[daip] WSRT Tsys correction

[Eric Greisen]

M.A.W. Verheijen wrote:
> Dear aips supporter,
> 
> thank you very much for your quick reply to my student's inquiry
> regarding the Tsys correction for our WSRT data.  There seems to have
> been some confusion about what we are trying to achieve, which is
> correcting the visibility data from a standard WSRT 21cm spectral line
> observation for the time-dependent variations in Tsys.  We started with
> a 30min observation of 3C48, followed by 12 hours on our source and
> finishing with another 30 min on 3C147.  Every visibility record has its
> own Tsys measurement, not only the visibilities of the two calibrators
> but also those of the source. 
> 
> With good old trusted AIPS we load the UVFITS files into AIPS with FITLD
> which creates the TY and GC tables.  Attached is a plot showing the Tsys
> behaviour as a function of time for all 14 antennas for LL pol of IF1 as
> stored in the TY table. 
> 
> Subsequently, we run APCAL that takes the GC and TY tables and produces
> a SN table.  This SN table contains the Tsys corrections and shows a
> similar behaviour as the points in the attached plot, be it with a
> different scale. So far, so good,
> 
> Next we try to run CLCAL that takes the CL-1 and SN-1 tables to produce
> a new CL-2 table which we can then apply with SPLAT to produce a new
> dataset in which the Tsys correction has actually been applied to the
> visibility data.  Our problem lies with the behaviour of CLCAL.  Since
> all visibility records have their own entry in the SN table,
> interpolation in time is not necessary.  Therefore we use
> INTERPOL='SELN'.  We want to apply the Tsys corrections not only to the
> calibrators but also to the source itself so we select SOURCES='' and
> CALSOUR=''.  We also apply a smoothing with SAMPTYPE='MWF' and suitable
> BPARM settings.  The full set of parameter settings in included below. 
> 
> The problem is that CLCAL does what it is expected to do, but the new
> CL-2 table it produces only contains data for the two calibrators.  The
> 12-hour timerange that contains the visibility data of our source has no
> data in the CL-2 table.  Somehow, it seems as if CLCAL ignores any
> source that it can not identify as a known calibrator source. 
> 
> Any suggestion you may have to fix this problem would be very much
> appreciated. Many thanks in advance!

Thanks - this does clarify things some - but it is missing the INPUTS to 
CLCAL.  If SOUCODE or CALCODE are set, the result you report might happen.

Eric