[evlatests] WIDAR data of Friday, Saturday

[Vivek Dhawan]

Recent data sets are looking very good. Not all the juice has been
wrung from these, but a quick update follows.

1. C band 'baselines' for ~4 hours, slewing around the sky to examine
   the phase connectivity. The residual phases are very well behaved,
   as good as the VLA. (VLA data were taken in parallel.) There is no
   sign of a frequency offset, other than the known phase change with
   elevation which affects many antennas, and is being fixed. The worst
   case change is < 1 deg phase over 1 deg elevation throw, not enough
   to damage the coherence in a detection experiment, and it self-
   calibrates out on a strong source. Ed Fomalont looked at this data
   as well, and reported the same to a subset of us.

   To demonstrate the phase transfer in the image domain, just for fun
   I calibrated on a single scan, applied it to all 4 hours, and imaged
   all 64 sources spread over 10-120 degrees elevation, and 360 azimuth.
   The resulting images show all targets within ~1/5 beam of their
   expected positions even in extreme cases, e.g. 0440-435 calibrated
   by 0927+390 observed 2.5 hours later.

/home/casa-dev-14/mrupen/WIDAR0/
widarbase04-08.55127.37700130787.uvfits (27 GB)


2. A long C-band run on 3C84, 3C147, plus others. Phase connection is
   very good, except at low elevation - shadowing was occurring as well.
   I am still working on this for image quality, noise averaging down
   as sqrt(Nchan) etc., and will report later.  This too has a matching
   VLA dataset.

/home/casa-dev-14/mrupen/WIDAR0
C0555plus23_000.55128.36695758102split.uvfits (70 GB)