What aspects of galaxy environment matter?

From NYU CCPP Wiki

Figure 1: Color-magnitude diagram of galaxies versus host group luminosity.

Figure 2: Color-magnitude diagram of galaxies versus groupocentric distance.

There are a huge number of different ways of quantifying "galaxy environment" --- local density (either unsmoothed or smoothed, in various ways), local group environment, nearest neighbor, etc. Here we compare a simple version of group environment to the local density defined on various scales, and ask which measure of environment contains more information (not in an information-theory sense, of course). For Sloan Digital Sky Survey galaxies, we measure use the Berlind group catalog and determine the host group luminosity and the distance from the group center (hereafter, ``groupocentric distance) for each galaxy. For comparison, we also measure the surrounding density field on scales ranging from 100 kpc/h to 10 Mpc/h. All of this is described in more detail in the paper.

Figures 1 and 2 show the basic effect of group environment on the colors of galaxies. In each panel, we list the fraction of galaxies considered "blue" (below the solid line). Taken together, these plots demonstrate four primary things:

1. In larger groups a smaller fraction of galaxies are blue.
2. Even at a fixed group size, the fraction of blue galaxies depends on groupocentric distance.
3. The positions of the blue and red sequences do not change with environment.
4. The effect of environment is not limited to the largest groups, and kicks in even for the small and medium sized groups.

Of course, none of these results are particularly new, and the basic outline of the environmental dependence of galaxy properties has been long established.

Figure 6: Dependence of blue fraction on large-scale local density, at fixed group environment.

The question is, is there any aspect of environmental dependence that isn't captured by these two parameters? One simple thing to compare to is the density in cylinders around each galaxy. We have measured this for cylinders with radii varying from 100 kpc/h to 10 Mpc/h. We can then ask the following question. At a fixed group environment, we can ask whether the blue fraction is a strong function of the local cylinder density. Figure 6 shows this dependence, for the local density measured on large scales, 10 Mpc. There is very little dependence, as long as the group environment is fixed. As we show in the paper, this is true almost uniformly except on the smallest scales, of a couple hundred kpc. We also describe in the paper the effect of this bias on the correlation function, which is less than 10% or so.

The overall conclusion from the point of view of galaxy formation is that the main things that affect galaxy formation appear to be the size of the group a galaxy is in, and the position within that group. Where that group is in the larger-scale structure (in a super cluster or in a void) appears to be pretty irrelevant based on Figure 6.

Another interesting point to make based on these results, which has indeed been made before but still bears repeating, is that the environmental effects kick in for groups. One does not have to be in a large cluster to see environmental differences between galaxies. Thus, explanations of the morphology-density relation that rely primarily on cluster-specific mechanisms cannot explain all of what we see in the galaxy population.

The overall conclusion from the point of view of cosmology is that the halo occupation distribution description of galaxy cluster appears to hold up well in the observed Universe. A halo model description can explain the large-scale projected correlation function of galaxies and its variation with galaxy type at the 5% level. It has been argued that the halo model is not a good description of some predictions, such as the Millenium simulation, which is likely an indication of incorrect assumptions in those simulations.