Quick and easy background subtraction

Setting up a background model which takes into account the 37-day cycle requires a substantial amount of effort. However, a previously determined set of universal background models is made available, each of which is derived from off-source observations made during a period 3-4 months (the background models are located in the directory $XACAL; see appendix A ). The background model closest in time to the source observation should be used. These models can give reasonable background estimates but with the following limitations:

1. It should be possible to compress the detector identification of the coefficient files (of the background model) so that it mimics the detector identification of the source observation file. Generally, the coefficient files have the following detector identification (even and odd numbered positions corresponding to the top- and mid-layer respectively):

   qcl>list id_mask
    id_mask(2)
       0101010101010101
       1010101010101010
   qcl>

   Occasionally, detectors 5 & 6 were switched off because of electronic noise. In this case, the coefficient files have the following detector identification:

   qcl>list id_mask
    id_mask(2)
       0101010100000101
       1010101000001010
   qcl>

   These detector identifications can be compressed to the following:

   qcl>list id_mask
    id_mask(2)
       1111111111111111
   qcl>list id_mask
    id_mask(2)
       1111111100001111
   qcl>

   Therefore, MPC2 files, which has a mixture of data from the top- and mid-layers, must be compress to a single detector ident (see section 5.1).

2. The estimates of the background in the mid-layer below 10 keV are relatively poor.

3. The universal model in late 1988 (September-December) may be affected by the problems with the gain calibration on two of the LAC detectors (7 & 8) at this time.

Do not be discouraged by the above limitations, as the results obtained using the models can be superior to those obtained using a single local off-source observation. A qcl procedure called bgduniv.qcl has been written to provide easy and convenient access to these models; it uses the programs bgdsaa37, bgdsim and subcb. The only input it requires is a sorted cleaned data cube. It prompts for only four parameters: the name of the on-source cleaned data cube; the names of the two output files created (containing the estimated background and the background subtracted data cube); and the background model to be used. The following is an example of its use:

qcl>.bgduniv
 BGDUNIV.QCL is QCL procedure for subtracting Ginga background using
 standardised files.
 
 There are a number of sets of background coefficient files which can
 be used by this program. Each set is derived from a few months of
 background data. The following are currently available
 
 87b2      => Jun 87 - Sep 87
 87c       => Sep 87 - Dec 87
 88b       => May 88 - Aug 88
 88c       => Sep 88 - Dec 88
 89a       => Jan 89 - Apr 89
 89b       => May 89 - Aug 89
 89c       => Sep 89 - Nov 89
 8912_9001 => Dec 89 - Jan 90
 9002_9003 => Feb 90 - Mar 90
 9005_9006 => May 90 - Jun 90
 9007_9008 => Jul 90 - Aug 90
 9009      => Sep 9           (Detectors 5 and 6 off)
 9010      => Oct 9           (Detectors 5 and 6 off)
 9011      => Nov 9           (Detectors 5 and 6 off)
 9012_9101 => Dec 90 - Jan 91
 9102_9103 => Feb 91 - Mar 91
 9012_9101 => Apr 91 - May 91
 9106      => Jun 91
 9107      => Jul 91
 other     => User will be asked to specify directory containing data
files
 
QCL_CL Source cube = HDS_NODE =mkn335_91a.cube_cl_sq
QCL_CL Model background cube =  =mkn335_91a.bgen
QCL_CL Background subtracted source cube =  =mkn335_91a.sub 
QCL_CL Background files to be used = 87c =9107
 Background files from the following directory will be used:
 $XACAL/bgdstand_9107/
 
 Use BGDSAA37 to insert decay parameters into source cube
 
** BGDSAA37 Version 1.6 Wed Sep  3 20:13:49 1997 **
 
 Use BGDSIM to create model background.
 
** BGDSIM Version 1.1 Wed Sep  3 20:13:58 1997 **
 
 BGDSIM generates variances from the statistical errors scaled by a
user chosen multiplier
 
 Use SUBCB to subtract model background from source cube.
 
** SUBCB Version 1.0 Wed Sep  3 20:14:02 1997 **
 
 
     SCB_DATIN1: mkn335_91a.cube_cl_sq
      Shape :    393 x    48 x     2   =  37728
              TIME    PHA     IDENT   
 
     IDENT masks
       1  0101010101010101
       2  1010101010101010
 
 
     SCB_DATIN2: mkn335_91a.bgen
      Shape :    393 x    48 x     2   =  37728
              TIME    PHA     IDENT   
 
     IDENT masks
       1  0101010101010101
       2  1010101010101010
 
 DATIN2 is a CUBE.
 DATIN1 is a CUBE. ALL of DATIN2 can be subracted from DATIN1,
 or a single time series or PHAD extracted from DATIN2 can be
 subtracted from each time series or PHAD in DATIN1
 Add comment fo history records
 
** COMMENT Version 1.1 Wed Sep  3 20:14:03 1997 **
 
 
** COMMENT Version 1.1 Wed Sep  3 20:14:04 1997 **
 
qcl>

As mentioned earlier, the universal background subtraction method is not recommended for observations performed during the first six months and last three months of operation of Ginga.