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

["Stephane Guilloteau" <guillote@iram.fr>]

Hi everybody,

    Although I fully agree that from a scientific perspective, images should
the main data product of ALMA, I believe there are at least 2 important
issues which may change the way we present this

1) Cost: It may be easier for funding reason to move the imaging out of the
ALMA project. This allows recurrent costs to be shared between several
projects, and is logical for software which tend to be re-used as much as
possible. We are not at all sure of the available funding level, but can
certainly expect additional manpower on software issues from other sources.

2) Archiving: My impression is that archiving images may not be the best
thing to do. A number of scientific topics are better adressed directly in
the UV plane. I would tend to archive calibrated, adequately compressed, UV
data, from which images can be easily produced (with tapers, etc... whenever
needed).

>From that, I would focus our efforts in getting an automated calibration
pipeline. This is not such a trivial task in the mm range (will be soon, but
than the sub-mm will remain non trivial for sometime...).

Although dynamic scheduling implies automatic quality assessment (which
includes, but goes somewhat beyond, automatic imaging), the required
automated imaging capability could be relatively crude compared to "final"
imaging. My proposal would be to allow us to move this "final" imaging step
outside the scope of ALMA, if resources show limitations.

Whether an expert software or a "real" dedicated expert performs the imaging
for the astronomer community is also a financial problem. The human eye is
quite efficient at grasping things that we don't enough know yet how to
describe to an expert system. I think we should include the cost of a small
team for this purpose in the operation budget.

            Stephane