Skip to content

Respository for all things regarding the MICE2 mocks

License

Notifications You must be signed in to change notification settings

KiDS-WL/MICE2_mocks

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

MICE2_mocks

This Repository provides code to create galaxy mock catalogues based on MICE galaxy catalogues.

Requirements

The pipeline is written in Python3 and requires the following non-standard packages:

  • numpy and scipy
  • astropy>=3.0 (recommended for the improved astropy.table performance)
  • matplotlib>=2.0 for the plotting scripts

Additionally, the wrapper scripts in ./KV450 and ./DES make use of an external packages that provide convenience functions to handle data tables:

  • jlvdb/table_tools (script calls starting with data_table_) The path to table_tools must be included in $PATH and $PYTHONPATH.

To be able to compute photometric redshifts BPZ is requried.

Instructions

Starting from the MICE2 base catalogues, the pipeline allows to model various observational selection functions:

  • Spectroscopic surveys: GAMA, SDSS (main sample, BOSS and QSOs), 2dFLenS, WiggleZ, DEEP2, zCOSMOS and VVDS (2h field)
  • Photometric surveys: Examples to create 450 sqdeg of KiDS-VIKING (KV450, ./KV450) and the Dark Energy Survey (DES, ./DES) are included.

The pipeline allows to attach realistic photometry realisations to the MICE2 catalogues, photometric redshifts, galaxy weights, and spectroscopic success rates for some of the included spectroscopic selection functions.

Data Access

The MICE2 base catalogues can be downloaded from COSMO HUB. Recommended column selections (in SQL query stile using 'expert mode') for KV450 are

SELECT
    `unique_gal_id`, `ra_gal`, `dec_gal`, `z_cgal`, `z_cgal_v`,
    `sdss_u_true`, `lephare_b_true`, `sdss_g_true`,
    `lephare_v_true`, `sdss_r_true`, `lephare_rc_true`,
    `sdss_i_true`, `lephare_ic_true`, `sdss_z_true`,
    `des_asahi_full_y_true`, `vhs_j_true`,
    `vhs_h_true`, `vhs_ks_true`,
    `bulge_fraction`, `bulge_length`, `bulge_axis_ratio`,
    `disk_length`, `disk_axis_ratio`
FROM micecatv2_0_view WHERE `dec_gal` <= 30 AND `ra_gal` >= 30 AND `ra_gal` <= 60

where des_asahi_full_y_true is the best match for the VISTA (vhs_*_true) Y-band that is missing in MICE2. The VST ugriz-bands are covered by sdss_*_true.

For DES we can select

SELECT
    `unique_gal_id`, `ra_gal`, `dec_gal`, `z_cgal`, `z_cgal_v`,
    `sdss_u_true`, `lephare_b_true`, `sdss_g_true`,
    `lephare_v_true`, `sdss_r_true`, `lephare_rc_true`,
    `sdss_i_true`, `lephare_ic_true`, `sdss_z_true`,
    `des_asahi_full_y_true`, `vhs_j_true`,
    `vhs_h_true`, `vhs_ks_true`,
    `bulge_fraction`, `bulge_length`, `bulge_axis_ratio`,
    `disk_length`, `disk_axis_ratio`
FROM micecatv2_0_view WHERE `dec_gal` <= 30 AND `ra_gal` >= 30 AND `ra_gal` <= 60

correspondingly and make use of the fact that MICE2 comes by default with all DES model magnitudes.

This selection uses the most complete patch of MICE (30 <= RA <= 60 and 0 <= DEC <= 30). Some of these columns are only needed to additionaly select spectoscopic samples.

Creating Photometric Catalogues

The wrapper scripts in ./KV450 and ./DES show exemplary how to create mock catalogues matched to observational data. These steps include:

  1. Applying the MICE2 evolution correction.
  2. Correcting the model magnitudes for magnification.
  3. Computing observational galaxy sizes based on the point spread function (this allows a size contribution to the photometric uncertainties).
  4. Adding a photometry realization based on the observational limiting magnitudes
  5. Assigning galaxy weights by nearest neighbour matching between mock and data in magnitude space
  6. Computing photometric redshifts with BPZ

Creating Spectroscopic Catalogues

The pipeline bundles a variety of spectroscopic (target) selection functions:

  • 2dFLenS (Blake et al. 2016)
  • DEEP2 (Newman et al. 2013)
  • GAMA (Driver et al. 2011)
  • SDSS
    • main sample (Strauss et al. 2002)
    • BOSS (Dawson et al. 2013)
    • QSO sample (Schneider et al. 2010a, only attempting to match the redshift distribution)
  • WiggleZ (Drinkwater et al. 2010, missing UV information replaced by redshift distribution matching)
  • VVDS (LeFèvre et al. 2005, only 2h field)
  • zCOSMOS (Lilly et al. 2009, only bright sample)

These selection functions are defined in ./pipeline/specz_selection.py and have some adjustments applied in order to give a better match to the data colour and/or redshift distributions.

There are wrapper scripts in ./KV450 and ./DES that produce photometry and/or line of sight realisations of the deep spectroscopic catalouges (DEEP2, VVDS and zCOSMOS) as they are used e.g. in Wright et al. (2019) for tests of the SOM DIR redshift calibration method.