Self-calibration

[and460]

Hi guys,

I have a question about whether or not GPCAL can be used instead of SELFCAL on the CA if the source is at the phase centre, but not at the primary beam centre. I explain what I mean a bit more below.

I have some mosaic data which I am trying to improve the calibration for. In one of the pointings, there is a source that is showing some strange structure in the Stokes I SED, so I thought self calibration might improve matters. Instead of using selfcal for this source (which has a flux density of ~100mJy), I wondered if I could use UVEDIT to shift the source to the phase centre, then run GPCAL on it to get the gain corrections. This is far easier than using selfcal, so I'm trying to think if this is a valid thing to do or not, and whether the gain corrections thereby obtained can be propagated to the other pointings in the mosaic. Finally, I was wondering what your thoughts were on using the qusolve option when doing this. qusolve means that GPCAL will attempt to solve for the polarisation leakages assuming their time-independence, which will not really hold if the source tracks through the polarised beam over the course of the observation (as it will if it is not observed at the primary beam centre). However, I was thinking that it might find some sort of leakage solution which minimised the effective leakage in the data when taken as a whole over the course of the observation. Or maybe that is just wishful thinking.

Anyway, the field is quite bare of other sources. There is another source a fair bit further out in the field, which comes in at about 10mJy when primary beam corrected, but it is quite far out in the primary beam so gets attenuated heavily in the un-PB corrected data.


On another slightly unrelated matter, I tried running selfcal and I get an error message saying:

Calculating the model for /srcCal5/May/d6/test3/forn_29.2950.iclmod
### Warning: Error processing text file /observatories.dat
### Warning: Error opening file
### Fatal Error: No such file or directory

I checked the forum and it said the file in question should be located in $MIRCAT. echo $MIRCAT gives me /miriad/cat/, and the observatories.dat file is sitting in there. So it seems like miriad should find it OK. Any idea what I have got set up wrong?

Cheers,

Craig.

[Mark.Wieringa]

Hi Craig,

Yes, if you have a strong point source and you shift it to the phase centre you can try gpcal on it and you should be able to use the gain corrections for nearby fields. If the source is not really a point source though I would stick to phase only. You could also try gpscal - this lets you enter a point source model, so you don't have to phase shift the data.

I'd stay away from trying qusolve or leakage solutions on an off-centre source. The terms would be strongly time variable and disentangling changing q,u and leakages would not work well I think.

On your /observatories.dat issue - it appears that MIRCAT is not defined at the point where you are running selfcal from, otherwise it would complain about /miriad/cat/observatories. Make sure miriad/MIRRC is run in every terminal (preferably in the login) before you access miriad tasks.

Cheers,

Mark

[and460]

Thanks Mark!