Practical 2:

In this Practical you will use KSBf90 to analyse CFHTLS Deep data and produce a dark matter mass map!  It follows the same initial steps (1-6) as Practical 1.  Don’t forget to edit the fitsname in your scripts to the D1 data name D1_i_V1.6A_Paris.cut.fits!

1. 1)Download and unpack KSBf90.tar  if you haven’t done this already
   

This contains a Makefile and pre-compiled 32-bit Linux executables.  At the IAP only some of the computers have the cfitsio routines installed so the Makefile will not compile on all machines.  However you should be able to use the pre-compiled executables.

2. 2)Set your directories as follows (edit)
   

setenv dataDIR /Users/heymans/Paris_DUEL/images     # Where your D1 images are

setenv CATDIR /Users/heymans/Paris_DUEL/catalogues  # Where your D1 catalogues are

setenv KSBDIR /Users/heymans/Paris_DUEL/KSBf90      # Where your KSBf90 software is

3. 3)Run SExtractor on your CFHTLS D1 image. 
   

Put the resulting SExtractor ascii catalogue in your $CATDIR.  Note this time you will need to run SExtractor with a weight image.  Include the mask with the LDAC scripts and then output a masked ascii file to work with.

4) Download and unpack scripts.tar if you haven’t already

This directory has two example cshell scripts that you will need to edit to include the names of the data your wish to analyse.  It also contains your KSB parameter file KSBf90.param.  You also need to edit this file to update it with the ascii SExtractor catalogue format (see  KSBf90.param website).

5) Run runstars.scr

This script runs findstars.f90 and psffit.f90.   Remember your training with the SkyMaker simulations.  You need to be conservative enough with your star selection to reduce the noise.  Which polynomial order best fits the varying PSF pattern that you see?

6) Run rungals.scr

This script runs gal_correct.f90 and gal_select.f90.    For the mass reconstruction we’ll be using the catalogue output by gal_correct.f90, but look at the plots gal_select produces and write down which selection criteria you are going to use for your mass reconstruction.  This time we want to see <g> = 0 within the errors.

7. 7)Use imcat to make your mass reconstruction
   

8. 
   

9. Download mass_recon_scripts.tar and unpack.
   

10. Read the mass recon webpage to learn about the Kaiser & Squires 1993 imcat software we’ll be using.  You’ll need to set the following imcat links
    

11. 
    

12. setenv IMCATDIR /tmp_mnt/netpapeur/users_home2/visi09/software/imcat
    

13. setenv PATH ${PATH}:/tmp_mnt/netpapeur/users_home2/visi09/software/imcat/bin/linux
    

14. setenv IMCATCONVERTNANS
    

15. setenv IMCATSWAPFITSBYTES
    

16. 
    

17. Edit the runmass.scr in your mass_recon directory to include your selection criteria from gal_select with the lc -i option
    

18. Run the script to produce E and B mode mass maps. 
    

19. The script calls add_wcs_header.a which will add a world co-ordinate system so you can blink you mass map with your CFHTLS image.  Can you see any galaxy clusters at the peaks? Try changing the smoothing scale in imcats smooth.
    

20. Are you Bmodes significant? Try changing your selection crtieria, or your PSF model and star selection and re-run.  Does it improve anything?
    

21. Use lc -i option to add a redshift selection to your selection criteria to select galaxies behind the cluster
    

22. Research task:  how does your result compare to  Gavazzi & Soucail 2007
    

23. Feeling game?  Try analysing another section of the CFHTLS D1 field (copy from /tmp_mnt/netpapeur/users_home4/visi08/images) where Gavazzi & Soucail 2007 find a massive z=0.5 cluster.

Page maintained by Catherine Heymans, IfA, University of Edinburgh