Description of COMBO-17+GEMS Master Catalogues

Data Product

Two catalogues are provided: Each catalogue contains around 300 columns, roughly 250 from COMBO and the rest from GEMS.

Purpose:

  1. You can look at objects fainter than R~24, the limit of the proto master catalogue.
  2. You find up to three sources from the complementary catalogue within 2 arcsec, so you can also tell when objects are in a complex environment where identifying the true counterpart may be non-trivial.
  3. Alternative counterparts have chi-squares of their match quality from position and magnitude, so the most likely single counterpart can even be identified in a complex environment.
  4. You find multiplicity values (how many objects in 'the other list' point to this object in 'this list' as the most likely counterpart) for any given object in any of the two lists. This can be used as a warning flag for blends, where information on a 'master object' from one list may not apply to all its individual components found in the other list.
  5. You don't have to redo the match of two source lists if you want to consider possible counterparts at your desired level of positional match (matches above 2 arcsec positional difference do not need to be investigated, as even in nearby 30 arcsec large galaxies, the COMBO-GEMS offset never gets so high).
  6. You can restrict your desired object sample by selecting from this catalogue, but it is not likely that you need to go beyond the content of this catalogue. Sub-selecting from here is trivial for you to do, asking for a remake/rematch would involve more work and should ideally not be necessary (if everything was done right here). Details:

    The columns appended fall into two categories:
    • Columns describing up to three objects from 'the other list' which could possibly be related to the given source from 'this list', and holding positional differences and chi-square values for the match based on position and magnitude
    • The full column set holding all information from 'the other list' for the most likely counterpart
    The most likely counterpart is the one with the lowest chi-square, not necessarily the one with the shortest positional offset. About 10% of the sources have more than one possibly related object within the maximum search radius of 2 arcsec. In most cases, the second object is just a nearby, but unrelated source. However, a significant number of COMBO sources have two GEMS sources within 1 arcsec. In the COMBO image and photometry these objects are blended. If the two GEMS sources are at the same true redshift, then COMBO will see a mix of stellar populations and still get the redshift right. But if the two GEMS sources are physically at different redshifts, the COMBO photometry will see the sum of two SEDs which are redshifted with respect to each other. If one source is much brighter than the other, the COMBO photometry and interpretation is dominated by that one and should reflect its redshift and SED. If they have comparable brightness (maybe within a 1:3 ratio or so), COMBO will be confused and come out with the wrong redshift. We had such a case in CADIS, where a source at z=0.6 was blended with one at z=1.2, and the CADIS photo-z was 0.9, by chance the average. Depending on the application, it may be safer to eliminate all objects from a sample, where a blend could have occured.


    The table is provided as a FITS table in order to

    • remove the need for all future times to waste time on any sort of format description (just read the table with a FITS reader into an array of strctures, then use the object number as an array element index and the column header as a structure tag name, and you can access every entry regardless of its physical location in the table)
    • to allow the table to change format at any time without any warning to any potential user and still ensure that not any of your software using it needs to be changed (e.g. when including further information as further columns, or changing the order of columns by random chance - rights for this to happen are reserved by the producer)
    A description of the columns originating from the COMBO catalogue can be found here and one of those originating from the GEMS catalogue by Dan can be found just below. In order to conform with the FITS concept I have given column labels to the data entries from the proto_tot catalogue by Dan, which are:
    Byte-by-byte Description of file: proto_tot.cat.dp
    -----------------------------------------------------------------------------
       Bytes  Column Label  #   Description 
    -----------------------------------------------------------------------------
       1- 24  GEMSname      1   GEMS sexagesimal coord.-based name
      26- 35  r_a           2   Right ascension (2000)
      38- 48  dec           3   Declination (2000)
      51- 58  flux_best     4   Total flux (SExtractor FLUX_BEST)
      60- 65  fluxerr_best  5   Total flux error (SExtr. FLUXERR_BEST)
      68- 73  mag_best      6   Total magnitude (SExtr. MAG_BEST)
      75- 80  magerr_best   7   Total magnitude error (SExtr. MAGERR_BEST)
      83- 87  background    8   Background (SExtr. "local" BACKGROUND)
      90- 95  isoarea_image 9   Isophotal area (SExtr. ISOAREA_IMAGE)
      98-103  x_image      10   Image center x coordinate (SExtr. X_IMAGE)
     105-110  y_image      11   Image center y coordinate (SExtr. Y_IMAGE)
     112-116  theta_image  12   Position angle (SExtr. THETA_IMAGE)
     119-123  ellipticity  13   Ellipticity (SExtr. ELLIPTICITY)
     126-130  fwhm_image   14   Image FWHM (SExtr. FWHM_IMAGE)
     133-134  flags        15   SExtractor FLAGS
     137-140  class_star   16   Star/Galaxy class (SExtr. CLASS_STAR)
     148-160  source_name  17   ACS source image name
     162-175  mask_name    18   ACS mask image name
     178-181  exptime      19   exptime (sec)
     184-189  ABsys_zp     20   ABsys zeropoint
     192-197  Kron_radius  21   Kron radius (SExtr. Kron apertures)
     200-208  Cxx_ell      22   Cxx object ellipse parameter from SExtr.
     210-218  Cyy_ell      23   Cyy object ellipse parameter from SExtr.
     220-229  Cxy_ell      24   Cxy object ellipse parameter from SExtr.
     252-253  nr_overlap   28   Number of overlapping sources
    -----------------------------------------------------------------------------
    
    Some column names have been changed to allow for the presence of similarly labelled GEMS and COMBO columns alongside in the same master table. In the GEMS master catalogue the following columns originating from COMBO are renamed:
    r_a          renamed into   COMBOr_a
    dec          renamed into   COMBOdec
    mag_best     renamed into   COMBOmagR
    mag_best_err renamed into   COMBOmagRerr
    mag_I_best   renamed into   COMBOmagI
    class        renamed into   GEMSclass
    class_val    renamed into   GEMSclassval
    
    In the COMBO master catalogue the following columns from the GEMS table proto_tot.fits (as detailed above) are renamed:
    r_a          renamed into   GEMSr_a
    dec          renamed into   GEMSdec
    mag_best     renamed into   GEMSmag
    magerr_best  renamed into   GEMSmagerr
    flags        renamed into   GEMSflags
    class_star   renamed into   GEMSclass
    



    Last update Oct 27, 2004, CW