ENACS Survey of Southern Galaxies Indicates Open Universe
New Light on Rich Clusters of Galaxies and their Formation History
9 February 1996
In the context of a comprehensive Key-Programme , carried out with telescopes at the ESO La Silla Observatory, a team of European astronomers . has recently obtained radial velocities for more than 5600 galaxies in about 100 rich clusters of galaxies. With this programme the amount of information about the motions of galaxies (the kinematical data) in such clusters has almost been doubled. This has allowed the team to study the distribution of the cluster masses, and also the dynamical state of clusters in new and interesting ways.
An important result of this programme is that the derived masses of the investigated clusters of galaxies indicate that the mean density of the Universe is insufficient to halt the current expansion; we may therefore be living in an open Universe that will expand forever.
Clusters of galaxies as tracers of large-scale structure
About 40 years ago, American astronomer George Abell, working at the Palomar Observatory in California, was the first to perform a systematic study of rich clusters of galaxies, that is clusters with particularly many member galaxies located within a relatively restricted region in the sky. He identified several thousands of such clusters, and he numbered and described them; they are now known to astronomers as `Abell clusters'.
More than twenty years earlier, Swiss-American astronomer Fritz Zwicky, using the famous 100-inch Mount Wilson telescope above Los Angeles, concluded that the total mass of a rich cluster of galaxies is probably much larger than the combined mass of the individual galaxies we can observe in it. This phenomenon is now known as the `Missing Dark Matter', and many attempts have since been made to understand its true nature. Although the existence of this Dark Matter is generally accepted, it has been very difficult to prove its existence in a direct way. Rich clusters have several components: in addition to several hundreds, in some cases even thousands of galaxies (each with many billions of stars and much interstellar matter), they also contain hot gas (with a temperature of several million degrees) which is best visible in X-rays, as well as the invisible dark matter just mentioned. In fact, these clusters are the largest and most massive objects that are known today, and a detailed study of their properties can therefore provide insight into the way in which large-scale structures in the Universe have formed. This unique information is encoded into the distribution of the clusters' total masses, of their physical shapes, and not the least in the way they are distributed in space.
The need for a 'complete' cluster sample
Several of these fundamental questions can be studied by observing a few, or at the most several tens of well-chosen clusters. However, if the goal is to discriminate between the various proposed theories of formation of their spatial distribution and thus the Universe's large-scale structure, it is essential that uniform data is collected for a sample of clusters that is complete in a statistical sense. Only then will it be possible to determine reliably the distribution of cluster masses and shapes, etc.
For such comprehensive investigations, `complete' samples of clusters (that is, brighter than a certain magnitude and located within a given area in the sky) can be compiled either by means of catalogues like the one published by Abell and his collaborators and based on the distribution of optically selected galaxies, or from large-scale surveys of X-ray sources.
However, in both cases, it is of paramount importance to verify the physical reality of the presumed clusters. Sometimes several galaxies are seen in nearly the same direction and therefore appear to form a cluster, but it later turns out that they are at very different distances and do not form a physical entity. This control must be performed through spectroscopic observations of the galaxies in the candidate clusters. Such observations are crucial, as they not only prove the existence of a cluster, but also determine its distance and provide information about the motion of the individual galaxies within the cluster.
The ESO Nearby Abell Cluster Survey (ENACS)
Until recently, there existed no large cluster sample with extensive and uniform data on the motions of the individual galaxies. But now, in the context of an ESO Key-Programme known as the ESO Nearby Abell Cluster Survey or ENACS, the team of European astronomers has collected spectroscopic and photometric data for a substantial sample of more than one-hundred, rich and relatively nearby southern clusters from the Abell catalogue .
The extensive observations were carried out with the OPTOPUS multi-fibre spectrograph attached to the ESO 3.6-metre telescope at the La Silla Observatory, during 35 nights in the period from September 1989 to October 1993. With this very efficient spectrograph, the spectra of about 50 galaxies could be recorded simultaneously, dramatically reducing the necessary observing time.
In total, the programme has yielded reliable radial velocities for more than 5600 galaxies in the direction of about 100 rich clusters. The velocities were derived from a comparison of the observed wavelengths of absorption and emission lines with their rest wavelengths (the galaxy `redshifts'). Assuming a particular value of the `Hubble constant' (the proportionality factor between the velocity of a galaxy and its distance, due to the general expansion of the Universe), the distances of the galaxies can then be derived directly from the measured velocities.
The new observations approximately double the amount of data available for rich clusters of galaxies.
In combination with earlier data, the ENACS has produced a `complete' sample of 128 rich Abell clusters in a region centered near the south galactic pole (the direction which is perpendicular to the main plane of the Milky Way galaxy), and comprising about one-fifth of the entire sky. The sample extends out to a cluster distance of almost 1,000 million light-years (300 Mpc)
The space density of the 128 clusters is constant within the investigated volume, so that this sample is well suited to study, among others, the distribution of cluster masses. For a representative subset of 80 clusters, accurate information on the internal motions of galaxies in the clusters is available.
Most nearby and rich Abell clusters are real
In their pioneering work, Abell and his collaborators identified the clusters from visual inspection of photographic plates obtained with the Palomar telescopes . Some concern has frequently been expressed that an important fraction of the rich Abell clusters may not be real, but rather the result of chance superpositions in the sky of several smaller groups of galaxies. However, the data of the ENACS now prove conclusively that 90 percent of the rich, nearby Abell clusters are real: i.e. many of the galaxies observed in each of these clusters are indeed at the same distance and they form a physical entity.
Nevertheless, about one-quarter of the galaxies in the ENACS do not belong to the main clusters and reside in much smaller galaxy groups or are located in the vast space in between. This can be clearly seen in the distribution of the radial velocities in the direction of each of the clusters, shown in the diagramme attached to this Press Release.
When studying this distribution, it must be kept in mind, that the velocities of the galaxies in the clusters contain two components. The first is due to the general expansion of the Universe and depends only on the distance of the cluster; it is therefore the same for all galaxies in the cluster. The other reflects the individual motions of the galaxies within the cluster.
Cluster masses and the mean density of the Universe
The motions of the galaxies within a cluster makes it possible to estimate the total mass of the cluster: the greater the mass, the faster the motions must be in order to prevent the cluster from collapsing .
Using the data for the full sample of 128 clusters, the distribution of cluster masses has been derived. This distribution has been compared with predictions based on several models for the formation of large-scale structures in the Universe.
A very important result of the current work is that the observations do not support scenarios which are based on the assumption that the mean density of the Universe is equal to the `critical' value, i.e. the one which would correspond to a so-called `flat' Universe. The observed cluster masses are systematically smaller than those predicted in such models. Instead, the observed distribution of cluster masses seems to indicate that the mean density of the Universe is probably only a fairly small fraction of the critical value. This points to the Universe being `open' and ever-expanding.
Cluster formation may still be going on
The galaxies observed during the ENACS programme may be divided into two groups on the basis of their optical spectra, those that show clear emission lines and those that do not. The former are almost all late-type galaxies, that is spiral galaxies with ionized gas in their disks which gives rise to the emission lines. It appears that both the distribution within the cluster, as well as the velocities, of the galaxies with emission lines are significantly different from those of the galaxies without emission lines.
It seems that the emission-line galaxies have a tendency to avoid the central regions of their clusters, and their average radial velocities are about 20 percent larger than those of the non-emission galaxies. A plausible interpretation of these results is that a large part of the emission-line galaxies have not yet `mixed' with the other galaxies, and that they are approaching the central regions of their respective clusters for the first time. This may imply that the formation of at least a good fraction of the nearby, rich clusters is still going on.
If the mean density of the Universe is indeed much smaller than the critical density, as indicated by the cluster masses determined during this survey, then this is a quite unexpected result. One explanation may be that many clusters have only started to form fairly recently.
 The team is headed by Peter Katgert (Leiden Observatory, The Netherlands) and Alain Mazure (Laboratoire d'Astronomie Spatiale, Marseille, France); other members are Andrea Biviano and Roland den Hartog (Leiden Observatory, The Netherlands), Pierre Dubath (Observatoire de Geneve, Switzerland), Eric Escalera (SISSA, Trieste, Italy), Paola Focardi (Bologna University, Italy), Daniel Gerbal (Institut d'Astrophysique, Paris, France), Guilano Giuricin (SISSA, Trieste, Italy), Bernard Jones (Theoretical Astrophysics Centre, Copenhagen, Denmark), Olivier Le Fevre (Meudon Observatory, Paris, France), Mariano Moles and Jaime Perea (Astrophysics Institute of Andalucia, Granada, Spain), and George Rhee (University of Nevada, Las Vegas, U.S.A.).
 This Press Release is accompanied by eso9607a showing one of the rich clusters, as observed with the ESO 1-metre Schmidt telescope.
 The masses of the planets in the solar system are determined in a similar way from the motions of their moons. The faster the moon moves around the planet at a given distance, the heavier is the planet.
The Velocities of 5634 Southern Galaxies eso9608a; 9 February 1996
This diagramme gives an overview of the velocities of 5634 individual galaxies which were measured during the ESO Nearby Abell Cluster Survey (ENACS) in the directions of 107 rich and nearby Abell clusters. Each bar plot refers to an individual cluster, the designation of which is shown to the left, together with the total number of galaxy redshifts measured for the cluster.
Each vertical line represents one galaxy, and the position in the bar indicates the measured value of its velocity according to the scale at the bottom. Galaxies that show emission lines in their spectrum are indicated by dashed vertical lines; the others are fully drawn.
Rich Abell clusters appear as clumps of galaxies in which the individual velocities only differ as a result of the internal motions in the cluster, so that the galaxies are very close in the bar plots. Such highly clumped distributions are observed for the majority of the clusters studied here. This means that about that 90 percent of the rich nearby Abell clusters are real physical entities.
About three-quarters of the galaxies observed during ENACS reside in the dominant clump (the cluster); about half of the remaining one-quarter are found in small groups, and the last are in the space in between. The galaxies with emission lines in their spectra (the spiral galaxies) have a preference to be located in small groups or at the periphery of the rich clusters.
Since their average velocity is proportional to their distance, the distribution of the average velocities of the clumps in this diagramme also gives information about the spatial distribution of the clusters in the Universe.