ecf97c2eb8d711ca.tex
1: \begin{abstract}
2:   
3:   We present results of a convergence study in which we compare the
4:   density profiles of CDM dark matter halos simulated with varying mass
5:   and force resolution.  We show that although increasing the mass and
6:   force resolution allows one to probe deeper into the inner halo
7:   regions, the halo profiles converge at scales larger than the
8:   ``effective'' spatial resolution of the simulation.  This resolution
9:   is defined both by the force softening and by the mass resolution. On
10:   radii larger than the ``effective'' spatial resolution, density
11:   profiles do not experience any systematical trends when the number of
12:   particles or the force resolution increase further. In the
13:   simulations presented in this paper, we are able to probe density
14:   profile of a relaxed isolated galaxy-size halo at scales
15:   $r=(0.005-1)\rvir$.  We find that the density distribution at
16:   resolved scales can be well approximated by the profile suggested by
17:   Moore \etal (1998): $\rho\propto x^{-1.5}(1+x^{1.5})^{-1}$, where
18:   $x=r/\rs$ and \rs~ is the characteristic radius.  The analytical
19:   profile proposed by Navarro et al.  (1996) $\rho\propto
20:   x^{-1}(1+x)^{-2}$, also provides a good fit, with the same relative
21:   errors of about 10\% for radii larger than 1\% of the virial radius.
22:   For this limit both analytical profiles fit well because for
23:   high-concentration galaxy-size halos the differences between these
24:   profiles become significant only at scales well below $0.01\rvir$.
25:   We also find that halos of similar mass may have somewhat different
26:   parameters (characteristic radius, maximum rotation velocity, etc.)
27:   and shapes of their density profiles.  We associate this scatter in
28:   properties with differences in halo merger histories and the amount
29:   of substructure present in the analyzed halos.
30: 
31: \end{abstract}