One of my main projects in recent years has been the 2dF quasar redshift survey, which produced the largest ever homogeneous quasar survey, with 23,000 redshifts (the Sloan Digital Sky Survey now has more QSOs but is not yet complete). The survey only covers part of the sky: there are two surveyed regions, each of which is a slice through the universe, and one of those slices is shown here.
There has been debate for decades about whether such large structures exist
and can be detected in the distribution of QSOs. I can remember coming back
on a plane from
La Palma in 1989 with
Paul Mitchell
convinced that we'd found the largest structures in the Universe
(we hadn't, it was just noise). It has taken a huge survey such as 2QZ to
obtain a large enough statistical sample to enable structures to be seen.
And now we can quantify what it means: the excess numbers found are just
what is expected from the standard model for large-scale structure in the
universe, the "biased Lambda cold dark matter model".
One of the other, related, main quantitative results of this study is the ability to measure the spatial power spectrum of the distribution of quasars on the largest scales yet probed by any redshift survey. Because quasars can be seen at significant cosmological distances, we can also measure how the strength of clustering has varied with cosmic epoch. Surprisingly, to a naive observer, things were more strongly clustered in the past than they are now, despite the growth of structure by gravitational instability. This result tells us how the bias of massive galaxies, relative to the underlying dark matter distribution, evolves with cosmic time.