WSRT calibration problems

Got a calibration problem? Discuss it here.

Moderator: Mark.Wieringa

Post Reply
Arpad
Posts: 12
Joined: Thu Feb 24, 2011 1:50 am
Location: Germany

WSRT calibration problems

Post by Arpad »

Hi,

I am trying to reduce WSRT 92 cm data with miriad.

I came across something peculiar.

I have 4 calibrators, A,B,C,D. A and C are flux and leakage calibrators, B and D are polarization calibrators. C and D are more like backups in case A and B are bad.
To test the calibration, I derived the gains and then treated all the other calibrators as targets. I just want to look at stokes I, so I used gpcal to get the flux calibration and then copied the gains to the other calibrators.
Now, apparently gpcopy (or something else) has problems if the observation date of the data set is different to the date of the calibrator that was used.

A and B were observed on the 20th around noon, C and D were observed 12 hours later, which would lead to a different day, 21th.
If I use gpcopy to copy gains from A to B,C and D, I can only image A and B. Accordingly, if I copy the gains from C to A,B and D, I can only image C and D.

I tested this in uvplt. If I use options=nocal, I can plot the visibilities, without it, it says there are 0 good data points, as if it were flagged.
But I also may have found a workaround. When I use A to get the gains, and use puthd to change the date in the headers of C and D, I can actually image all of them.
(curiously, if I use puthd in=A/time it says it creates a new variable)

Why is that? Is this intended? Is this a Bug? Is the workaround ok or am I missing something?

I am also a bit confused on how to actually do the polarization calibration. WSRT has an equatorial mount, so which option would I need in calibration? options=xyref or xyvary, or none? Because of this mount, the phases should more or less remain constant over time, right?
If so, what would be the correct procedure if I know the fluxes the calibrators should have?

gpcal on A to get gains and leakage
gpcopy A to B
gpcal on B to get polarization
gpcopy B to target

would that be correct?

Thanks you,
Arpad
Mark.Wieringa
ATCA Expert
Posts: 297
Joined: Mon Feb 08, 2010 1:37 pm

Re: WSRT calibration problems

Post by Mark.Wieringa »

Hi Arpad,

The miriad manual explains how to change the default interpolation tolerance (0.5 days) for gain tables in the Calibration Strategies chapter:
http://www.atnf.csiro.au/computing/soft ... node99.htm

I have no experience with trying to do polarization calibration of WSRT data in Miriad, I'd suggest finding someone at Astron who has tried this before - from what I can see they recommend using their own software for polarization data. You can see how far you can get, but things like solving for polarization and leakages won't work if the feeds don't rotate on the sky.

xyphase - this is not related to the mount type, it is just the drift in phase between the x and y channel (due to electronics, cables etc), it should always be pretty small (few degrees) unless there is a problem.

Cheers,

Mark
Post Reply