about using the uvfit to get the radio flux

[hzhangshao2011]

hi , Miriad expert,
When I process the radio data with 'uvfit' command to get the radio flux, I usually got a negative flux value?
What did the negative value mean? What caused such kind of result and how to correct it? I have puzzled with
this problem for a long time. Its lucky for me to succeed to register on Miriad Forum after trying many times.
I expect the miriad expert to give me some suggestions and experiences about this problem. Thanks!

Best regards,
hui

[ste616]

Hi Hui,

Without further detail, I can only make a guess here at what might be going wrong.

I suspect that uvfit has failed to find the source you are interested in, and thus is essentially returning noise power. This happens in imfit as well. My recommendation is that you specify the initial guess for the position with the 'spar' parameter, which may help uvfit converge to a more useful solution.

[hzhangshao2011]

Hi...
I will attach jamie's reply to my email, to share the information with any other people.
Hui
---------------------------------------------------------------------------------------------------------
Hi Hui

I think the negative flux value is because it has not found the source you're looking for, and is therefore just returning noise power. I see that you have set spar (something I recommended when I just replied to your ATCA forum question), and that you expect the source to be at the phase centre (ie. position offset of 0).

In the special case where the source you are trying to measure is at the phase centre, I recommend that you use instead one of the dedicated uv flux measurement tools: uvflux, calred or closure.

uvflux vis=h1743-322.uv stokes=i

calred vis=h1743-322.uv stokes=i interval=9999

closure vis=h1743-322.uv device=/xs stokes=i options=amplitude

The last command, closure, has the advantage of using a method that, for strong enough sources, is immune from antenna-based phase calibration errors.

The other thing that concerns me is that you are using the option "noif" during your atlod step. This is not a recommended option anymore. I would recommend using the options "birdie,rfiflag,xycorr,noauto" only. I'm not convinced yet that this will be causing you problems, but it would be a good step to try if the other measuring tasks do not help. Other than this, your procedure looks good.

[hzhangshao2011]

Hi, everyone,
I think jamie is right. I get the negtive flux because I do not find source.
I can use difmap to map the visibilities and get the position of the source.
This positon is set to be the initial parameter in uvfit and can get the right radio flux.

But there is still a question puzzle me. I only care the data of LMXBs, such as gx339-4, h1743-322 etc.
There are several scans for these sources during one observation, each scan last about 15 minitues,
I can map the source for many scans by using difmap, but I can not map the source for only one scan, so how
could I get the position of the source? Why I can not map one scan, is it because the uv coverage is small?
Is there anyone can help me? I hope I have describled clearly.
Thanks!

Hui