[watsisname]

The history of astronomy is that of receding horizons, and repeated realizations that there are structures at larger and larger scales than previously thought.  (And of course particle physics does the same for smaller scales) 
Particles --> Atoms --> [various intermediate things] --> Planets --> Star Systems --> Galaxies --> Galactic Clusters ... where does it end?  What are the largest structures in the universe?  Is there an endless hierarchy of structures at ever increasing scales?  Is there a way to know?

Apparently, the results of the recent Dark Energy Survey demonstrate that the universe is indeed homogenous at scales beyond that of galactic superclusters.  There is no bigger structure¹.  If correct, this is a very significant validation of some key assumptions of General Relativity.

Article

We know that stars group together to form galaxies, galaxies clump to make clusters and clusters gather to create structures known as superclusters. At what scale though, if at all, does this Russian doll-like structure stop? Scientists have been debating this very question for decades because clustering on large scales would be in conflict with our 'standard model' of cosmology. The current model is based on Einstein's equations assuming everything is smooth on the largest scales. If matter were instead clumpy on very large scales, then the entire model would need to be re-thought.

Cosmologists agree that on 'small' scales (tens of millions of light years), matter in the Universe is highly clustered. So the 'standard model' can only hold true if the Universe transitions to an even distribution of matter (homogeneity) on larger scales, irrespective of the viewing direction. However, some scientists have recently argued that the entire Universe never becomes homogenous, and that it is clustered on all scales, much like one of Mandelbrot's famous 'fractals' (snowflakes and broccoflowers are good examples of fractals). If the Universe has properties similar to a fractal, our description of space and time is wrong, and our understanding of things like Dark Energy is deeply flawed.

New data from a recently completed galaxy survey was published last night by a CAASTRO PhD student at the International Centre for Radio Astronomy Research (ICRAR) and The University of Western Australia in Perth and her colleagues. This paper might finally put an end to this long running debate.

Using the Anglo-Australian Telescope, Ms Morag Scrimgeour has found that on distance scales larger than 350 million light years, matter is distributed extremely evenly, with little sign of fractal-like patterns.

“We used a survey called WiggleZ which contains more than 200,000 galaxies, and probes a cosmic volume of about 3 billion light years, cubed,” Ms Scrimgeour explains “This makes it the largest survey ever used for this type of measurement of the large scale Universe.”

This finding is extremely significant for cosmologists as it confirms that the tools being used to describe the Universe are the right tools for the job after all. Had evidence been found confirming large-scale fractals, it would have left cosmologists without a working model for the Universe, sending them back to the drawing board to painstakingly adjust theories. [etc]

1:  Disclaimer:  This might not disprove inhomogeneity beyond Hubble-Volume scales; I'm actually not sure if that's possible or not.  Either way it doesn't matter much for the purposes of understanding and describing our own Hubble Volume.

[Mongoose]

probes a cosmic volume of about 3 billion light years, cubed

That is a ****-load of space.

[Nuke]

i was right! everything is a fractal!

[The E]

i was right! everything is a fractal!

Except that these findings seem to prove that theory wrong...

[Nuke]

dont argue with a drunk

[Dilmah G]

I'm not an astronomer, but I'm going to chime in here and express delight at the fact that my university is actually doing something. UWA rep!

[Dragon]

Seems like physics are getting awfully predictable nowdays. Higgs was where we expected it, universe is homogeneous on large scales, as predicted...

[Colonol Dekker]

Next stop, black hole warheads which accidentally destroy us all

[Nuke]

i want mushroom clouds that look like boobs

[Mika]

3 billion light years cubed? So the side length of the equivalent volume cube is 3*10^9 ly or the cubic root of 3*10^9 ly^3?

[headdie]

without other qualifiers 3 billion light years cubed means the volume of a cube with 3 billion light year long sides

H*W*D so 3 billion light years * 3 billion light years * 3 billion light years.

the thought of that number makes my head hurt

[Mongoose]

I think it'd be more plausible that she meant "cubic light-years," though I guess you never know.

[watsisname]

I can't figure out where the '3 billion light years cubed' is coming from in that article; either I don't understand how cosmic volumes are calculated, or that might be a misquote.  The abstract of their paper says "a cosmic volume of ~1 (Gpc/h)^3".  h is about 0.73 (give or take about 0.03) and a Gpc is 3.26 billion light years.  That would imply a volume of 89 GLY³, or a cube with side length ~4.5 billion light years.

For comparison, the observable universe has a co-moving volume of 4/3*π*(46GLY)³ or ~410000 GLY³.  So this study examined about 0.02% of the universe. 

Holy **** space is big.

[Mongoose]

And that's just the part we can see.  Who knows what the hell is going on beyond that.

[karajorma]

For comparison, the observable universe has a co-moving volume of 4/3*π*(46GLY)³ or ~410000 GLY³.  So this study examined about 0.02% of the universe. 

For comparison how much space does a supercluster take up? Without knowing the ratio it's hard to tell if there really is homogeneity or if they just didn't study a big enough area.

[Dragon]

I can't figure out where the '3 billion light years cubed' is coming from in that article; either I don't understand how cosmic volumes are calculated, or that might be a misquote.  The abstract of their paper says "a cosmic volume of ~1 (Gpc/h)^3".  h is about 0.73 (give or take about 0.03) and a Gpc is 3.26 billion light years.  That would imply a volume of 89 GLY³, or a cube with side length ~4.5 billion light years.

For comparison, the observable universe has a co-moving volume of 4/3*π*(46GLY)³ or ~410000 GLY³.  So this study examined about 0.02% of the universe. 

Holy **** space is big.

Yes, sounds about right. 3 billion LY cubed most likely comes from the writers not knowing what "h" is supposed to stand for and just going with gigaparsecs (also somewhat obscure, but easier to google) cubed. Space is indeed pretty big, and mostly empty to boot. Makes you feel rather small in comparison.

And that's just the part we can see.  Who knows what the hell is going on beyond that.

It's more complicated than that. With distance, you also move backwards in time. You don't even know what's going on Mars right now, but you do know what was going on two minutes ago (as soccer fans know, two minutes are more than enough to change everything ). If you go into advanced physics, you need to start thinking in terms of spacetime, because separation of those concepts is only an illusion.

[watsisname]

For comparison how much space does a supercluster take up? Without knowing the ratio it's hard to tell if there really is homogeneity or if they just didn't study a big enough area.

Very good question.  Let's see:  Superclusters and filaments are typically on the order of a few hundred million light years across, and the largest ones can be almost a billion light years long.  It's not obvious how to compute the volume of these things, since they're rather aspherical, but a sphere containing one a billion LY across would have a volume of 0.5 GLY³, which is still quite a bit smaller than what the survey looked at.  The survey also examined several different portions of the sky, so I imagine it's a fairly representative sample of the universe at large.

Apparently the method for determining homogeneity is to see if the number of galaxies increases as the cube of the distance from a specific point (presumably Earth), and the data shows that indeed it does for sufficiently large distances.  I assume that has to take the expansion of the universe into account, since the universe gets less dense on average over time and the farther out we look the farther back in time we see.