1934-638 at 20 cm

Got a calibration problem? Discuss it here.

Moderator: Mark.Wieringa

Post Reply
len067
ATCA Expert
Posts: 66
Joined: Mon Feb 08, 2010 2:35 pm

1934-638 at 20 cm

Post by len067 »

Hi All,

I was wondering if anyone has tried to image the field around 1934-638 at 20 cm to see if there are any sources nearby that may affect the quality of the bandpass calibration at this wavelength? I just noticed that one of my phase calibrators (at 20 cm) had ringing in the bandpass on some baselines and I've come to suspect that this is caused by neighbouring sources in the field. Anyway, this got me into thinking - what if there are similar neighbouring sources around 1934-638? Would the bandpass calibration try to "calibrate" this out and so in turn apply this to the calibration/target sources when the bandpass is copied over? This may affect the overall dynamic range of the observation.

Cheers,

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

Re: 1934-638 at 20 cm

Post by Mark.Wieringa »

Hi Emil,

yes this will certainly be the case at some level. At lower frequencies (<1GHz) it is essential to have a complete model of the field, not just the central calibrator source. At Westerbork they have been building up source models for their main calibrator fields for many years. Of course miriad doesn't let you use anything except a point source for most calibration operations (except selfcal), which makes it a lot harder to do things correctly for high DR imaging.

Cheers,

Mark
shane
Posts: 10
Joined: Mon Oct 25, 2010 8:54 am

Re: 1934-638 at 20 cm

Post by shane »

Hi Emil,

As part of a campaign to find suitable polarization calibrators for ASKAP, we will be imaging a 1.2 deg^2 field centred on 1934 with the ATCA at 20cm. But I guess it is surprising if no one has taken a bunch of archive data for 1934 and tried to make a deep image.

Cheers,
Shane
len067
ATCA Expert
Posts: 66
Joined: Mon Feb 08, 2010 2:35 pm

Re: 1934-638 at 20 cm

Post by len067 »

Hi All,

Hmm, I rather sheepishly came to the realisation that I already had a pile of pre-CABB data for 1934-638 as a result of all of the ATLAS observation (~10 min observations taken over many 10's of days using a variety of configurations). Unfortunately the observations were usually made around the same time for all of the ATLAS observations so that didn't help with the overall u-v coverage (it has a few major holes) but nonetheless it produced an interesting image (attached).

The image is made using uniform weighting and has a dynamic range of ~200,000:1, there are a number of double-lobed radio galaxies within the primary beam at approximately the 0.1% level of the 1934-638 flux. I get about 150,000:1 with a naturally weighted image - the beam is fairly cruddy because of the gaps in uv.

1934-638 is VERY well modelled by a point-source down to the ~90,000:1 level (the residuals seen in the image are at a very low level but are obvious because I used a highly logarithmic intensity scale to show up all of the sources). There is approximately 308 mJy of flux in the field besides that coming from 1934-638.

This could be a really interesting field to image using the piles of historical data available in the archive.

Cheers,

Emil.
Attachments
1934-638.eps.zip
1934-638 image made from 20 cm ATLAS observations.
(441.39 KiB) Downloaded 4148 times
ste616
Site Admin
Posts: 220
Joined: Thu Feb 04, 2010 3:27 pm
Location: Paul Wild Observatory Narrabri NSW

Re: 1934-638 at 20 cm

Post by ste616 »

That's very very cool :) Great job Emil! And thanks for sharing on the forum!
cheers
Jamie Stevens
ATCA Senior System Scientist
len067
ATCA Expert
Posts: 66
Joined: Mon Feb 08, 2010 2:35 pm

Re: 1934-638 at 20 cm

Post by len067 »

Hi All,

Woo hoo! I've finally hit the 500,000:1 mark for the first time! Thanks to Shane for letting me know of a stash of pre-cabb 1934-638 data that he and Dominic collected from the archive. I've attached a new image ... which I'm afraid isn't any more awe-inspiring compared to the last image I posted as I changed the intensity scale to pick up more of the weaker sources ... so the noise looks similar even though it is actually quite a bit lower (x 2.5 times).

I'd have to say that there is probably potential to do much better than this (though probably not by me any more). After loading up the data I realised that it also included all of the pre-cacal data too so I just did a rather nasty triage-style flagging (cutting everything about 10% above and below the 1934-638 flux) - I've also only taken the 128 MHz continuum data centred on 1384 MHz to simplify imaging. I also only performed a single bandpass calibration ... which is probably a bit silly for a decade+ worth of data ;-) Some of this is probably fixed by self-cal because of the way I split the channels up to effectively treat them as independent IFs. Careful flagging and proper initial calibration would probably achieve a much better result.

Cheers,

Emil.
Attachments
1934-638_5e6.eps.zip
1934-638 Field, 500,000:1 dynamic range
(406.55 KiB) Downloaded 4159 times
Post Reply