[Paleopsych] Science: Mapping the large-scale structure of the universe

Fri Jul 22 19:08:01 UTC 2005

Mapping the large-scale structure of the universe
http://www.sciencemag.org/cgi/content/full/309/5734/564
Science, Vol 309, Issue 5734, 564-565 , 22 July 2005
[DOI: 10.1126/science.1115128]
[Thanks to Eugen for this.]

Mapping the Large-Scale Structure of the Universe
David H. Weinberg*

In a large-scale view of the universe, galaxies are the basic unit of 
structure. A typical bright galaxy may contain 100 billion stars and span tens 
of thousands of light-years, but the empty expanses between the galaxies are 
much larger still. Galaxies are not randomly distributed in space, but instead 
reside in groups and clusters, which are themselves arranged in an intricate 
lattice of filaments and walls, threaded by tunnels and pocked with bubbles. 
Two ambitious new surveys, the Two-Degree Field Galaxy Redshift Survey (2dFGRS) 
and the Sloan Digital Sky Survey (SDSS), have mapped the three-dimensional 
distribution of galaxies over an unprecedented range of scales (1, 2). 
Astronomers are using these maps to learn about conditions in the early 
universe, the matter and energy contents of the cosmos, and the physics of 
galaxy formation.

Galaxies and large-scale structure form as a result of the gravitational 
amplification of tiny primordial fluctuations in the density of matter. The 
inflation hypothesis ascribes the origin of these fluctuations to quantum 
processes during a period of exponential expansion that occupied the first 
millionth-of-a-billionth-of-a-trillionth of a second of cosmic history. 
Experiments over the last decade have revealed the imprint of these 
fluctuations as part-in-100,000 intensity modulations of the cosmic microwave 
background (CMB), which records the small inhomogeneities present in the 
universe half a million years after the big bang. Although the visible 
components of galaxies are made of "normal" baryonic matter (mostly hydrogen 
and helium), the gravitational forces that drive the growth of structure come 
mainly from dark matter, which is immune to electromagnetic interactions.

By combining precise, quantitative measurements of present-day galaxy 
clustering with CMB data and other cosmological observations, astronomers hope 
to test the inflation hypothesis, to pin down the physical mechanisms of 
inflation, to measure the amounts of baryonic and dark matter in the cosmos, 
and to probe the nature of the mysterious "dark energy" that has caused the 
expansion of the universe to accelerate over the last 5 billion years. The 
2dFGRS, completed in 2003, measured distances to 220,000 galaxies, and the SDSS 
is now 80% of the way to its goal of 800,000 galaxies (see the figure). The key 
challenge in interpreting the observed clustering is the uncertain relation 
between the distribution of galaxies and the underlying distribution of dark 
matter. If the galaxy maps are smoothed over tens of millions of lightyears, 
this relation is expected to be fairly simple: Variations in galaxy density are 
constant multiples of the variations in dark matter density. Quantitative 
analysis in this regime has focused on the spatial power spectrum, which 
characterizes the strength of clustering on different size scales (3, 4). The 
power spectrum describes the way that large, intermediate, and small 
structures--like the mountain ranges, isolated peaks, and rolling hills of a 
landscape--combine to produce the observed galaxy distribution. The shape of 
the dark matter power spectrum is a diagnostic of the inflation model, which 
predicts the input spectrum from the early universe, and of the average dark 
matter density, which controls the subsequent gravitational growth. Recent 
analyses have also detected subtle modulations of the power spectrum caused by 
baryonic matter, which undergoes acoustic oscillations in the early universe 
because of its interaction with photons (4, 5).

To go further, one would like to know the precise amplitude of dark matter 
clustering, not just the variation of clustering with scale. Unfortunately, the 
factor relating galaxy and dark matter densities depends on aspects of galaxy 
formation that are difficult to model theoretically. One observational approach 
to isolating dark matter clustering uses weak gravitational lensing, in which 
the dark matter surrounding nearby galaxies subtly distorts the apparent shapes 
of the distant galaxies behind them. Another approach uses the relative 
probabilities of different triangular configurations of galaxy triples, picking 
out the characteristic filamentary geometry produced by anisotropic 
gravitational collapse. Applications of these two methods to the SDSS and the 
2dFGRS, respectively, imply that the clustering strength of bright, Milky 
Way-type galaxies is similar to that of the underlying dark matter (6, 7).

     Figure 1 The big picture. Large-scale structure in the SDSS. The SDSS uses 
a mosaic charge-coupled device camera to image regions of the sky and a 
fiber-fed spectrograph to measure distances of galaxies selected from these 
images. The main panel shows the SDSS map of 67,000 galaxies that lie within 5° 
of the equatorial plane; the region of sky obscured by the Milky Way is not 
mapped. Each wedge is 2 billion light-years in extent. Galaxies are color-coded 
by luminosity, and more luminous galaxies can be seen to greater distances. 
(Inset) SDSS image of a cluster of galaxies, showing a region roughly 1 million 
light-years on a side.

     CREDIT: IMAGES COURTESY OF THE SLOAN DIGITAL SKY SURVEY COLLABORATION

On smaller scales, the relation between galaxies and dark matter becomes more 
complex, and it is different for different types of galaxies. Redder galaxies 
composed of older stars reside primarily in clusters, the dense urban cores of 
the galaxy distribution. Younger, bluer galaxies populate the sprawling, 
filamentary suburbs. Current efforts to model galaxy clustering in this regime 
focus on the "halo occupation distribution," a statistical description of the 
galaxy populations of gravitationally bound "halos" of dark matter. Depending 
on its mass, an individual dark halo may play host to a single bright galaxy, a 
small group of galaxies, or a rich cluster.

By combining theoretical predictions for the masses and clustering of halos 
with precise measurements of the clustering of galaxies, one can infer the halo 
occupation distribution for different classes of galaxies empirically. 
Theoretical models of galaxy formation predict a strong dependence of halo 
occupation on galaxy luminosity and color, and the initial results from the 
2dFGRS and the SDSS show good qualitative agreement with these predictions (8, 
9). Increased precision and measurements for more galaxy classes will test the 
predictions in much greater detail, and they will sharpen our understanding of 
the physical processes that produce visible galaxies in the first place and 
determine their observable properties. By deriving the relation between 
galaxies and dark matter from the clustering data themselves, halo occupation 
methods also allow new cosmological model tests that take advantage of precise 
measurements on small and intermediate scales.

The large-scale clustering results from the 2dFGRS and the SDSS, in combination 
with CMB measurements and other cosmological data, support the predictions of a 
simple inflation model in a universe that contains 5% normal matter, 25% dark 
matter, and 70% dark energy (10). However, several analyses that incorporate 
smaller scale clustering suggest that either the matter density or the matter 
clustering amplitude is lower than this "concordance" model predicts, by 30 to 
50% (11-13). This tension could reflect systematic errors in the measurements 
or the modeling, but it could also signal some departure from the simplest 
models of primordial fluctuations or dark energy. For example, if inflation 
produces gravity waves that contribute to observed CMB fluctuations, then naïve 
extrapolation of these fluctuations would overpredict the level of matter 
clustering today. Alternatively, evolution of the dark energy component can 
affect the amount of growth since the CMB epoch. As the SDSS moves toward 
completion, improved clustering measurements and analyses may restore the 
consensus on a "vanilla" cosmological model, or they may provide sharper 
evidence that our theoretical recipe for the universe is still missing a key 
ingredient.

References and Notes

    1. M. Colless et al., Mon. Not. R. Astron. Soc. 328, 1039 (2001). [ADS]
    2. D. G. York et al., Astron. J.120, 1579 (2000). [ADS]
    3. M. Tegmark et al., Astrophys. J. 606, 702 (2004). [ADS]
    4. S. Cole et al., http://arxiv.org/abs/astro-ph/0501174.
    5. D. J. Eisenstein et al., http://arxiv.org/abs/astro-ph/ 0501171.
    6. E. Sheldon et al., Astron. J.127, 2544 (2004). [ADS]
    7. L. Verde et al., Mon. Not. R. Astron. Soc. 335, 432 (2002). [ADS]
    8. F. C. van den Bosch, X. Yang, H. J. Mo, Mon. Not. R. Astron. Soc. 340, 
771 (2003). [ADS]
    9. I. Zehavi et al., http://arxiv.org/abs/astro-ph/0408569.
   10. Convergence on this model from several independent lines of evidence was 
the 2003 Science "Breakthrough of the Year," C. Seife, Science 302, 2038 
(2003).
   11. F. C. van den Bosch, H. J. Mo, X. Yang, Mon. Not. R. Astron. Soc. 345, 
923 (2003). [ADS]
   12. N. A. Bahcall et al., Astrophys. J. 585, 182 (2003). [ADS]
   13. J. L. Tinker, D. H. Weinberg, Z. Zheng, I. Zehavi, Astrophys. J., in 
press; preprint available at http://arxiv.org/abs/astro-ph/0411777.
   14. I thank the National Science Foundation for support.

The author is in the Department of Astronomy, Ohio State University, 140 West 
18th Avenue, Columbus, OH 43210, USA E-mail: dhw at astronomy.ohio-state.edu