[daip] [!4381]: aips - Changes in VLA U-Band gain curves

[NRAO Helpdesk]

Heidi Medlin updated #4381
--------------------------

          Department: AIPS Data Processing (was: VLA Observing and Data)
       Staff (Owner): -- Unassigned -- (was: Emmanuel Momjian)

Changes in VLA U-Band gain curves
---------------------------------

           Ticket ID: 4381
                 URL: https://help.nrao.edu/staff/index.php?/Tickets/Ticket/View/4381
           Full Name: Andy Biggs
               Email: abiggs at eso.org
             Creator: User
          Department: AIPS Data Processing
       Staff (Owner): -- Unassigned --
                Type: Issue
              Status: Open
            Priority: Default
                 SLA: NRAO E2E
      Template Group: Default
             Created: 03 February 2014 04:33 PM
             Updated: 03 February 2014 04:33 PM
                 Due: 05 February 2014 04:55 PM (2d 0h 0m)
      Resolution Due: 11 February 2014 04:33 PM (7d 23h 37m)


Hi. I'm trying to derive accurate gain versus elevation curves for VLA 15-GHz data taken in late 1992 - the earliest gain curve that ships with AIPS is from March 1995. I found a few data sets in the archive that seem suitable for gain-elevation derivation using ELINT. Running this task seems pretty straightforward and for most antennas I get quite nice looking polynomial fits to the data. If I then compare these to the March 1995 data, I'm surprised by how much they differ. So far, I've only done a detailed comparison for one antenna that has an exceptionally nice polynomial fit, but it is presumably instructive.

An example of my ELINT output is:

ELINT1: ANT=13 STOK=R IF= 1:0.2994E+01 - 0.259E-02*ZA + 0.549E-04*ZA**2

VLA.GAINS in AIPS has:

U 13 19750101 19960601  9.7093E-01  8.4600E-04 -5.8547E-06

Note that the second and third coefficients have swapped signs between 1992 and 1995. One can probably see this better in a plot (attached). The two curves from 1992 are from two different datasets and obviously agree quite nicely. The other two curves are actually identical (there's an arbitrary vertical offset in all curves) and was a result of me confirming that the VLA.GAINS coefficients (a function of zenith angle) were the same as those in the original ELINT output (function of elevation - the original ELINT output files are available from the "VLA/VLBA Polarization Calibration Page").

If you could confirm that such a large change is not actually that surprising, that would be very reassuring. Alternatively, I'm perhaps doing something very wrong.

Thanks,

Andy

------------------------------------------------------
Staff CP:  https://help.nrao.edu/staff