[alma-config] short spacing

[John Conway]

Hi,

 I looked at the web site 
http://www.cv.nrao.edu/~awootten/mmaimcal/zlines.html
with the deep field ALMA simulations. Rather than trying to dig out a FITS 
image I  think that it would be simplest to remake my dots
image with the source counts specified 
N(>S) = 7900 S^{-1.1} on that webpage  cutting off in flux 
at a few times below the expected sigma 
in 4hrs continnum obs (i.e. 3 microJansky),

Using a CC file computed as above to generate the uv data
is probably  a bit more robust than taking a FITS image with 
points  on a regular grid.  I can also automatically 
apply the primary beam atennuation to the model
image directly. Keeping the galaxies  as points may not 
be completely realistic but I guess there are lots 
of ultracompact (approx 100pc scale starbursts) 
out there. I guess right now we don't know how big they will 
be. I think a fullly realistic simulation
is too complicated and unneccesary - as long as we are 
in the right ball park then it serves our purpose. 

Mark (below) points out that the source count index does of
course effect the image in important ways  - which 
is of course true. Too flat an index and the image has low 
complexity - too steep and you have a sea of confused 
unresolved giving pseudo-smooth contiuum. I guess in the 
absence of anything else maybe we should go with the 
value which is realisic i.e. index -1.1 ? or pehaps
a tad steeper to make the image a little more complex 
i.e. -1.3?. Anyway I can easily make both, and select 
then.

   John.



J.Conway wrote
  
> > field perhaps he has also something to hand?).  If not I have a
> > simple model callled 10000DOTS which is just a random  distribution
> > of 10000 point sources drawn from a N \propto S^-3/2 cumlative
> > distribution down to some limit. We could easiliy include this in the 
> > simulations that Steven is doing. Since its just a CC file its 
> > very quick to create dummy data using UVCON and CLEAN it. I doubt 
> > from the imaging point of view the exact details of the flux 
> > cutoff or power law exponent are vitally  important  - the main 
> > point is to include -something- as an antidote to all the other 
> > diffuse emission dominated sources - and to represent a imaging
> > area of ALMA science which is not at all  yet represented in the 
> > test set.

Mark wrote

> 
> The exponent may actually be of concern.  A flat exponent
> will basically drive most of the weak sources to a level below the
> deconvolution error limit, effectively reducing the number of imagable
> sources and the source complexity.  A steep exponent will put that
> "sea of weak sources" at a higher flux level, and at 10,000 sources,
> there will be one in each beam (for a 256 x 256 image), so this would
> qualify as a complex source.
> 
> I don't have a strong opinion about what is more appropriate, so I
> would advocate the s^-3/2.
> 
> 	-M
> 
>