Astron. Astrophys. 355, 848-862 (2000)

5. Discussion and conclusions

From the analysis of our data obtained through multi-object spectroscopy, we have found that Abell 521 is a moderately distant cluster () with an apparently very large velocity dispersion ( km s^-1). The velocity distribution of cluster members is consistent with sampling from a parent Gaussian population, and the high dispersion does not seem to result from trivial superposition effects. If it was real, this velocity dispersion would be the signature of an extremly massive cluster. However, our analysis suggests otherwise, and indicates that Abell 521 is a highly unrelaxed system. In fact, there do exist several hints that the large apparent velocity dispersion obtained may be due to a superposition of several distinct populations of galaxies. The projected distribution of the galaxy positions in the cluster can be described with a mixture of three two-dimensional Gaussians, a model which is significant at the 99% level. Analysis of the velocity distribution for these partitions shows that the velocity dispersion is very high ( km s^-1) in the central North-East/ South-West "ridge" corresponding to KMM2 North, although the dispersion for the smaller samples of galaxies associated with the KMM1 and KMM2 South partitions are more typical of most rich clusters ( km s^-1). We also find that the velocity distribution is rather different for subsets of the galaxies selected according to color, with the bluest objects (spirals) showing a high velocity dispersion ( km s^-1) and the redder objects (ellipticals) exhibiting a much smaller dispersion ( km s^-1). This is another indication that this cluster is dynamically young, with its population of spirals not yet relaxed to the cluster potential. Moreover, the high value of the velocity dispersion as compared to the one expected from the X-Ray Temperature and the relation suggests that this cluster is far from dynamical equilibrium.

On the basis of our results, we can outline a tentative picture of the dynamical state of Abell 521. In the frame of hierarchical models, the cluster formation process proceeds by merging of smaller units (Frenk et al. 1996). Several pieces of evidence that merging processes are occurring in this cluster have been obtained. First, one can note that the various clumps evidenced on the NW/SE axis (KMM1, KMM2 South, KMM3) are well-defined concentrations. The foreground/background hypothesis is quite improbable, as redshift measurements have shown that all the groups contain several cluster members (although in the case of KMM3 only two redshift measurements exist). We are then probably seeing the early phase of infall of these various groups. According to numerical simulations (Schindler & Bohringer 1993), we should expect that groups are strongly accelerated and their velocity distributions diverge as the merger process proceeds. However, we do not detect any offset of the central velocity locations of the various groups (except possibly KMM3, which could be at a slightly higher velocity of km s^-1 based on the two measured velocities). This implies that either we are seeing the very beginning of the merging and the clumps are still nearly at rest with respect to one another, or we are witnessing a more advanced state but fail to detect the shift of the velocity distributions because the collision axis happens to be mostly in the plane of the sky.

The apparently large velocity dispersion of Abell 521 is due in great part to the contribution of the KMM2 North structure. Its very high velocity dispersion suggests that we are witnessing the collision epoch, at which point numerical simulations show that the dispersion reaches its maximum value, which can be twice the value after the cluster approaches dynamical equilibrium (Schindler & Bohringer 1993). The present data could be explained in a scenario whereby two (or more) subclusters have just collided along an axis which is projected on the sky in the direction of KMM2 North, but with a substantial component along the line of sight. A detailed merging scenario taking into account the whole set of optical and X-Ray properties of this cluster is presented in Arnaud et al. 2000.

The Brightest Cluster Galaxy of Abell 521 is particularly interesting. Its large luminosity and effective radius are characteristic of a cD galaxy, but no extended halo is detected down to our limiting magnitude of 27 mag arcsec^-2. The BCG exhibits multiple nuclei, and a particularly puzzling configuration of several knots superposed on a diffuse arc-like structure. Although we failed to measure the redshift of the diffuse structure, the spectra and colors of the knots indicates (at least for two of them) that they belong to the cluster. The diffuse curvilinear structure would then probably result from previous collisions and further interaction between the knots. Several other signs of interactions are present within the BCG- twisting of the isophotes, and the presence of a clearly interacting nucleus (A1). From our dynamical analysis, the BCG and its various nuclei appear to be a bound system. This is probably a young system of small spatial extensions, belonging to KMM1, in which the merging rate is very fast but the halo of the cD has not had yet time to form. The BCG's peculiar velocity of 194 km s^-1 with respect to the main cluster is judged not to be statistically significant.

From this analysis, Abell 521 seems to be a galaxy cluster currently forming at moderate redshift. If this kind of object is common, it would represent a severe constraint for low value in the hierarchical model of galaxy formation (Richstone et al. 1992; Kauffmann & White 1993). Additional multi-object spectroscopy in order to obtain a complete set of velocity information for galaxies in Abell 521 down to magnitude is planned. This will allow a better determination of the complex dynamical state of this cluster.

© European Southern Observatory (ESO) 2000

Online publication: March 21, 2000
helpdesk.link@springer.de