[daip] [!5234]: aips - Daynumbers from VLBA datasets

[Adam T Deller]

Adam T Deller updated #5234
---------------------------

              Status: Open (was: Response Overdue)

Daynumbers from VLBA datasets
-----------------------------

           Ticket ID: 5234
                 URL: https://help.nrao.edu/staff/index.php?/Tickets/Ticket/View/5234
           Full Name: Adam T Deller
               Email: deller at astron.nl
             Creator: User
          Department: AIPS Data Processing
       Staff (Owner): Eric Greisen
                Type: Issue
              Status: Open
            Priority: Default
                 SLA: NRAO E2E
      Template Group: Default
             Created: 28 July 2014 01:13 PM
             Updated: 08 April 2015 01:37 PM
                 Due: 10 April 2015 01:37 PM (2d 0h 0m)
      Resolution Due: 15 April 2015 03:32 PM (7d 1h 54m)



To help characterize the problem, I ran a short test where I correlated a dataset which spans a day boundary twice: the first time I included two scans, one on either side of the day boundary, and the second time I included only the later of those two scans.  So there are two datasets:

1) reference day 20150326, one scan on 20150326, one on 20150327; and
2) reference day 20150327, only one scan on 20150327

So both datasets have exactly the same data for the latter scan, but they have different reference days.  When I load them into AIPS using refdate = '20150326', I see that running CLCOR with opcode='EOPs' gives phases which differ on the latter scan by a fraction of a degree.  Worse, though, is that running TECOR gives DDLY values which differ by several nanoseconds on that second scan.

I confirm that if I leave the refdate empty, then TECOR produces identical DDLY values for the two datasets (although then the day numbers are of course different, which was the whole reason for setting the refdate in the first place, so I could copy SN tables between datasets).  So I suspect that the fact that the RDATE values are wrong in various tables is indeed the root of the problem, and it is affecting (at least) CLCOR and TECOR.  I expect that the solution requires propagating the refdate specified in FITLD through to all of the RDATE values in the MC table, CT table, etc.

The fits-idi files can be found in /home/swc/difx/projects/dayspan-test/fromcorrelator/ and are called VLBA_BR198V5_test-1day_BIN0_SRC0_0_150407T215822.idifits and VLBA_BR198V5_test-2day_BIN0_SRC0_0_150407T215827.idifits.

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