Introduction/ High resolution Figure 1 for astro-ph/0201193/ Zoom out : 50 Mpc/h wide and 15 Mpc/h thick region centred on Figure 1/ Nearest neighbour statistics with denser reference sample/ Test galaxy and halo populations/ links/

The next plots illustrate the effect of using a denser set of reference galaxies (new sample, with relaxed selection criteria) on the nearest neighbour statistics compared to the results obtained using the astro-ph/0201193 sample (initial sample). Recall that we are computing all statistics in real space.




Click the plots to view them in full extent (corresponding .ps.tar.gz files can then be downloaded).




---





• The left series of 4 panels below correspond to a new set of reference galaxies selected by their luminosity : -20.5 < M_B < -18.5 and the presence of a bulge, of any size. We find 26457 such galaxies in the high-resolution zone : comoving density 0.012 (h/Mpc)³. Note that at a factor of three down from the top of the distribution, the indices associated to the curves decrease monotonically from left (test sample #7) to right (test sample #1), as for the initial nearest neighbour statistics. We redirect to the preprint for the description of the test samples.




• The right series of 4 panels recall the nearest neighbour statistics for the initial set of reference galaxies used in the preprint : -21.8 < M_B < -19.8 and morphological type Sa/Sb/Sc as obtained with the Simien and de Vaucouleurs (1986) criterion : 1 < M_B,bulge-M_B,total < 2.2. We found 1471 such galaxies in the high-resolution zone : comoving density 0.00069 (h/Mpc)³.





The cumulative distribution of the distances Dto from the test galaxies of the series of samples to their nearest reference galaxy (dashed lines), the cumulative distribution of the distances Doo from the new reference galaxies (selected in broad luminosity range with a bulge of any size) to its closest new reference galaxy (solid line), and the cumulative distribution of the distances Dto measured from the nodes of a regular mesh to their nearest new reference galaxy (dotted line). The sample indices associated with the curves increase from the right to the left.

Same as the left picture, but with reference galaxies selected as in the preprint.








---




We now focus on the samples associated with the two outermost galaxy property bins : samples #1 and #7, with the exception of the 6th "halo mass" sample for reasons of noise. We also plot the result of randomizing within the simulation the positions of one third of the test objects.



• The left and right pictures correspond to the new and old(initial) set of reference galaxies respectively.



• Let us compare nearest neighbour distances at a factor of three down from the top of the distribution in both samples. As expected, using more reference galaxies substantially reduces D_oo : at this population fraction :
  
  
  • D_oo,new=2.5 Mpc/h
  
  
  
  • D_oo,old=8.8 Mpc/h



• Distances from test samples to the nearest reference galaxy decrease as well, but the ratios D_to/D_oo increase. We quote D_to,new for the first sample of each set:
  
  
  • D_{to,new,halo mass 1}= 3.1 instead of D_{to,old,halo mass 1}= 9.2 Mpc/h, and a 18% increase in the ratio D_to/D_oo
  
  
  
  • D_{to,new,luminosity 1}= 2.8 instead of D_{to,old,luminosity 1}= 9 Mpc/h, and a 10% increase in the ratio D_to/D_oo
  
  
  
  • D_{to,new,colour 1}= 3.3 instead of D_{to,old,colour 1}= 9.45 Mpc/h, and a 23% increase in the ratio D_to/D_oo



• Because our algorithm does not allow D_to=0, test samples overlapping in luminosity with the new reference sample will be biased toward higher distances. However, there is no overlap in luminosity between the new reference galaxies and the galaxies in our first, faintest luminosity test sample, where we will therefore gauge the effect of denser sampling (some 20 times denser).

  In this case, the measured 10% increase in the ratio D_to/D_oo proves that the effect is quite limited and does not change our conclusions that GL₁ does not constitute a homogeneous population with respect to the new reference objects. We stress that the new reference sample has 20 times more objects.





• Note the sharp decrease in the plots for D_to in the right hand side windows of the plots with the new reference population. This is because a large faction of the test objects in the samples HM₆, GL₇ and GC₇ are galaxies in massive haloes. They find their nearest new reference galaxies as satellites or central objects within the same halo . The same trend was already apparent in the old sample of reference objects (compare to the corresponding plots in the Figure on the right), but was much weaker. Recall that HM₆, GL₇ and GC₇ contain the most massive haloes, the brightest and the reddest galaxies respectively.





The dashed line gives Dto for the extreme bins of each sample (1 and 7, with the exception of HM6). The dash-dotted lines show the corresponding Dto after randomizing the positions of one third of the objects in each test sample, and the solid curve gives Doo for the new set of reference galaxies. The dotted line is the same as in the Figures above.

Same as for the left picture, but with the initial set.

Download the (galaxy and halo) void test populations.