8. Clusters and rich systems
Within our survey lie the centers of 9 ACO clusters, 5 ACOS clusters and 12 EDCC clusters. Several entries of the three lists correspond to the same cluster. Taking into account multiple identifications, there are 20 clusters listed within one or more of the three catalogs that lie within ESP.
In our catalog we find at least one counterpart for 17 out of the 20 clusters. The three clusters that do not correspond to any of our systems are ACO 2860, ACOS 11, and ACOS 32, all of Abell richness R = 0. ACO2860 is a very nearby object with a redshift, z = 0.0268, close to our minimum redshift. ACOS 11 and ACOS 32 are both distance class D = 6 objects that may be either projection effects or real clusters located beyond our redshift limit. We select the ESP counterparts among the groups that are close to the clusters on the sky and that have a redshift compatible with the distance class and/or magnitude of the cluster. If more ESP groups are counterparts of a cluster, we identify the cluster with the richest counterpart.
In Table 3 we list the name (column 1), and the coordinates (columns 4, 5) of the 17 clusters with ESP counterpart together with their richness (column 6) and, if available, their redshift as estimated by Zucca et al. (1993) (column 7). In the case of clusters listed in both EDCC and ACO or ACOS, we give the ACO or ACOS identification number. In the same Table 3 we also list the ID number of the cluster counterparts within our catalog (column 2), their number of members (column 3), redshift (column 8) and velocity dispersion (column 9).
Table 3. Clusters within ESP
There are 8 clusters with redshift estimated by Zucca et al. (1993) The measured redshifts of 6 of these clusters are in good agreement with the estimated redshift: the difference between the two redshifts is of the order of 10%, less than the 20% uncertainty on the estimated redshifts. For the remaining 2 clusters, ACO 2828 and ACO 4068, the estimated redshift is significantly smaller than our measured redshift. The projection of the foreground systems ESP 175 and ESP 178 within the Abell radius of ACO 2828 could explain the inconsistency between estimated and measured redshift for this cluster. In the case of ACO 4068 we do not find any foreground/background system within ESP. ACO 4068 is very close to the northern declination boundary of the ESP strip. An inspection of the COSMOS catalog just outside the boundary of the OPTOPUS field containing ACO 4068 shows that a significant part of this cluster lies outside our redshift survey and therefore background/foreground projection could still be the cause of the inconsistency between its estimated and measured redshifts.
We also note that EDCC163 and ACOS1055 are among the most incomplete systems in our catalog. In the fields of EDCC163 ( = 3) and ACOS1055 ( = 9) the number of objects without redshift is 16 and 63 respectively. We will not consider these two clusters in what follows.
In panel a) and panel b) of Fig. 10 we plot, respectively, and as a function of cz . As expected, clusters (represented by large dots) populate the highest part of both diagrams at all redshifts. In both diagrams, mixed with clusters, there are also ESP groups that have not been identified as clusters.
The completeness of bidimensional cluster catalogs is an important issue for cosmology (van Haarlem et al. 1997) since the density of these clusters and their properties are used as constraints on cosmological models (e.g. Frenk et al. 1990, Bahcall et al. 1997, Coles et al. 1998). It is therefore interesting to determine whether there are other systems selected in redshift space that have properties similar to those of the cluster counterparts but that have escaped 2-D identification.
We limit our search for "cluster-like" groups to the velocity range 25000 cz 45000 km s-1 . Within this range the selection function is rather stable and relatively close to its maximum. In our catalog we identify the counterparts of 8 2-D clusters within this redshift range. Two of the eight clusters are of richness class R=1 (ACO2840=ESP183 and ACO2874=ESP216). The minimum number of members of these clusters is 6 and the lowest velocity dispersion of the 8 clusters is 280 km s-1 .
Apart from the counterparts of the 8 clusters, we find 11 additional ESP groups that satisfy all three conditions 25000 cz 45000 km s-1 , , and km s-1 . These groups are 10% of all groups in this redshift interval and we list them in Table 4.
Table 4. Cluster-like groups
In a - plane, Fig. 11, the eleven "cluster-like" groups occupy a "transition region" between clusters and groups. First we note that, in this plane, the two counterparts of the R=1 ACO clusters (ESP183 and ESP216) are very distant from the "cluster-like" groups. The same holds true for the only rich EDCC cluster that is not an ACO cluster, EDCC519. We conclude that no rich cluster is missing from 2-D catalogs in the region of the sky covered by the ESP survey. This conclusion is reassuring, even if it does not allow us to discuss the problem of the completeness of rich 2-D clusters in general because it is based on a small number of objects.
In the case of the more numerous poorer clusters, Fig. 11 shows that several systems could be missing from the 2-D list. The boundaries of the cluster and group regions in the - plane are blurred by the OPTOPUS mask and by the narrow width of the ESP survey. It is therefore difficult to give a precise estimate of how many "cluster-like" groups should be considered "missing" from bidimensional catalogs.
That poor 2-D clusters and "cluster-like" 3-D groups are probably the same kind of systems is confirmed by the fact that they have the same fraction of -galaxies , a higher value than it is typical of richer clusters. The 11 "cluster-like" groups have a total of 110 members, 43 of which are -galaxies : = 0.39. The 4 poor clusters that have 17 include 39 members and have = 0.41. We remember here that for all ESP counterparts of clusters we find = 0.25.
In conclusion, the comparison of ESP systems with ACO, ACOS and EDCC clusters indicates that the "low mass" end of the distribution of clusters is poorly represented in 2-D catalogs; on the other hand, the 2-D catalogs appear reasonably complete for high mass clusters.
© European Southern Observatory (ESO) 1999
Online publication: December 22, 1998