8 Surface density contrast

Next we combine information about multiplicity and environment and compute the surface density contrast of CGs ( $\Sigma_{\rm CG}/\Sigma_{\rm env}$), a parameter that quantifies the excess of surface density within the CGs as compared to that of their environment. It is defined as the surface number density of galaxies within and area of radius $r_{\rm max}$ with respect to the surface number density of galaxies within an area of radius 1 h^-1 Mpc. A space number density contrast constraint (defined with respect to the mean of the entire sample), has previously been coupled to the FoF algorithm to identify loose galaxy groups in the CFA (Ramella et al. 1989) and SRSS (Maia et al. 1989) surveys. If we were to adopt a similar criterion, CGs would correspond to even higher overdensities, because the vast majority of the systems defined on a 200 h^-1 kpc scale turn out to be single galaxies, which, as shown in Fig. 10, are associated with environments typically sparser than those of CGs. If surface density contrast is plotted against $\sigma _{v}$ one expects field systems to occupy low velocity dispersion-high density contrast regions and systems which are subclumps embedded within larger structures to occupy high velocity dispersion-low density contrast regions. In Fig. 11 the region occupied by CGs in a $\Sigma_{\rm CG}/\Sigma_{\rm env}$ vs. $\sigma _{v}$ plot is displayed. Whilst Ts are located predominantly near the field-systems area, Ms are typically associated with the embedded-systems area. Figure 11 shows that multiplicity is a rather robust parameter to discriminate between field structures and embedded structures, and indicates that, to reduce scatter in CG properties, Ts should not be included among higher multiplicity CGs, as this roughly would correspond to sampling together field-CGs and embedded-CGs.

In Fig. 11 the CGs that have been excluded from the sample because they are ACO subclumps are also plotted. ACO subclumps occupy a distinct region on the diagram. Whilst presenting a velocity dispersion similar to Ms, ACO $_{\rm CGs}$ are generally less overdense structures. On one side this might confirm that several Ms are structures that constitute the central core of large-groups/poor-clusters. This interpretation nicely matches observations indicating that, concerning X-ray properties, the distinction between compact and loose groups is not a fundamental one (Heldson & Ponman 2000). On the other hand the embedded status of many Ms could indicate that these are actually temporary chance alignments within a structure much larger than the CG (Mamon 1986; Walke & Mamon 1989; Hernquist et al. 1995). If this is the case, the characteristic $\sigma _{v}$ associated with Ms are probably too high an estimate, and all dynamically derived parameters, such as M/L or the dynamical time $H_{0}t_{\rm d}$ would strongly reflect the same bias.

To underline that the properties and differences between Ts and Ms are not to be attributed to random properties of the large scale distribution of UZC galaxies we show in Fig. 12 the position occupied by pseudo-CG samples extracted from simulated UZC catalogues (see Sect. 5) on a $\Sigma_{\rm CG}/\Sigma_{\rm env}$ vs. $\sigma _{v}$ plot. As pseudo-CG samples typically include few systems, to match the numerical dimension of the real sample we have grouped together 20 pseudo-CG samples.

Figure 12: Surface density contrast versus $\sigma _{v}$ in pseudo CG samples. To achieve numerical consistency with the real CG sample 20 pseudo-samples have been grouped together. No relation appears to exist between $\sigma _{v}$ and surface density contrast in CG samples identified in random UZC catalogs.