[mmaimcal]Minutes of Science IPT Meeting, 9 March 2004

[Al Wootten]

Minutes of the Science IPT Telecon 9 March 2004
Hogerdeijde, Wootten, Bacmann, Hasegawa, Kawabe, Tatmetsu, Conway, 
Laing, DiFrancesco, C. Wilson, Mangum,
Lucas, Pety, Guilloteau, Holdaway, Hills

ACTION: Wootten get with DTE on satellite.  Wilson wrote an email introducing 
the topic to him.
ACTION: CSV plan and canvass.  Laing provided the plan to Japanese colleagues.
ACTION: John will work on the antenna location plan
ACTION: Specifications for overlap of tunable filter.  Unassigned, pending
further information.
ACTION: Date change to third Tuesday.  20 Apr  Stay constant in local time, 
change in UT.  Wootten.

Wootten gave a quick summary of action items from the last 
meeting, followed by a terse summary
of milestones, updated for the AMAC meeting.  For most Level 
2 Milestones the Science IPT is
doing well, but many of the Level 3 and below milestones have 
suffered delays.  These are not
critical path milestones, but they are significant secondary goals.

Wootten briefly discussed the Calendar.  

The Operations Group draft has been passed to the Executives by 
the JAO.  The Board is expected
to receive a version to pass to the ASAC by the end of the 
month, when the Board meets in Socorro.
A subgroup has been tasked with identifying 'modes' for Early 
Science.  In conjunction with the
Operations Plan, Robert Laing has drafted a commissioning and 
science verification plan.  IPT
members interested in participating in the evolution of this 
activity should contace Laing or Wootten.

Plans for the ASAC meeting 10-11 May in Cambridge are firming up.  
The Science IPT will be asked
to provide some support for them as they frame responses 
to the Charges which the Board has presented the ASAC.

Wootten discussed the System Design Review conducted 
during ALMA Week. A report will be made
available when the committee finishes its drafts.   A 
table of modes for the correlator with the
new tunable filter is attached to the agenda.  There 
are clearly benefits for science, but
studies continue on the design.  To suppress aliasing, 
subchannel windows need to be overlapped.
Science specifications should be set to determine the 
allowable errors in the overlap region.
Baudry showed an example: Overlap of two spectral 
points of each subchannel resulted in
0.5% amplitude and 0.3 degree phase maximum errors, 
ending up with 93% usable Bandwidth.
Additional overlap can suppress these errors further 
at the expense of usable bandwidth.

At the ATF antenna characterization continues. 

Kawabe summarized discussions during miniminiALMA 
Week in Tokyo.  Key issues included the antenna
specifications and the ACA design. Japan proposed 
a modified LO system--a photonic hybrid option which
replaces warm multiplier assembly.  Also discussed 
was a power generator for whole system with 
presentation from Canio Dichirico on this.  Hasegawa 
noted that the Japanese had considered how to
optimize the OTF capability of the 12m antennas.  
A simulation result was presented.  It was found that
tuning antennas to efficiently do OTF results in somewhat 
different specs from those of the main 
array; a memo is being written describing the results 
of the study.  Can this requirement be 
accommodated in parallel with requirement for fast switching 
on the main array.  Another issue on 
first LO.  A possible concern on photonic hybrid has been 
examined with no outstanding concern.  
Some higher risk thought to exist by the LO Group; this is 
still under study.  In the Diet, the
House of Representatives has passed the 2004 budget, sent to 
house of councilors.  Passage expected 
by 4 April.  

Chris Wilson asked about the utility of a calibration satellite 
at mm wavelengths.  CA PS for Gemini 
has interest for putting a source on a satellite.  This 
would serve as an absolute calibration device.  
It was suggested that this be discussed with Darrel Emerson.  
This would be useful for holography.
It could also have some uses for flux calibration.  

Hogerheijde noted that a dedicated website will be made for 
DRSP.  He needs to make sure that script 
will work.  DiFrancesco asked how to modify the DRSP.  
Hogerheijde is the contact for this. Laing is
developing some polarization projects for eventual inclusion.  
The ASAC is considering asking
DRSP contributors to supply additional information on the level 
of calibration accuracy needed 
for their experiments.

Robert reported on SSR activity--meeting tomorrow.  Some tests 
of obs prep system and offline testing.  
These worked quite well, please examine the reports.  Memo 
underway on data rates.  Also under 
discussion is the amount of work required for the tests.

Conway has submitted draft early science configurations on 
the web, jump frog triangles of
configurations.  This assumed 172 pads available by the 
end of 2007, whereas some will not be ready.
Configuration 6 will not be possible, some changes on 
configuration 5.  This will have some
repercussions on the inner configurations also.  Will submit 
it to almaedm at that point.
Eventually plans needed to build this up.  

Conway also discussed the cycle time of whole array and 
constraints on the transporter.  There
is a strawman in the comments on the transporter spec 
document in the absence of a proposal from 
science.  We will come up with a strawman now that we 
know the total number of pads.  We think a 
9 month cycle is good, which has an average of 4 moves 
every 2.5 days.  This will be specified in
the antenna movement document which awaits further comment 
in the almaedm DAR.

Conway described the baseline calibration plan he began 
using monte carlo simulations and 
kolmogorov turbulence.  Then he attacked it analytically 
in the case of one moved antenna and a 
single antenna with which its baseline is calibrated.  
Occasionally the whole array is calibrated.  
The question is the balance between these.  The biggest 
problem is getting the z component, as only 
one hemisphere is being sampled.  One finds that the residual 
phase scales at a strong power of 
baseline length.  Math is analogous to two antennas 1 km 
apart with a 36 km atmosphere.  So one must
use short baselines, <1km perhaps 500m. Once array is largest, 
the calibration needs to be
done with the nearest unmoved antenna, probably in night, 
probably takes one hour.  Just takes
a few antennas from the larger array and does not cause a 
large loss of baseline hours.
Errors increase as one goes further out.  One builds in a 
guard against systematic error.
Is the 65microns on each coordinate or the total.  Can 
afford to have some antennas with larger
than this error.  Will run tests to show the effect of 
baseline accuracy on image quality.
Also worried about dry path effects owing to altitude effects.  

Respectfully submitted,
Al Wootten