Galaxy Clusters: Definition

Clusters of galaxies are larger than groups, typically containing from about 50 to 1000 members. We refer to those clusters near the upper end of this range as rich clusters, while those near the lower end are termed poor clusters. We may further classify clusters according to whether they are regular (spherical with a central region of higher density) or irregular (amorphous in shape without a condensed central region).

An Abell Cluster

Galaxy Clusters: Description

Typical Properties of Clusters

Clusters typically have masses from 1014 to 1015 solar masses, the diameters are typically 8 Mpc, the velocity dispersions are 800-1000 km/s, and the average separation from other clusters is about 10 Mpc. In clusters the typical mass to light ratio is ~ 400 in units of solar masses divided by solar luminosities. Thus, they contain large amounts of dark matter.

Galaxy Clusters

Most galaxies are found in groups or clusters which contain a few to thousands of galaxies bound together by their mutual gravity. These clusters contain not only galaxies, but also contain the material which lies between the galaxies. The space between the galaxies in a cluster is filled with hot gas as well as invisible dark matter which has not yet been identified. According to studies of how the galaxies in a cluster move (which is influenced by the total mass in the cluster), this dark matter could have about five times the mass of all the galaxies and hot gas in a cluster combined. Galaxy clusters are classified by their appearance as being either regular or irregular clusters. They are also classified as rich or poor clusters depending on the number of galaxies they contain.

Regular Clusters

Regular clusters are spherically symmetric, with the galaxies concentrated toward the center of the cluster. They usually contain at least 1,000 galaxies which are brighter than an absolute magnitude of -16. Most of galaxies in regular clusters are elliptical galaxies. An example of a rich, regular cluster is the Coma Cluster (shown to the left), which contains thousands of galaxies. Almost every object in this image is a galaxy.

Irregular Clusters

Irregular clusters do not have a well-defined center, but are often made up of loose groups of small clusters. They contain all types of galaxies: spirals, irregulars and ellipticals. Irregular clusters can contain just a few to over 1,000 galaxies. The Local Group of galaxies, the cluster which contains our Milky Way galaxy, is an example of a poor, irregular cluster. Our Local Group contains about 30 galaxies: 3 spirals, 13 irregulars and 15 ellipticals.

A Galaxy Cluster showing Gravitational Lensing

Einstein's general theory of relativity shows that a very large amount of mass can bend the path of light and warp spacetime. This effect is seen in many rich, massive clusters of galaxies. The huge mass of large clusters acts as a gravitational lens, bending light which enters the cluster from objects lying far behind it. This bent light is then focused by gravity to create one or more images of the light source. The image might be distorted, or even magnified by the cluster's gravitational lens, depending on the position of the light emitting source relative to the cluster and to the observer. Galaxy clusters usually create these images in the form of distended arcs. Cluster gravitational lenses allow us to observe objects that are much too far or too faint to be seen directly, helping us view the very distant, early universe.

Galaxy Clusters: Distribution

The gist of the cosmological principle is that there are no preferred directions or locations in the universe. Present observations suggest that the cosmological principle is valid but only on distance scales in excess of 150 Mpc or so. On smaller scales, we do see spatial variations in galaxy properties (and possibly even in certain cosmological parameters) which are presumably tied to substantial fluctuations in the underlying distribution of matter in space. Early maps of the galaxy distribution on large scales, generated prior to the 1980s, were two dimensional— reproducing the distribution of galaxies on the plane of the sky. The third dimension, distance from our Galaxy, was only available for a small fraction of those galaxies. Nonetheless, these maps showed the distribution of galaxies to be highly non-random: when one galaxy is found, the likelihood of finding a nearby companion galaxy is significantly higher than Poisson statistics would predict.

Clustering in the universe is detected on many different scales. Our Milky Way Galaxy is a member of the local group of 24 galaxies spread over a region about 1 Mpc in radius. The Local Group contains a total mass in the range 3 × 1012 to 1013 solar masses. Groups like this are relatively common: on average, there is approximately one group for each 1800 cubic Mpc volume. Clusters of galaxies are approximately comparable in size, spanning 1–5 Mpc, but are considerably more massive and rarer than galaxy groups. Clusters contain hundreds of galaxies in their central regions. The total cluster mass is usually at least 1014 solar masses and the richest systems can be 50 times more massive than that. The mass density within a cluster is typically at least 100 times the universal closure density1. Clusters of galaxies are the largest systems known to have reached dynamic equilibrium, a state in which the cluster’s gravitational potential energy is twice its internal kinetic energy. Gravitationally bound structures may exist on larger scales still. Although quite rare, clusters of clusters (known as Superclusters) have been detected. These systems can contain in excess of 1016 solar masses, including up to 10 individual galaxy clusters spread across regions 10–30 Mpc in extent. Some superclusters are suspected of being bound systems but it is highly unlikely they have had sufficient time to reach dynamical equilibrium.

Structure on scales in excess of 100 Mpc was not discovered until wide-area, three-dimensional galaxy surveys were first undertaken in the early 1980s. Three dimensional surveys are essential for discovering the true nature of the galaxy, cluster, and supercluster distributions because projection effects hide substantial structures in two-dimensional. Indeed, an amazing bubble-like network of voids and sheets of galaxies and clusters, extending for hundreds of Mpc, is revealed in these spatial surveys. The clustering patterns that we see in the universe, from galactic scales and up, are believed to be governed by the gravitational potential of ubiquitous dark matter.

The galaxies, which comprise less than 5% of all the matter in the universe, are simply tracers of the dark matter distribution. This remarkable realization is based on observations which infer the total mass in galaxies and in clusters of galaxies from the distribution of orbital velocities of the material inside these systems. The dynamical mass estimates derived in this way are typically 5–10 times larger than the masses estimated by summing up the luminous components (e.g., stars in the case of galaxies; galaxies and hot gas in the case of clusters). The implication is that most of the matter in the universe has not yet been directly detected by conventional astronomical instrumentation.

The nature of this dark matter is still not well understood. The hypothesis that dark matter consists solely of Baryons (e.g., neutral gas, dilute plasma, massive planets, white dwarfs, neutron stars, and black holes) appears to be in conflict with observations of hot gas in clusters of galaxies (the gas comprises only about 10–20% of the total cluster mass), the tendency for dying stars to eject a substantial fraction of their mass back into interstellar space (and, hence, locking baryonic matter in stellar remnants is a very inefficient process), and limits on the number of massive planets, brown dwarfs, and very low mass stars from both gravitational microlensing surveys and deep Hubble Space Telescope images (the space density of such objects is too low to account for a significant fraction of the inferred total mass density in the universe).

The spatial distribution of galaxies and clusters can help constrain the nature of the dark matter because dramatically different configurations are predicted depending, for example, on whether or not the dark matter particles are highly relativistic (dubbed hot and cold dark matter, respectively). Candidates for nonbaryonic dark matter include neutrinos, axions, and the weakly interacting massive particles predicted by supersymmetry models (photinos, higgsinos, zinos, and solar cosmions). Programs to detect the more exotic of these cosmological non-baryonic particles are underway.

Voids, sheets, filaments, and spikes

The frothy spatial distribution of about 10 000 galaxies in two nearly opposing directions is shown in figure 3. The data were obtained at observatories located in Arizona and Chile. In this figure, known as a cone diagram, each galaxy’s position is shown as a point. Our own galaxy is located at the point where the wedges meet. The distance represented by the length of each wedge is about 120 Mpc. Several striking features are evident in this galaxy map (none of which were predicted by theory prior to the observations): large regions of very low galaxy density, somewhat misleadingly referred to as voids, extend for tens of Mpc. The voids are surrounded by a sheet-like network of galaxies often extending for several hundred Mpc. Indeed, one such coherent feature can be seen in figure 3 crossing from the mid-region of the northern cone diagram into and across the southern cone diagram. This feature has been dubbed the ‘Great Wall’. Clusters of galaxies can also be seen in the map as highly elongated clumps oriented towards the center of the diagram. This elongation is an artifact of the large, orbital velocities of galaxies about the cluster centers which can often approach 1000 km s−1. These orbital velocities distort the spatial positions determined from the redshift measurements and result in the elongated appearance of clusters in such diagrams. If we could observe the cluster galaxies using a map based on distances not dependent on the Redshift, such elongation would vanish.

Figure 3. The spatial distribution of approximately 10 000 galaxies extending out to a distance of 120 Mpc. Galaxies from a small portion of the northern galactic hemisphere are shown in the upper cone diagram and those from a comparable region in the southern galactic hemisphere are shown in the lower cone diagram. These data are from a large redshift survey coordinated by the Center for Astrophysics at Harvard University.

The structures in this map are clearly comparable with the depth of the survey and, thus, wider and deeper surveys were undertaken in the 1990s to assess just how large the largest structures in the universe are. One technological development which made very large redshift surveys feasible was the multi-object spectrograph (MOS). The 10,000 redshifts above were measured one at a time on telescopes which had narrow fields of view (1/7th the area subtended by the full Moon). Multi-object spectrographs today allow up to 600 redshifts to be measured simultaneously from galaxies spread over an area of a few square degrees (roughly 15 times the area subtended by the full Moon).

Figure 4. The spatial distribution of approximately 24,000 galaxies extending out to a distance of 600 Mpc. These are the results from the Las Campanas Redshift Survey conducted by an international team of astronomers using a Chilean telescope equipped with a multi-object spectrograph. The 3 bands going across the top and bottom show the projected distribution of the galaxies on the sky. Although the depth of this survey is substantial, it only subtends 0.3% of the entire sky.

One of the first large MOS surveys to be completed, known as the Las Campanas Redshift Survey (LCRS), measured about 24,000 redshifts in two directions 132 degrees apart. The distribution of galaxies from this survey is shown in figure 4. The depth of the LCRS was nearly 5 times greater than that in figure 3. Yet the structures found were comparable in size suggesting that perhaps there are no substantial inhomogeneities much larger than 120 Mpc. However, the LCRS only includes distances for a fraction of all the galaxies in this part of the sky due to time and telescope constraints. As a consequence, the structure in the LCRS map is not as clearly delineated as that in a survey which completely samples the galaxy distribution, such as the one in figure 3. Any conclusions regarding the existence of structures on scales larger than 120 Mpc or so must therefore be tempered by the fact that the survey has only done a partial sampling of the galaxy and cluster distribution.

The results of the most extensive redshift survey of clusters of galaxies is shown in figure 5. The radius of the survey is 240 Mpc and includes about 50% of the ~1000 clusters actually residing in this volume.

Figure 5. Three views of the spatial distribution of the richest ~480 clusters within a radius of 240 Mpc, centered on our Galaxy. Each dot represents a cluster and the dot size is proportional to the number of bright galaxies in the central region of the cluster. The empty wedges in the XZ and YZ projections are due to the blockage by the disk of our Galaxy.

Galaxy Clusters: Morphology

Zwicky distinguished compact, medium-compact, and open clusters, with or without strong central concentrations. The Rood and Sastry (RS) classification is based on the projected distribution of the brightest 10 members (being appropriately cautious about possible interlopers). They recognize these types:

The Rood and Sastry Classification

cD - single dominant cD - (galaxy A2029)
Cluster dominated by a single giant galaxy, often with a gigantic halo, perhaps a megaparsec in diameter. Very often an oval object, at least twice the size of the next brightest galaxy, with a dense halo of dwarf galaxies relatively nearby.

C - single core of galaxies
Cluster has a core containing at least three or four of the ten brightest cluster members which, on photo-plates, appear to dwarf the other cluster members.

B - dominant binary -- like Coma
Cluster dominated by a pair of relatively close giant galaxies, much larger and brighter than the other cluster members.

F - flattened -- (IRAS 09104+4109)
Some of the ten brightest cluster members are distributed in a flat configuration.

L - linear array of galaxies --(Perseus)
At least three of the cluster's ten brightest members are distributed in a line or chain across the sky.

I - irregular distribution -- (Hercules)
The ten brightest members are randomly distributed.

The Bautz-Morgan classification is based on brightness contrast between first- and second-ranked galaxies (i.e. the slope of the luminosity function at the bright end)


central cD galaxy (A2199)
Cluster dominated by a single supergiant galaxy


intermediate E/cD (Coma)
The brightest galaxies are intermediate between cD galaxies and ordinary giant elliptical galaxies.


no dominant galaxy (Virgo, Hercules)
The cluster contains no dominant galaxies.

Intermediate stages I-II,II-II are recognized.

Galaxy Clusters: Additional Information

First giant structures of the universe discovered

Astronomers using the South Pole Telescope (SPT) report that they have discovered the most massive galaxy cluster yet seen at a distance of 7 billion light-years. The cluster, designated SPT-CL J0546-5345, weighs in at around 800 trillion Suns and holds hundreds of galaxies. "This galaxy cluster wins the heavyweight title. It's among the most massive clusters ever found at this distance," said Mark Brodwin, from the Harvard-Smithsonian Center for Astrophysics in Cambridge, Massachusetts.

Located in the southern constellation Pictor the Painter, the cluster has a redshift of z=1.07. This puts it at a distance of about 7 billion light-years when the universe was half as old as now, and our solar system didn't exist yet.

Even at that young age, the cluster was almost as massive as the nearby Coma cluster. Since then, it should have grown about 4 times larger. If we could see it as it appears today, it would be one of the most massive galaxy clusters in the universe. "This cluster is full of old galaxies, meaning that it had to come together very early in the universe's history — within the first 2 billion years," said Brodwin.

Galaxy clusters like this can be used to study how dark matter and dark energy influenced the growth of cosmic structures. Long ago, the universe was smaller and more compact, so gravity had a greater influence. It was easier for galaxy clusters to grow, especially in areas that already were denser than their surroundings.

As the universe expanded at an accelerating rate due to dark energy, it grew more diffuse. Dark energy now dominates over the pull of gravity and chokes off the formation of new galaxy clusters.

Brodwin and his colleagues spotted their quarry in the first 200 square degrees of data collected from the new SPT. The SPT is currently completing its pioneering millimeter-wave survey of a huge swath of sky covering 2,500 square degrees.

They're hunting for giant galaxy clusters using the Sunyaev-Zel'dovich effect — a small distortion of the cosmic microwave background — a pervasive all-sky glow left over from the Big Bang. Such distortions are created as background radiation passes through a large galaxy cluster.

Surveying for this effect has significant advantages over other search techniques. It works just as well for very distant clusters as for nearby clusters, which allows astronomers to find rare, distant, massive clusters. Further, it provides accurate measurements of the masses of these clusters, which are crucial to unraveling the nature of dark energy. The main goal of the SPT survey is to find a large sample of massive galaxy clusters in order to measure the equation of state of the dark energy, which characterizes cosmic inflation and the accelerated expansion of the universe. Additional goals include understanding the evolution of hot gas within galaxy clusters, studying the evolution of massive galaxies in clusters, and identifying distant, gravitationally lensed, rapidly star-forming galaxies.

Once this distant cluster was found, the team studied it with the Infrared Array Camera on the Spitzer Space Telescope to pinpoint galaxies within the cluster. Detailed observations of the galaxies' speeds with the Magellan telescopes in Chile proved that the galaxy cluster was a heavyweight.

"After many years of effort, these early successes are very exciting," said Brodwin. "The full SPT survey, to be completed next year, will rewrite the book on the most massive clusters in the early universe."

Galaxy Members of the Local Group





l, b

D (kpc)

vr (km/s)

Milky Way

Sbc I-II



0 0







280 -33







303 -44



Sgr I




6 -14







237 -65



Sculptor dE




286 -84



Leo I




226 +49



Leo II




220 +67



Ursa Minor




105 +45







86 +35







260 -22







243 +42







121 -22



M32=NGC 221




121 -22



NGC 205




121 -21



NGC 185




121 -14



NGC 147




120 -14



And I




122 -25



And II




129 -29







119 -26



Cas=And VII



109 -09



Pegasus=DDO 216

Ir V



94 -43



Peg II=And VI




106 -36







126 -41







134 -31



NGC 6822




025 -18



IC 1613




130 -60



Sgr dI




21 +16







76 -74



IC 10




119 -03



DDO 210, Aqr

Im V



34 -31







272 -68







323 -48



Leo A=DDO 69

Ir V



196 52







101 -72