On effects of the cosmic variance in dissipationless cosmological simulations
Abstract
Two sets of numerical simulations are presented. Each set contains ten individual runs differing only in the choice of the random phases for generating the initial conditions. The two sets themselves differ only in the box size: 4 Mpc/h and 16 Mpc/h, respectively. The analysis of these twenty simulations focuses on the scatter introduced by both the various representations of the power spectrum and the box size. The "direct" variance of the internal properties such as the mass and concentration of galactic halos is then compared against the variance derived from averaging over the ten most massive halos in a single "reference" simulation of box size 64 Mpc/h. There are two primary conclusions from that analysis. First, there are no indications that the commonly used technique for setting up cosmological simulations contains serious flaws. For the two sets, each containing 10 representations of the same input power spectrum in the same cosmological volume, there were no oddities encountered that could not be explained and the fluctuations in the particle distribution actually agree with those described by the input power spectrum. Second, the scatter introduced by the cosmic variance when using small simulations boxes is comparable to the scatter found when analyzing similar objects in a single large-volume simulation. I conclude that it is safe to simulate boxes with scales (marginally) smaller than today's non-linear scale without creating results that are dominated by the cosmic variance.Keywords
All Related Versions
This publication has 0 references indexed in Scilit: