too little difference in GP to change the wavelength anywhere near as much as the inflationary redshift, or even noticeably. The lensing really has nothing to do with that, since the light will come back “up” after going “down” into the galaxy cluster’s gravity well.

    The frequency of a photon will change by a factor of e when the photon traverses a gravitational potential difference of c^2. That is, f1/f0 = e^(ΔGP/c^2). If we observed light from the sun at an infinite distance (so that our GP is 0), the change in frequency would be a factor of e^((-GM/r)/(c^2)) = 0.999999783 (where M and r are the mass and radius of the sun).


Discover's Newsletter

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


See More

Collapse bottom bar