This page contains data concerning a galaxy cluster simulation that was run the week of Sept. 23, 2004. This simulation was run using the enzo code, which is an Eulerian adaptive mesh cosmology code which combines an adaptive particle-mesh n-body solver with the piecewise parabolic method (PPM) hydrodynamic scheme. Significant other physics is included in the code but is not relevant here.

The simulation was initialized as follows: In order to find a galaxy cluster of appropriate size we initialized a simulation using only dark matter (no baryons) in a volume of space 256 Mpc/h on a side at z=30 (when fluctuations were still linear on scales relevant to this problem) using a 256^3 root grid (1 Mpc/h spatial resolution) with 256^3 dark matter particles. Each of these particles has a mass of 1.19e11 solar masses. Standard cosmological parameters are used: Omega_total = 1, omega_matter = 0.3, Omega_lambda = 0.7, Hubble parameter = 70 km/s/Mpc, sigma_8 = 0.9 with a power spectrum with index n=1. Adaptive mesh refinement (AMR) was turned on with an additional 4 levels of refinement (doubling the root grid resolution at each level) and the simulation was run to z=0. At this point, we stop the simulation an use the Hop halo-finding algorithm (Eisenstein and Hut, 1998) to find the most massive halo in the simulation. This halo has a mass of 1.4e15 solar masses.

The simulation was then reinitialized at z=30 with both dark matter and baryons (Omega_baryon = 0.04) using the same random seed (producing the same volume of space) with a 128^3 root grid and 2 nested static grids in the lagrangian region where the cluster forms, which gives an effective root grid resolution of 512^3 cells (0.5 Mpc/h) and dark matter particles of mass 1.49e10 solar masses. Adaptive mesh refinement is allowed in the region where the galaxy cluster forms, with a total of 7 levels of refinement beyond the root grid, for a maximum spatial resolution of 15.625 Kpc/h. Note that while the baryons are resolved at higher and higher spatial and mass resolution at higher levels, the dark matter particles maintain constant mass so as not to add any additional noise. The volume was also shifted so that the cluster would form at the center of the spatial volume.

This simulation was then evolved to z=0.06 (not to 0.0 due to computer issues) following the evolution of the dark matter and using adiabatic (ie, non radiating) gas dynamics. The galaxy cluster that we were simulating at high resolution was relocated with the HOP algorithm and data cubes, radial profiles and images were created, as described below.

Data Cubes

4 sets of files have been extracted at redshifts z=1, 0.5, 0.25 and 0.06. Each set of files contains a dataset for baryon density, temperatature and x, y and z-velocity at two different resolutions: one being a box that is 4 Mpc/h on a side with 15.625 Kpc/h resolution (the "l7" dataset), and the other box being 8 Mpc/h on a side with 31.25 Kpc/h resolution. Both datasets are cubes with 256^3 cells and the data is ordered in the fortran (column major) style. The datasets are in the HDF 5 data format and the code is in the enzo AMR code's internal code units. An example C++ code is provided here which will read the datasets and convert them into proper cgs coordinates. In order to get this code to compile you will have to download and install the HDF 5 library, version 1.4.5-post2, which can be obtained via ftp at ftp.ncsa.uiuc.edu/HDF/HDF5/prev-releases/hdf5-1.4.5-post2.tar.gz. There is documentation in the HDF5 library which explains how to install it, and the sample reader has a makefile with some instructions on how to link the HDF 5 library. NOTE: This is not the most current version of the HDF 5 library. We have experienced compatibility issues with later (1.6.x) versions of the libraries and are using the 1.4.5-post2 version until these issues are resolved.

All cubes at z=0.06 - 670 MB
All cubes at z=0.25 - 670 MB
All cubes at z=0.5 - 670 MB
All cubes at z=1.0 - 670 MB

Data cubes in gzipped text format

The cubes are arranged in columns, in this order: Cell number, Dark matter density, Baryon density, baryon temperature, baryon x-velocity, y-velocity, z-velocity. All data is in proper CGS units: density is in g/cm^3, temperature in Kelvin, velocity components are in cm/s. Note that the output is in column-major format (see the source of the hdf reader for an explanation of this). Cubes are here:

z=0.06, L6 - 66 MB
z=0.06, L7 - 60 MB
z=0.25, L6 - 64 MB
z=0.25, L7 - 60 MB
z=0.50, L6 - 63 MB
z=0.50, L7 - 60 MB
z=1.00, L6 - 49 MB
z=1.00, L7 - 45 MB

All files are gzipped. Warning: all expand to roughly 1.3 GB apiece when you unzip them!

Images

The links below lead to pages containing images of the dark matter density, baryon density and baryon temperature for the same x-ray cluster at different redshifts: z=1, 0.5, 0.25 and 0.06. On each page the cluster that data is available for is imaged at various resolutions, as described in the picture c aptions.

z=1 images
z=0.5 images
z=0.25 images
z=0.06 images

Radial profiles

A gzipped tar file which contains text files with spherically-averaged radial profiles of various quantities in the simulations can be found here. This file also contains the parameter files from the simulation at the redshifts in question. These parameter files can be interpreted using the enzo user's manual and the enzo parameter list, though for the redshifts provided the example code will convert everything to proper CGS units.

Last modified 19 October 2004
by Brian O'Shea