| |
 About Us
Search
Contact Info |
 University
of Washington |
Department
of Astronomy |
Computer
Science and Engineering |
Department
of Physics |
|
|
|  |
Directions on how tobuild a catalog of subhalos |
|||||||||||||||||
Example of SKIDcommand
lines for a cluster halo in an Omega=0.3,
Lambda=0.7 universe:
The linking length
tau is the most crucial parameter. For this run the softening
length is eps=2.5e-05; according
to my experience tau should be
set to at least 3 times the softening eps.
In order to make sure that the subhalo properties are independent of
SKID's linking length it is best to run SKID for 3 values of tau
-
for this case choose
tau=10
and
15e-05;
tau=10 is the most balanced choice (adequately identifies both tiny
and biggy subalos -- I consider this the fiducial SKID run) .
[Considerations based on the analyses I have
carried out for the "Virgo" and "SuperVirgo" cluster runs]
For instance the command line for a SKID run with
minimal tau is:
Once you have the coordinates of the center of the halo, you start building
the catalog of halos with distances relative to the center (you can also
find
the index of the central halo as found by SKID running find_neighbour.f
for the given coordinates. This straightforward code reads the
subhalo positions from a default file skid.stat,
so remember to copy your .stat
file to skid.stat.
Also, once you have the index of a halo, you can easily look at its
data in the skid.stat file, using
popup.f.
The program reads some standard info about the simulation from a .sim
file. This information is then used to convert the data from PKDGRAV code
units
to physical units. Here is an example of a .sim
file for a LCDM simulation in a box of 100 Mpc a side:
| This program can also be fed the orbital parameters of
the subhalos, but they have to be computed separately (and before this stage) using orbits.f
(see
Step 4 below).
The output of make_halo_table.f is a table file listing all the relevant properties of the subhalos (within the selected radial distance). Units are physical kpc, km/s and M_sun (units of (km/s)^2 for energy per unit mass Ene and kpc*(km/s) for angular momentum angM). The 'circularity' eps is the ratio between the angular momentum of the orbit and that of a circular orbit with the same energy. [If you do not provide orbital data the columns for peri,apo, eps, etc.. will show zeros] Here is an example of a table of subhalos for a LCDM run: #Generated from: cl1c6.432.00512.0.stat
Step 4How to compute the orbital parameters.... The orbits are computed in the spherically simmetric static potential approximation.Use code orbits.f. Here is an example of input file for a Virgo-sized cluster run at redshift zero: orbits.inp The input
files are a SKID .stat file
for the subhalo properties and a tipsy profile
of
the (main) halo. The latter file must give the density profile over a wide
radial range
The outputs consist of two
files. One tabulates the spline fitted potential profile (but
it is scarcely of interest). The main
output is the file listing the
1 SKID group index
An example output for a LCDM virgo-sized cluster: cl1c6.432.00512.0.orb As said at point 3 above, the .orb file can be input into make_halo_table.f so that the final halo table contains also the orbital parameters. Don't be put off (too easily) by error messages from orbit.f. They warn you that some of the orbital parameters could not be computed (typically the orbital period for very radial orbits). All these pathological cases have negative or 0 values in the output table. Step 5What to do with the subhalo data in the .table files? ... Things to look at are the mass function or the the circular velocity function of the subhalos (the latter is the distribution of the peak values of the circular velocity profiles for the subhalos, as computed by SKID). Here is an SM macro to plot the mass function dMDF.normalized.sm ("normalized" in the sense that masses and volumes are measured in units of the virial mass and the virial volume of the parent halo). Here is another for the circular velocity function: dVDF.normalized.sm (normalized to the circular velocity of the parent halo at its virial radius).Typically, in order to make the plots, you will use the fiducial subhalo .table for the fiducial SKID run. According to my experience this is enough (I have also tried fancier solutions combining tables for different SKID runs but the results were essentially unaffected). Here is an SM macro that I used to compare different subhalo catalogs for different SCDM cluster runs: SCDM.checks.sm; here is another for LCDM cluster runs: LCDM.checks.sm; and one more for various galaxy runs: Galaxies.checks.sm. The SM macros listed above may need this collection of additional macros my_macros.sm, which could be useful for other occasions too. Just in case, here is a list of other SM macros to draw plots for halo and subhalo properties:
|
|
|
N-Body Shop
University of Washington Box 351580 Seattle, WA 98195-1580 (206) 543-9849 voice, (206) 685-0403 FAX [comments to ghigna@astro.washington.edu] |
