The Expanding Universe

Looking at a map of the distribution of galaxies in the local Universe, the eye is surprised to find that the galaxies do not seem to be laid down in a uniform or random fashion, but seem to cluster. It is not too surprising that the galaxy distribution is not uniform, for this would imply that the matter in the Universe was organized like a huge crystal and there are no known laws of physics that would lead to such a result.

To understand clustering consider a distribution of ducks on a pond. Some ducks will be wandering off by themselves, but most of the ducks, most of the time, are found in small groups of two or three or more that tend to travel around together. Ducks tend to cluster and this clustering is not a result of chance. Ducks have a tendency to want to be near one another.

Galaxies also tend to cluster. When looking at maps of the galaxy distribution it is generally the case that if you find one galaxy, it is very likely that there is another galaxy close by, as with ducks in a pond. Why galaxies are clustered and to what degree, is a very important question in cosmology.

Answering the question of whether the matter in the Universe is distributed randomly is subtler. The reason for this is that even if the galaxies were placed randomly throughout the Universe, some clustering would appear simply by chance even in the absence of any physical process that would tend to cluster galaxies. One physical process that functions to cluster galaxies is gravity. Since gravity is attractive, it will attract galaxies to one another. So as the Universe ages and evolves, it is expected that galaxies will become more and more clustered.

Clustering is understood and measured in terms of statistics. For example, if a research biologist wants to study the clustering behavior of ducks, it is necessary to study many groups of ducks at different times on many different ponds. On the other hand, if an astronomer wants to study the clustering of galaxies, it is necessary to have a very large systematic survey or map of the distribution of galaxies. This is why a survey like the Sloan Digital Sky Survey is so important to understanding how matter is distributed in the Universe.

A fundamental question in cosmology is at what scale do galaxies, or clusters or galaxies, appear to be randomly distributed? In other words, how large does a sample volume of the Universe have to be in order that our measurements of the mean density settle down and become well defined? This is a necessary condition for the Cosmological Principle to hold on large-scales.

It is a very modern notion that the Universe could have random properties at all. Until the last fifty years or so, astronomers and physicists still looked to the Heavens to find an unblemished example of the laws of Nature operating directly on the primordial stuff of creation. They didn't expect to find the secrets of the formation and evolution of the Universe by studying the Earth or the Solar System, or even the Milky Way. All of these each had their own idiosyncratic histories of formation and evolution, and it was obvious that in cosmological terms, the scale of these objects was actually very small.

But when looking on the largest and grandest scales, cosmologists expected to find well-ordered and sensible phenomena. Surely, they did not expect the Universe to be perfectly organized and ordered like the ancient Greeks, but they believed that once they gazed out beyond their local neighborhood, within a few hundred million light-years that is, that the average properties of the Universe would become quite predictable.

Astronomers were not very surprised to find that our own galaxy, The Milky Way, was a member of a group of some twenty galaxies. They were also not too surprised to find that that our local group was a member of a cluster of some two thousand galaxies. But as larger and deeper sky surveys were done in the 1980s and 1990s, astronomers were surprised to find that there were clusters of clusters, or superclusters, of galaxies that formed huge walls and thin sheets that surrounded large voids, areas with very few galaxies. On the largest scales seen so far, the distribution of galaxies looks like a gigantic foam of bubbles with the structure of a sponge.

Presently, one of the basic questions of cosmology is, "When does this clustering stop?" There are clusters and superclusters, and clusters of superclusters, but does this organization continue to superclusters of superclusters, etc.?

This question is of central importance to understanding the birth and evolution of the Universe. Some of the most basic predictions of theories of the early Universe concern how the matter was initially laid down, or distributed. Since the distribution of galaxies observed today evolved out of this initial distribution, a knowledge of the distribution and clustering properties of galaxies on very large scales today is one of the few strong tests of theories of the early Universe.

The Sloan Digital Sky Survey was designed precisely to make this fundamental measurement. By observing a very large, deep area in a very systematic fashion, scientists should be able to measure the level of clustering on all these scales and use their results to constrain theories of the early Universe. How do Cosmologists understand Galaxy Clustering?

It is often asked, "If the distribution of galaxies is different everywhere astronomers look, how is it possible to understand galaxy clustering?" This question is most easily answered by comparing it to other phenomena, which many people find very familiar, called random noise processes.

Examples of random noise processes are the sound of static on an old radio, or the sound of a waterfall, or the distribution of waves on the surface of the sea. In each of these cases, every time you listen or look, what you hear or see is different than what you heard or saw before. However, it is also obvious that, in some sense, you are hearing the same waterfall or radio, or looking at the same sea.

What is the same from time to time in each of these cases are the statistical properties of the sound or waves. Taking the sea as an example, although the surface or the sea is always changing, the distribution of the number of waves and their heights have some well-defined average properties. By observing a very large portion of the sea at once, or a small portion for a long period of time, it becomes possible to characterize the general properties of sea waves.

The astronomers working with the Sloan Digital Sky Survey will do the same type of analysis on the distribution of galaxies revealed in their map of the Universe. And just as the depth of the water, or the strength of the wind determines the characteristics of the waves in a part of the sea, the strength and amount of galaxy clustering tells cosmologists a lot about how matter was laid down in the early Universe and what physical processes have evolved the clustering since that time.

Knowledge of galaxy clustering can give cosmologists information on other fundamental properties of the Universe as well. For example, cosmologists will be able to use this data to measure the density of the Universe, constrain theories of Dark Matter, and predict the ultimate fate of the Universe, that is, will it expand forever or is it inevitably destined to one day recollapse upon itself.