[daip] Re: CLCOR

[Chris Phillips]

On Sun, 1 Dec 2002, Leonia Kogan wrote:

> I do not understand what does "modified version of CLCOR" mean.
> I guess you added the print of JD.

Yep, I took the code and added a few write(*,*) lines to see what values
were being passed.

>
> If it is right then you need to print 'OBSDAT' also and compare it with ...
> OBSDAT is the only parameter which control calculation of JD
>
>             CALL JULDAY(OBSDAT, JD)

This confirms my suspicions. The Epoch of the apparent->mean is calculated
at the beginning of the OBS date (in this case 20010315). The date => JD
calculation is done correctly (as 2451983.5). Then JPREC does the correct
conversion for this date/time.

But is this the right epoch/time to do the conversion? With the original
data (before running CLCOR) the mean<=>apparent conversion had been done
at the time when the source was first observed in the observation, in this
case 26.5 hours after the OBSDATE.

It looks to me as either FITLD is doing the mean<=>apparent conversion at
the wrong date/time or CLCOR is.

An important thing here is that if I do a CLCOR with a shift of 1mas in
RA/DEC the apparent position changes by the correct amount, but there is
about a 0.5" shift in the mean position (mostly in RA).

Cheers
Chris