Clean and MEM

Got an image problem? Let us help!

Moderator: Mark.Wieringa

Post Reply
fuj006
Posts: 2
Joined: Tue Apr 22, 2014 8:37 pm

Clean and MEM

Post by fuj006 »

I have a question about the imaging of the ATCA HI line mapping data.

I am now concerning the imaging of HI mosaic data without continuum subtraction.
Imaging by "mosmem" doesn't work well around the strong, point-like background sources (the shape of the spectra was somewhat winding), may be due to the weakness of MEM method to the point sources.
I think it is better to do the "clean" first for these point sources, and then do "mosmem" for the residual image, but unfortunately I could not figure out how to do such kind of Clean and MEM combination.
Could you tell me how to do this on MIRIAD?

I would be happy to be told if there are any better way to do this kind of imaging.
Mark.Wieringa
ATCA Expert
Posts: 297
Joined: Mon Feb 08, 2010 1:37 pm

Re: Clean and MEM

Post by Mark.Wieringa »

Hi,

The alternative to doing continuum subtraction is a lot more work for mosaic data. I think something like the following recipe should work:
1.Image the data with invert with options=mosaic
2. Use mossdi to clean the cube to the level where all the strong point sources are in the model.
3. Use demos on the clean model cube to create a model for each mosaic pointing.
4. Split the visibility file by pointing
5. Use uvmodel to subtract the matching clean components from each visibility file.
6. Image the residual data with invert options=mosaic and all the residual visibility files as input
7. Use mosmem to model the extended emission.
8. Use restor to subtract/restore the mosmem model
9. (optional) Use restor with options=convolve to on the mossdi model and use maths to add the mosmem and mossdi images together

Cheers,

Mark
Last edited by Mark.Wieringa on Fri Jan 16, 2015 10:16 am, edited 1 time in total.
fuj006
Posts: 2
Joined: Tue Apr 22, 2014 8:37 pm

Re: Clean and MEM

Post by fuj006 »

Thank you very much for your kind advice.

I have a trouble with 5th process (uvmodel).
I have performed "uvmodel" as follows, but the output image seems to show no difference before model subtraction.

Task: uvmodel
vis = one point visibility data
model = model data after "demos"
options = subtract
out = output filename

Is this a problem with previous modeling process, or do I need to set the other parameters?
Mark.Wieringa
ATCA Expert
Posts: 297
Joined: Mon Feb 08, 2010 1:37 pm

Re: Clean and MEM

Post by Mark.Wieringa »

Hi,

I don't understand why there is no difference in the before/after subtraction image - I can only suggest to double check the input and output files and make sure the correct ones are used for each step.

Another thing to try would be to use mossdi to do the point source clean as before but then feed the resulting model straight into mosmem as the starting model. That avoids the uv subtraction stage giving you problems and should be quicker way to achieve a similar result.

Cheers,

Mark
Post Reply