Astron. Astrophys. 343, L29-L34 (1999)

3. Results

Table 2 gives the original cluster candidates coordinates, the new flux-weighted positions as described above, the significance of the detection, the Abell richness and the estimated redshift from the EIS catalog, as listed in Olsen et al. (1998b) and the new redshifts derived below using the CM-diagrams from the combined VLT-TC and SOFI data. All coordinates are in J2000. Note that the estimate of the Abell richness for distant clusters is quite uncertain, but it serves to indicate their relative richness.

Table 2. Cluster properties

3.1. EIS 0046-2930

In the EIS candidate clusters catalog this object was identified only in I-band, and assigned a redshift of . However, visual inspection of the original survey images of this field showed the presence of foreground "blue" galaxies and of a fainter red population, not detected in the band. Using the deeper optical (reaching ) and the IR catalogs produced from the VLT-TC and from the SOFI images, we can study in greater detail this cluster candidate. The resulting four optical and IR CM-diagrams are shown in Fig. 1, for all galaxies within the VLT-TC field of view. The upper panel shows the optical CM-diagram, where there is a suggestion for a concentration of galaxies at , just beyond the reach of the EIS color data. However, the scatter is large compared with that seen in clusters at intermediate redshifts (Olsen et al. 1998b), and cannot be explained by photometric errors in the color which at are 0.3 mag. This scatter prevents a secure identification of the red sequence. By contrast, the diagram shows a clear early-type sequence in the interval at . Using the above magnitude range, the CM-relation is well-fitted by a linear relation with an estimated scatter of (), comparable to the estimated error in the color and in agreement with the color dispersion of morphologically classified early-type galaxies in high-z clusters (Stanford, Eisenhardt & Dickinson 1998). The infrared diagram shows an even tighter sequence at , with a scatter of 0.1 mag, again comparable with the estimated error in the color. The ten brightest galaxies (in the Ks band) for which and are represented by filled circles in the CM-diagrams and are also numbered in Fig. 2 according to their magnitude ranking. These objects are the most likely early-type galaxy members of EIS 0046-2930. The flux-weighted position of the "cluster" is also shown.

Fig. 1a-d. Optical and infrared CM-diagrams for all galaxies in the VLT-TC field of EIS 0046-2930. The filled circles indicate the ten brightest, most likely cluster members based on a color selection as described in the text. These galaxies are also marked in Fig. 2.

Fig. 2. Combined VLT-TC images (see text) for EIS 0046-2930 (, left panel) and EIS 0046-2951 (, right panel). The field of view of these images is 90 arcsec. The computed flux-weighted centers are marked by . The numbers indicate the magnitude ranking of the ten brightest, most likely cluster members as described in the text.

The projected radial distribution of objects brighter than and within the color range , is shown in Fig. 3, in annuli 0.3 arcmin wide. The contrast of this bright red-sequence population relative to the background is clearly seen, while there appears to be no appreciable clustering for galaxies outside this color range. Even though the statistic is poor, the scale and amplitude of the overdensity associated to this population, a factor of 7 within the innermost 0.3 arcmin, are similar to those observed by Dickinson (1996) for the cluster surrounding 3C 324 at . To test the robustness of this results flanking fields with the same size of the VLT-TC field of view were extracted from the same SOFI image and used to obtain CM-diagrams and radial density profiles. None of these fields showed the presence of a concentration of galaxies both in color and in position. This suggests that the concentration of galaxies in both color and projected separation seen in the field of EIS 0046-2930 is significant and that this object is likely to be a real cluster. Further support to this conclusion comes from the matched-filter algorithm which applied to the band data detects a "cluster" at the level and at .

Fig. 3. Projected radial distribution of galaxies brighter than within the SOFI field of EIS 0046-2930. The figure shows separately the distribution of galaxies within (solid line) and outside this color range (dashed line), both normalized to their respective backgrounds.

On the presumption that EIS 0046-2930 is a real cluster, the color of the red sequence can be used to estimate its redshift. This can be achieved either by using synthetic stellar population models, or purely empirically using the colors of the red sequence of clusters of known redshift. Even though the available data are sparse, we have adopted the latter approach because it is model independent. We have used the spectroscopic redshifts and the CM-diagrams given by Stanford, Eisenhardt & Dickinson (1998) for their clusters at and the cluster of Stanford et al. (1997) to estimate the location of the early-type galaxies sequence in different passbands for clusters at . Interpolating these relations to the colors of the red sequence of EIS 0046-2930 (, , and ) we consistently estimate its redshift to be (statistical uncertainty only).

3.2. EIS 0046-2951

In the EIS catalog this object was estimated to have a redshift of , being detected only in the I-band (Table 2). However, visual inspection of the and band EIS images suggested that this system could be an overlap of two concentrations at different redshifts. Using the deeper band image obtained with the VLT-TC we are now able to investigate the optical CM-diagram shown in Fig. 4. Indeed, we find two concentrations of galaxies: one seen at and another at . These colors correspond to redshifts and , respectively. However, in the and CM-diagrams only one sequence is seen, located at and . These values lead to redshift estimates of in both cases, in good agreement with the original estimate based on the matched-filter algorithm. In contrast to the previous cluster, the scatter of the red sequence in both colors is significantly larger (0.21 in and 0.19 in ) and cannot be fully accounted for by the photometric errors in our data ( 0.15 mag). The larger scatter may be due to a larger fraction of spiral galaxies in the "cluster", or to a stronger contamination by foreground galaxies. As in the previous case, the most likely early-type cluster galaxies have been selected adopting a color-selection criterion similar to that described above. These galaxies, chosen to have and , are identified in Fig. 4 and in the right panel of Fig. 2.

Fig. 4a-d. Optical and infrared CM-diagrams for galaxies in the VLT-TC field of EIS 0046-2951. The filled circles represent the ten brightest, most likely cluster members (marked in Fig. 2) as described in the text.

Fig. 5 shows the projected radial distribution of color-selected candidate cluster members. In this case we find that the overdensity of the red sequence galaxies is 5, over the same radial distance as for the previous cluster. The smaller overdensity of this candidate cluster (and perhaps the larger fraction of spirals) is consistent with the lower original estimate of its richness (Table 2). Note that a 3 detection at approximately the same redshift was obtained applying the matched-filter algorithm to the Ks data. As for the previous object, the analysis of flanking fields from the SOFI image gives further support to the reality of the observed concentration in color and projected separation, suggesting the existence of a physical association.

Fig. 5. Projected radial distribution of galaxies brighter than within the SOFI field of EIS 0046-2951. The figure shows separately the distribution of galaxies within (solid line) and outside this color range (dashed line), both normalized to their respective backgrounds.

© European Southern Observatory (ESO) 1999

Online publication: March 1, 1999
helpdesk.link@springer.de