Researchers based at the Institute for Computational Cosmology (ICC) in
Durham and at Caltech in California, have found striking proof that their
computer simulations of the universe can accurately predict how galaxies are
clustered, so helping to reveal the distribution of dark matter throughout
the universe. Using a computer simulation to follow the formation and
evolution of galaxies in a universe filled with dark energy and dark matter,
they predicted that the most luminous galaxies should be associated with the
most massive concentrations of dark matter and, as a consequence, these very
bright galaxies should be clustered more tightly than average galaxies.
After testing the predictions using data on thousands of galaxies from the
Anglo-Australian telescope 2-degree Field Galaxy Redshift Survey (2dFGRS),Dr
Peder Norberg of the University of Durham, working with other members of the
2dFGRS team, found that their predictions were spot on, proving that the
Cosmology Machine supercomputer is a powerful tool for understanding how the
universe works. These results will be presented on Wednesday 10 April 2002
at the National Astronomy Meeting in Bristol by team member Dr Ofer Lahav of
Cambridge University.

For a long time, cosmologists have held the belief that the bulk of the mass
in the universe is in the form of “cold dark matter”. This material does not
emit any light but astronomers are convinced it really is there because
something invisible exerts a gravitational pull on luminous objects
astronomers can see, such as stars and galaxies. Several experiments,
including one operated by the UK Dark Matter Collaboration at the Boulby
mine in North Yorkshire, are racing to be the first to detect the exotic
elementary particles that are the likely candidates for the cold dark
matter. On top of that, recent observations have suggested that the universe
is filled with a mysterious “dark energy”, which produces a repulsive force.
Currently, dark energy is winning out over the gravitational pull of the
matter, both dark and luminous, with the result that the expansion of the
universe is speeding up.

Theorists expect the dark matter in the universe to be clumpy due to the
influence of gravity, which boosts small primordial ripples in the
distribution of mass. The formation of structure in the dark matter – dubbed
“the cosmic web” – can be followed with computer simulations. However, one
of the biggest challenges facing cosmologists is to predict where galaxies
should form and light up the cosmic web of dark matter. The team from the
Institute for Computational Cosmology (ICC) at the University of Durham (Dr
Carlton Baugh, Dr Shaun Cole, Prof. Carlos Frenck and Dr Cedric Lacey) and
the California Institute of Technology (Dr Andrew Benson) have now come up
with a model that can follow the formation and evolution of galaxies in a
universe filled with dark energy and dark matter. The results, involving
some of the largest simulations run on the Cosmology Machine supercomputer
at the ICC, predict how galaxies are arranged in space and how tightly they
are clustered. One basic prediction is that the most luminous galaxies
should be associated with the most massive dark matter structures. This in
turn means that these very bright galaxies are expected to cluster together
more than average galaxies.

Because these luminous galaxies are extremely rare, the theoretical
predictions have been hard to test until now. The problem of finding enough
very bright galaxies has been solved by an Anglo-Australian collaboration
that has produced a map of the local universe of unprecedented size. The
Two-Degree Field Galaxy Redshift Survey (2dFGRS), named after the instrument
used to make the map on the Anglo-Australian Telescope at Siding Spring in
New South Wales, Australia, contains over 250 000 galaxy redshifts, more
than ten times more than any survey completed in the last millennium, and
has uncovered plenty of the precious bright galaxies. The 2dFGRS team
measured the degree of clustering of very luminous galaxies and found that
these galaxies have a much clumpier distribution than more typical galaxies
in the survey. Remarkably, the level of the enhancement in clustering was
exactly that predicted by the computer model.

Aware that galaxies in which relatively few new stars are forming tend to be
more numerous among the brightest galaxies, the team wanted to be quite sure
their observational result was really due to changes in luminosity and not
simply a case of ‘comparing apples with oranges’. So they went on to use the
huge wealth of data from the 2dfGRS to see whether the clustering of
galaxies depends as much on their type as their brightness. “This is only
feasible in a survey with as many galaxies as the 2dFGRS” says 2dFGRS team
member Carlton Baugh. “Whilst we found a small difference in the clustering
strength of different types of galaxy with the same luminosity, the
overriding trend is that the clustering signal is most sensitive to the
luminosity of the galaxy. These exciting results mean that cosmologists are
starting to get to grips with the problem of lighting up their dark


Dr Carlton Baugh, Institute of Computational Cosmology, Dept of Physics,
University of Durham, DH1 3LE. Tel: (+44) (0)191 374 2142

Dr Peder Norberg, Institute of Computational Cosmology, Dept of Physics,
University of Durham, DH1 3LE. Tel: (+44) (0)191 374 1662

Dr. Ofer Lahav, Institute of Astronomy, University of Cambridge, Madingley
Road, Cambridge CB3 0HA. Tel. (+44) (0) 1223 337540.

2dFGRS coordinators:

Dr Matthew Colless, Research School of Astronomy and Astrophysics,
National University. Tel. (+61) 2 6125 8030.

Prof John Peacock, University of Edinburgh Tel. (+44) (0)131 668 8390.
Fax (+44) (0) 131 668 8416. email

Dr Steve Maddox, School of Physics & Astronomy, University of Nottingham,
Tel. (+44) (0)115 9515 133, Fax (+44) (0)115 9515 180


1.The work described in this release is based on two papers, one published
recently and the other accepted but not yet published:

The 2dF Galaxy Redshift Survey: luminosity dependence of galaxy clustering
by P. Norberg et al. (the 2dFGRS team). Monthly Notices of the Royal
Astronomical Society, Volume 328, pages 64-70, November 2001.

The 2dF Galaxy Redshift Survey: The dependence of galaxy clustering on
luminosity and spectral type by P. Norberg et al. (the 2dFGRS team). Monthly
Notices of the Royal Astronomical Society, in press.
Abstract available at

2. Designed and built by the Anglo-Australian Observatory, the 2dF
instrument is one of the world’s most complex astronomical instruments, able
to capture 400 spectra simultaneously. A robot arm positions up to 400
optical fibres on a field plate, each to within an accuracy of 20
micrometres. Light from up to 400 objects is collected and fed into two
spectrographs for analysis. The expansion of the Universe shifts galaxy
spectra to longer wavelengths. By measuring this ‘redshift’ in a galaxy’s
spectrum, the galaxy’s distance can be determined.

The 2dF survey covers a total area of about 2,000 square degrees, selected
from both northern and southern skies.

The 2dF galaxy redshift survey website, including a fly-through movie of the
survey, is at

3. Other sources of information and images

Images of the simulations and of the 2dFGRS can be found at

Further images of the computer simulations are available at:

The Institute for Computational Cosmology at Durham:

The UK Dark Matter Collaboration:

4. UK National Astronomy Meeting Web site:

5. RAS Web site: