Summary of October 27/28 meeting

[Robert Lucas <lucas@iram.fr>]

-- 
Robert LUCAS,            Institut de Radioastronomie Millimetrique
300 rue de la Piscine,  F-38406 St Martin d'Heres Cedex   (FRANCE)
Tel +33 (0)4 76 82 49 42 - Fax +33 (0)4 76 51 59 38 - Tlx 980753 F
E-mail: mailto:lucas@iram.fr                http://iram.fr/~lucas/
ALMA Science Software Requirements Committee
--------------------------------------------

A brief summary of discussions between European Members, in Garching 
on October 27/28, 1999.

- Current membership of Committee is:

Barry Clark (NRAO)              bclark@nrao.edu 
Mark Holdaway (NRAO)            mholdawa@nrao.edu
Robert Lucas (IRAM), chair      lucas@iram.fr
Jeff Mangum (NRAO)              jmangum@nrao.edu 
Michael Olberg (OSO)            olberg@oso.chalmers.se
John Richer (MRAO)              jsr@mrao.cam.ac.uk
Peter Schilke (MPIFR)           schilke@mpifr-bonn.mpg.de 
Joe Schwarz (ESO)               jschwarz@eso.org 
Steve Scott (OVRO)              scott@ovro.caltech.edu 
Francois Viallefond (Paris)     fviallef@maat.obspm.fr 
Mel Wright (Berkeley)           mwright@astro.berkeley.edu 

Ex officio:
Stephane Guilloteau (IRAM)      guillote@iram.fr
Al Wootten (NRAO)               awootten@nrao.edu

- We proposed to have the e-mail list open to any one interested in
listening/contributing (since these discussions have to be open to the
science groups). I'll send an email in this direction to S.G. and
A.W., to be forwarded to the science groups / calibration and imaging
groups.  We nevertheless keep the current members as responsible for
the final reports and for telephone/face-to-face meetings.

- We proposed to set up a conference phone call between all members on
Nov 17th or 18th at 16h UT (Europe=17-18h; USA=09-12h). Please send
suggestions for the agenda. As opposed to the e-mail list, I think
that for efficiency, we need to keep the attendance to
phone/face-to-face meetings limited to Committee members.

- I think we shall probably have a face-to-face meeting in the US
(along with the AIPS++ data pipeline meeting in January? any other
suggestions ?)

- A short presentation was made to all software group members
(basically a summary of what had been discussed on the email list
before). Some questions were raised: do we include provision for the
nutating subreflectors ? Do we need as much as 64 sub-arrays ?

- Then we met for approx. two 1.5 hour sessions. Discussion issues:


1) Observing modes
------------------ 

- Does it make sense to observe in on the fly simultaneously in single-dish
mode and interferometric mode ?
  . not for continuum since the dump times are going to be very
 different
  . may be for line but then it may not be worthwhile since on would spend
 generally much less time in single dish mode to recover short spacing data than
 in interferometry. mode, additionally an Off position is needed in total power, which is
 useless in Interferometric mode;
  . may be one would like to do frequency switch + on the fly in which case one can
 possibly do both SD and Int. at the same time.
 
- in some projects one would like to be able to switch
frequencies/receivers quite fast; note that this is needed for
calibration.

- add holography measurements, beam shape measurements (for thermal,
gravitational variations) to the list of observing modes.

- add measurements of offsets between receivers (specific observing
procedure, probably with planets).

- calibration operations: F.V. proposed that one could self-calibrate
the pointing if the source is strong enough to obtain a map in quasi
real time, by comparing the map obtained with all baselines related
with one antenna to the map obtained with all baselines, or the
current source model. Clearly more work (simulations) are needed on
this point.

- polarisation calibration; at Plateau de Bure, new dual polarisation will 
be installed; but so far no experience with polarisation calibration. 
Can experience at cm wavelengths be transposed at mm/sub-mm  wavelengths ?
Any experience with this in the US ?

- number of sub-arrays: we tried to see how many separate sub-arrays
should be needed. One needs two to calibrate the atmospheric phase
screen in real time while observations go on, and three during
reconfigurations if one wants to use separately antennas in the
previous (e.g. compact) configuration, antennas already in the new
(e.g. extended) configuration) and antennas just moved to calibrate
the baselines. For single dish operations, how many sub-arrays ?

- Do we need a pulsar mode ? ask for specs to the interested scientists
(see science group work plan)

- Baseline calibration: not needed as soon as moved; we may wait till
more are available provided the antennas without known precise
positions are flagged as such in the on-line phase calibration.

- Operators' needs should be included as well, not put in only later as
an afterthought (B. Glendenning).


2) Astronomer input at proposal submission and after acceptance/Scheduling
--------------------------------------------------------------------------
- Astronomer input should be in two steps: one before proposal
evaluation, one after the proposal is accepted to specify actual
observing modes, position and frequency data, ... This `Phase I, Phase
II' concept is is use on, e.g., HST and the VLT.

- interactive scheduling should be used only for maintenance/tests and
time specific observations (ephemeris sky events). Here interactive
scheduling means the mode in which the astronomer is given real-time
control of the instrument.

- automatic scheduling: 
 . Phase II tools should produce typ. 15-minute blocks to enable
interruptions due to weather, priorities changing with LST, ...
 . the PI may set breakpoints where he/she wants to see the data
before going on. However there is the danger of `cheating' by the
astronomer (changing objectives if observations are not
successful). Probably we need some control by one staff astronomer.
In that case the should be fast enough (at most a few days), if we
have continuously evolving configurations.

- does automatic scheduling imply guaranteed data quality ? probably yes.
sensitivity limit should be specified at input, and considered in
proposal evaluation.

3) On line data pipeline: How far can this be automated ? 
---------------------------------------------------------

The consensus seemed to be that one could probably do more with
interferometric data than with single dish data (spectral baseline
removal) and for mixing short-spacing with interferometer data.

- How often should interferometer maps & total power maps be produced ?

4) Other
--------

- The system must be able to dump raw correlator data for
test/engineering use, when needed.

- Will Doppler tracking be done in hardware or software ?