4 The candidate cluster sample

By applying the cluster finding method described in the previous section, we obtained a sample of 171 candidate clusters associated to NVSS radio sources identified with $d_{\rm r-o} \le 7\hbox{$^{\prime\prime}$ }$ with EDSGC galaxies of magnitude $17.5~\le~b_{\rm J}~\le~20.0$. Among these candidates, 123 are associated to NVSS pointlike sources, 23 to NVSS "close'' double sources, and 25 to NVSS "wide'' doubles. The sample covers an area of $\approx$550 sq degrees at the South Galactic Pole and the uncertainty on the candidate cluster position is $30\hbox{$^{\prime\prime}$ }$.

As a further step in the compilation of the cluster sample we looked for candidates in common with other cluster catalogues, considering a candidate as already "known'' if the radiogalaxy and the optical centroid are found inside an Abell radius from the cluster centre.

Out of 171 candidates, 76 were found to be associated with a known cluster according to the above definition. In six cases two candidates are associated with the same known cluster. We detected 2 Abell poor clusters with measured redshift, 40 out of the 128 ACO/Abell (Abell et al. 1989) rich clusters present in this sky region (31 of them being listed in the EDCC as well), 16 EDCC (Lumsden et al. 1992) clusters, 9 clusters from the APM catalogue (Dalton et al. 1994, 1997), 2 groups selected from the ESO Slice Project survey (Ramella et al. 1999). Finally, in one case the candidate corresponds to a cluster identified with an X-ray source in the Einstein Medium Sensitivity Survey (Stocke et al. 1991).

To evaluate $R_{\rm A}$, we used the measured cluster redshift when available; otherwise, we used the estimated redshift or, in the case of EDCC clusters, values of $R_{\rm A}$ estimated by the catalogue authors. In thirty-six cases the association with previously known clusters has been made on the basis of a measured redshift in the literature, while for the other 40 only an estimated redshift is available.

The detection of these known clusters can be interpreted as a further indication that this radio-optical selection method is powerful in the search of cluster candidates.

Among the 76 known clusters, 57 host a pointlike NVSS radio source, while 11 and 8 are respectively associated to "close'' and "wide'' double radio sources.

In Fig. 2 the sky distribution of the known candidates (empty symbols marked with the cluster name) and of those candidates without a counterpart in the literature (filled symbols) is shown.

The use of this bivariate radio-optical selection method, based on the condition that an optical excess is considered as a reliable candidate cluster only if it is associated with a radiogalaxy, makes it possible to detect cluster candidates whose reliability, in terms of their optical surface density alone, would be in many cases too low to be included in catalogs based on pure optical selection methods.

The total contamination present in the cluster sample, due to the probability of chance coincidence between a radiogalaxy and an excess in the optical surface galaxy density, has been estimated as follows: we repeated the search of cluster candidates coupling the smoothed matrix relative to each UKST plate with the radiogalaxies belonging to another plate. By applying the same cluster finding criteria described above, we found a contamination percentage of $28\%$ in our sample of candidate clusters at intermediate redshifts. As the actual radiogalaxy sample instead of a random-generated one has been used, this is an estimate of the effective contamination, that is also the contamination term due to the presence of spurious radio-optical associations among the radiogalaxies is taken into account.

An assessment of the reliability of the method is provided by the association with known clusters and measurements on new candidates. Therefore, in the next two Sects. we briefly report on those candidates hosting a radiogalaxy whose redshift is known from previous surveys. This search has been made using the NASA Extragalactic Database. In Sects. 5 and 6 the results obtained from a spectroscopic observative run for a first set of new cluster candidates are presented. The whole candidate cluster sample will be presented in a following paper.

  
4.1 Previously known clusters

The naming convention for cluster candidates in our sample is as follows: first digits identify the number of the UKST plate on which the candidate has been found; letters are used to distinguish among the various radio morphologies (pointlike or double sources) and to identify candidates associated to more than one radiogalaxy. Last digits in the name are the sequential number of the radio source on that UKST plate.

295BN07: this candidate is found to correspond to A2860, whose measured redshifts is z=0.105800 (Struble & Rood 1999). The radiogalaxy lies at $z = 0.10757 \pm 0.00018$ (Vettolani et al. 1998) so that 295BN07 is considered actually coincident with a known cluster.

295D24: the candidate is found inside one Abell radius from the ESP Group 193 (Ramella et al. 1999). The radiogalaxy has been detected at $z = 0.05547 \pm 0.00007$ in the ESP survey (Vettolani et al. 1998). ESP group 193 lies at redshift $z=0.05444 \pm 0.001$ so that the association is considered real.

352N25: the candidate has been found to be associated with the ACO cluster A2871. In A2871 the presence of 2 galaxy systems at different redshifts has been detected (Katgert et al. 1996) respectively at <z> = 0.114 (14 galaxies) and <z> = 0.122 (18 galaxies). The radiogalaxy has measured redshift from the ENACS survey z = 0.11415$\pm$0.0003 (Katgert et al. 1998) so 352N25 seems to be associated with one of these two substructures.

411N35: the radiogalaxy in this candidate lies inside an Abell radius from the center of the EDCC cluster E482, whose measured redshift is z = 0.108040. The radiogalaxy has $z = 0.07673 \pm 0.0003$ (Collins et al. 1995); 411N35 is thus not actually coincident with E482. However, measured velocities in the E482 field include a set of galaxies with $cz \sim 22\,700~{\rm km~s}^{\rm -1}$, indicating the possible presence of a superimposed structure at about the same redshift as the radio source.

  
4.2 New cluster candidates

From the publicly available data of the Las Campanas Redshift Survey (Shectman et al. 1996) we obtained the redshift of two radiogalaxies associated with the new candidate clusters 297BN04 ( $cz = 53\,460 \pm 58~{\rm km~s^{-1}}$) and 293D22 ( $cz = 41\,695 \pm 95~{\rm km~s^{-1}}$). At $\sim$ $2\hbox{$^\prime$ }$ from the radiogalaxy in 297BN04 we found a galaxy with measured velocity $cz = 53\,319~{\rm km~s^{-1}}$. Finally, a further redshift has been found in the ESP survey for the radiogalaxy associated to the candidate 294N04: $cz = 43\,097~{\rm km~s^{-1}}$ (Vettolani et al. 1998). No redshift data for other galaxies near these radiogalaxies are available, so these velocity measurements cannot be used to confirm or not the presence of a cluster.