[alma-config] DOTS image

[Al Wootten]

>From Mel Wright:

- One of the model images I used to study CARMA's imaging was 
8 point sources. It's written up in http://bima.astro.umd.edu/memo
memo 73 with a summary in Imaging 1999 conf.

		-----------
"We imaged a model composed of eight compact sources with flux densities
from 1 to 0.01 Jy with randomly sampled  $uv$ data.  The image noise
was 1 mJy/beam.  The images were deconvolved using the CLEAN algorithm,
which is well matched to deconvolving a field of compact sources.  With
only 1000 $uv$ samples there were many spurious sources at the 1\%
level and a 20\% error in the measured flux density of the 50 mJy
source.  With 2000 $uv$ samples there were several spurious sources at
the 0.3\% level.  With 4000 $uv$ samples there were no spurious sources
at the 0.3\% level, although several appear at the 0.2\% level and
there was 20\% error in the measured flux density of the 10 mJy
source.  In all cases, the RMS noise was close to the theoretical noise
level of 1 mJy/beam, and the dynamic range was 1000:1.  The level at
which spurious sources appear, or the maximum difference between the
image and model are useful measures of the image fidelity, but these
are hard to use if the real source distribution is unknown.  Empirical
estimates of the image fidelity may be obtained by varying the imaging
parameters and measuring the RMS differences between the sources on the
ensemble of acceptable images."
		-----------

Although I used clean, one could argue that an expert would identify
the stronest point sources and subtract them from the uv-data and
iterate this process to enhance the image fidelity.

Anyway, see how the point source model works for ALMA.

			cheers,

				Melvyn.