[bananas] 31DEC23, 31DEC22 AIPSLetter

[egreisen]

The 31DEC22 final AIPSLetter is now available at the web page

     http://www.aips.nrao.edu/index.shtml

It contains two recent items of particular interest.  The first
is the issue of applying polarization D terms to the parallel-hand
visibilities as well as the cross-hand.  The 31DEC22 version did
this from 30 July through 7 December, at which point we determined
that this is not a correct procedure.  The code was then reverted
to the previous usage in which D terms (when DOPOL > 0) are
applied only to the cross-hand visibilities.

The other, more insidious issue at frequencies less than about 3
GHz, concerns the ionospheric Faraday rotation.  A geometry error
was found in TECOR which caused the IFR values for positive hour
angles to be somewhat in error.  SNPLT shows a jump in IFR at zero
hours as a symptom of the problem.  The usage of the IFR in PCAL
was also in error, with no IFR applied to spectral window 1 and the
IFR correction appropriate to channel 1 of the previous spectral
window applied thereafter.  It is not clear how much the D terms
were affected by this issue.

The TECOR bug was only corrected at the end of December 2022.  Users
with the 31DEC22 and 31DEC21 versions are encouraged to run a MNJ
to correct their installations.  31DEC23 may now be installed.


Eric Greisen