Re: ALMA Science Software Requirements (SSR) -- Step 1

[Al Wootten <awootten@NRAO.EDU>]

I'm not sure how this discussion began but there are three topics which
seem germane to what I have seen.

1) Special purpose use of individual elements
2) Extended configurations may need special treatment
3) Observer interaction with the instrument in real time

1) One element of the MMA design was that one antenna at least would be
available for use of novel equipment, access to frequencies outside the
scope of the array or other traditional single antenna sorts of uses.
Monitoring programs have, for instance, shown the site to be capable of
producing interesting science in the 205 micron window.  Specifications for
the array are a 25 micron surface (7% efficiency), with 20 microns 
(16% efficiency) a target.  I imagine we might achieve better than that 
with some antennas and that one antenna (or more?) might be specifically
designated for this work.  A mode such as this would probably require
more human supervision than the norm, especially if novel equipment were
installed.  Is such a mode an ALMA requirement also (WBS 10.5.6.1)?

2) For any 10km configuration we have studied, the uv coverage is ratty
enough to require inclusion of data from smaller arrays to get a good image,
as Mundy showed in his protoplanetary disk simulation at the ALMA conference.
For some experiments especially on the longer baselines, for which, as Robert
points out limiting conditions will be more frequently encountered, 
automatic generation of maps may not provide the information which a remote
astronomer needs to asses the success of his/her experiment.

3) For many experiments, of course, the short spacing data from total power
observations must be added in.  At present, I'd say we don't know how to
do this in an automatic fashion; an observer might want interferometric
and total power images separately particularly at the higher frequencies
for which total power flux would dominate much of the time.  The exact 
location, for example, of an interesting bit of an object (e.g. a shock feature 
in an outflow, for example)  may not be known
to a few seconds' precision until some data is obtained; should the observer
be able to interact with the datastream enough to modify a field center on the
fly.  Being a cowboy-hatted observer from wild western open range country,
I like the latter model.

Clarity,
Al