Several presentations from the Leiden Meeting are available at
ALMAEDM
.
New Business -- Agenda
Level Two Milestones
- Soon-to-be-final Level 3
milestones from Bryan Butler: "Included are the modifications based on the calibration
group telecon on December 16 and the Leiden meeting."
- Project news/updates;
Calendar
Meeting of the group formerly known as the ACC 24-25 Feb Washington.
The Agreement will be signed, creating the ALMA Board.
26 Feb 2003 Charlottesville meeting on integration center design.
ASAC Face-to-face Meeting 2-3 April 2003 in Grenoble, France at IRAM.
May 26/27 ALMA Board face-to-face (Europe)
ALMA Week 2 June Victoria
Several of the below items were also discussed at length during the recent
mmaimcal and calibration telecons, but it is important to have the input
and comments of the entire science IPT before recommendations are made to
the project/ASAC.
See ImCal
and
Calibration
for background information on some of the above items.
Matters Arising
- ASAC telecon feedback (Stephane/Al/Ewine) ASAC
Telecon Agenda
5 Feb
Minutes of meetings may be found at
the ASAC ALMAEDM
page. Wilson: Agenda draft with names to be circulated soon. Interim
report to new ALMA Board under construction. We will include front
end stability specs, calibration, configuration, perhaps correlator, software/aips++
on the agenda. Science IPT should present correlator, calibration,
front end specs. EvD: ACTION item to summarize these in a page and
provide to the ASAC.
- Polarization widget: central frequency? See material at ImCal, above.
(Stephane/Al) This is one FE Spec which needs to be defined. There
is one slot, one frequency. Three options--high frequency option, low
end of Band 7, center Band 7, high end of Band 6. Scientific tradeoff
is S/N considerations as we increase frequency, also perhaps spectral line
special cases, e.g. the CO spectral line. SG: It may be better to have
the plate at band 7, where the plate will enhance ALMA's ability in a poor
spectral window to do polar.ization, as we expect stability to be excellent
at the lower end of Band 6. Wright: At BIMA, we are 50-50 lines and
continuum. CW: Carlstrom polarizer should be considered, see
new Nature. Me: Steve Myers should consider whether this device
has ALMA applications. ACTION item. EvD: In terms of brightness,
350 GHz loses somewhat owing to higher system temperatures. One must
balance this. Butler: read Brogan's line list. SG: Can
we reach a decision? CW: will canvass ASAC and send in an email. We
need a summary sheet.
- Scientific benefits of a single antenna design. See material at
ImCal, above. (Stephane/Al). SG: Getting the same diffraction
pattern is the same antenna from a scientific point of view. This means
the layout of the quadrupod is the same with the same shadowing, etc. Should
not count on cancellation of first order effects. EvD: Polarization
and mosaicking could be most difficult. Mel: Different diffraction
patterns should perhaps enable us to disentangle effects. SG: Even
with the same design, as soon as they are spearated by a large amount, different
environmental considerations dominate. Many: Costs could be put
to better use scientifically. SG: We have to comply with many
different regulations to buy the antennas and it is not obvious that we can
do that and keep them exactly the same. EvD: Make sure this does not
impact science. MH: Phase errors won't cancel out, which I think is
a very important effect. Focus changes in different way provoke phase
errors. SG: But they will see different environments on long baselines.
BB: Much of mosaic science will occur in compact arrays, where
this will be important. SG: Focal length and magnification is specified
and are the same in both designs. MW: Backup structure however
will affect how the dish performs. BB: I think we would always want
one design. EvD: We are uneasy about two designs, but it is hard
to quantify. We should argue from the scientific point of view against
it. SG: Can everyone summarize comments please?
- Fast switching tests. See material at ImCal, above. (Mark Holdaway)
I made a plan to evaluate this on the antennas. First phase depends
only on the OPT. S/N will let you get tenth arcsec accuracy at 20 Hz
so we can see the antenna ringing. We will do fast switching between
two stars to evaluate this, tests to be done by July. Radiometric tests
probably cannot be done until Fall, November possibly. We also need
to verify that fast switching works interferometrically, but this will be
hard owing to lack of sensitivity on a single baseline. Some source
pairs can probably be used to evaluate this when we have a prototype interferometer,
perhaps by the end of 2004. EvD: Comments are to include bright sources,
such as the Moon, or bright planets radiometrically. MH: Yes, we will
do this for the Fall milestone. Further feedback sought. ACTION
Post current version.
Brief reports from subgroup leaders
The Calibration subgroup leaders held a telecon 2003-Feb-07; the
agenda
is available. Note particularly the calibration scheme proposed by Jack
Welch, and a
modification
of it proposed by Stephane Guilloteau. BB: Bock reported on dual load device;
they find a variation with temperature, being written into a memo. Martin-Pintado
presented 30m vane test plan, unclear if it can be completed given the 30m
schedule and the press of bolometer science. Final report expected
end of May. Welch proposal was discussed, ma de a proposal for a similar
system to ALMA. Butler described this in summary. Stephane proposed
a modification which addresses some problems. SG: Briefly, please
read the proposal. Note that the most difficult hardware in the proposal
is the waveguide switch between horn and receiver, perhaps not reuired for
stable receivers and calibration. We can use standard interferometric
gain, ensuring that decorrelation is controlled--this needs more elaboration,
particularly at high frequency. If we can do this, an antenna with
no primary, just horns attached to the receivers, becomes the calibration
antenna. Expensive, but not too much so; we could use the mount from
the unselected prototype. Development would be of horns, not of as
many components. EvD: How to test. SG: Not trivial, so follow
the 98 GHz experiment, probably just build the device and see if it works.
BB: Welch plans to build a 98 GHz version at BIMA. Also
at telecon we discussed a face to face cal meeting in Victoria in June.
Minutes
, including Action Items, from the Leiden meeting have been provided by Bryan
Butler and Ewine van Dishoeck.
The Configuration Group has delivered the configuration plan to the
Site Group, and surveyors have staked the Chajnantor Plain. Some small adjustments
to pad locations will be needed. Plan 'C' is available at
ALMAEDM
; Plan 'D' is in production.
Jérôme Pety reports:1. ACA simulations:
Tak Tsutsumi succeeded to make many simulations in Japan. There have
been difficulties in the visualisation part but that should be solved
since yesterday. I think Tak is now pretty much independant. So the week
spent in Grenoble has been very useful. Joined is a very preliminary
result : a
PS plot
of (image and uv plane) fidelities as a function of the SNR for m51ha.
From the fidelity definition, fidelity values will always be lower than
the SNR. To futher interpret this plot (and others simulations with noise)
we need to think about what are the SNR that can be expected for the simulated
scientific cases . These information are what I know and probably Hasegawa
san and Morita san will have more to say about that.
JC: Version D is produced, but perhaps needs more data on CBI avoidance zone.
Spec on separations of 15.15m is a decision of the JAO. AW thinks
this should not just be oral, but should be written down and justified. So
I made a new version, stating the specification, with a breakdown of how
this was arrived at. 5 cm construction tolerance for each pad, and
a component of 2.5cm per pad for post construction movement of the pad, on
top of the 15m separation specification of long standing. This document
will then be sent for approval, approval to signify acceptance of this specification.
Critical problem is if this is enough for pad movement after construction.
Is the spec enough for settling? I do need a number. If
an IPT thinks this insufficient, they should say so. We don't want
antennas colliding! SG: 15.15m is our position, but if after
construction we find this is not OK we can re-engineer pad location. EvD:
We can then close this issue. JC: Comments on file please. Also
other items, such as the CBI exclusion region. I had made a 100m zone
before, but Readhead has plans which impact a larger zone, perhaps 300m,
which hits six pad locations. To shift them all would cause significant
degradation of the beam. One pad is 101m away, where ACR wants to build
a new instrument. Move that pad 300m SSW then he can build his instrument,
and he has a 200m avoidance. Has there been a decision to accommodate
him? EvD: Talks ongoing. JC: For 200m, I only need to move
one pad. SR: PvdB and I had phone meeting, asked JC to evaluate
the impact on the configuration of exclusion zones, to understand the impact.
JC: I make comments on this. 300m is a big redesign, 200m
is a minor redesign. SR: Don't put this in the document, just
tell PvdB. JC: I will put reconfiguration sequence and try to incorporate
a discussion of more than sixty antennas-where would the extra four go? Also
the four TP antennas need to be discussed. There ends up being a lot
of detail in the spec document! SG: The document should have
the configuration with some info on how we use it but details are in a separate
document.
MH: Y+ array was happy last week, with 51 pads, with a strategy of optimization
of outermost configuration. However, 51 thought to be too many additional
stations (42 was last number). Now I am working on more shared stations,
with 44 now, 16 shared stations, 11 moves of four stations. I was considering
four moves per day for two transporters. (agreement from SG and JC).
Good, I am having success at getting low inner sidelobes now around
8-9%. But code has bug which lets one antenna run away. Under
51 antenna plan, sidelobes were lower (~5%). This is the price for
more shared stations. SG: A good compromise is to have some continuity
between arrays, so minimizing moves is a goal. JC: But we will pause
in the large configuration. SG: But it takes time to get there. EvD:
But seven pads makes a difference of a factor of two in sidelobe level for
the inner ones. MH: I have only been working on this in a few days.
MW: How do they look in intermediate configurations? MH: TBD,
but will probably be lower . Work continues. I hope to optimize
the intermediate configurations still, working with largest intermediate
configuration. I should have the single outermost configuration
within a few days. SG: This approach of optimizing longest baseline
is probably best. BB: High resolution is the driver (quotes ASAC report).
MH: I am comfortable doing things this way and will seek to do this
and share results, then we will ask the question of 'is this good enough?'.
SG: This is a sound approach.
The Site Characterization Group. Reports from Lars-Ake Nyman and Simon
Radford.
Riccardo Giovanelli visited Charlottesville on 2003-Feb-07. AW and SM attended;
AW reports:
Apropos the proposal from Mark and Angel:
Until talking to Giovanelli today I had not realized that he has an installed
weather station on Cerro Negro. They have data, retrieved two weeks ago.
Apparently the clock was not working properly, so it is not clear that the
timing is sufficiently accurate to compare to Chajnantor. Actually, I guess
it is clear that it is not--he said that he was able to determine times by
matching the diurnal profiles to the standard. One question which is not
answered is the relationship of the inversion layer, which normally sits
above Chajnantor, to the 5100m peak of Cerro Negro. He suggested that if
a tipper could be stationed there, we could determine whether the inversion
layer is at a fairly constant elevation above sealevel, or whether it is
at a fairly constant elevation above ground level, in which case Cerro Negro
might penetrate it and become a more interesting astronomical site. I suspect
the former but some discussion of this point would be interesting. He said
it takes at least 4.5 hours to get equipment up there after appropriate acclimatization.
I would think it not worthwhile to go to great lengths to compare the non-time-stamped
data with ours. He may be interested in launching radiosondes from lower
altitudes but AW did not talk to him directly about doing this. MH:
PLB is only 8 minutes upwind, not directly, so correlation there is not so
clear. SG: We also have good data from Sarazin.
Nyman report: Personnel will be shifted around. Juan Pablo is
a consultant, work with Rivera on the actual site under Bronfman guidance.
Our work package was cut drastically, including travel budget, limiting
times we can go to the site, only once per month now rather than preferred
once per two weeks. ALMA Memo ready to be submitted, error sources
on water vapor and delay from 183 GHz, originally from Guillermo Delgado.
Radford: Quiet data collection continues. Hygrometer needs
cleaning. No outstanding new data analysis, now running in maintenance
mode. A student is reorganizing interferometer data. Richer:
We have not done much yet but will review the memo.
The SSR held a telecon this month. Minutes of meetings can be found
at Robert's SSR page.
You have received a copy of a message from Steve Scott on behalf of the
SSR, requesting comments from the Science IPT on the ALMA data rates:
"The SSR has decided to drop its proposal for increasing the ALMA data rates
in favor of a new proposal that takes into account the capabilities of the
enhanced correlator. We would like input from the Science IPT before making
this proposal and hope that you will help us. I have attached a document (
EnhancedCorrelator.pdf
) that attempts to make some assumptions to distill out reasonable constants
in the data rate equation. If these are accepted at face value, the main
unknown in the average data rate is the number of channels. The second attachment
(DataRates.pdf
) is our previous request for a rate increase. It contains more background
material and examples that are not duplicated in the newer document. The
newer document contains recommendations, some with more emphasis than others,
as it seems easier to iterate from a given position rather than starting
from scratch. But feel free to examine any aspect of the problem. Please
don't hesitate to contact me or any SSR member if there are questions."
Jérôme Pety reports:2. Phase 2 of AIPS++/PdBI test: Phase
2 of this test is to test the calibration and imaging tools inside AIPS++
on different PdBI data sets. I have been asked to take the lead of this
phase. There are 7 testers (C.Wilson, D.Shepher, L.Testi, A.Baker, A.Coulais,
F.Gueth and myself) and as many data sets as testers. Almost everybody
has his/her data set and a working version of AIPS++. The final document
is due March, 10.
Lucas: Read material from Scott, what is the typical data rate? We
have a guess for the peak rate, 72 MB/s, from the proposed enhancement plus
our guess on the integration times. EvD: Science IPT will then
endorse the peak rate, and give a rec on the average rate, now 4 MB/s, proposed
to increase to 8 MB/s. SG: Benchmarking the last few weeks make
it obvious that the bottleneck is in storage, which is going down faster
than the cost of computing. Always advantageous to have higher data
rate, enriches the archive, I strongly support increasing the average data
rate. EvD: Comments please. Perhaps you would need to justify
in proposal going significantly over the data rate. ACTION: Send
comments to SG.
Imaging Gueth not present. Pety asks about ACA imaging (see
discussion above). Results available at or before ASAC meeting. ACTION:
comment on Pety's query.
Operations SG: Three people--Sramek, Dilva, Guilloteau--are
working to provide preliminary set of conclusions in June to project. First
top-down, reviewing existing documents, second bottom up by asking
IPTs for advice, such as reconfiguration strategy from this group. Impact
on RSCs will await a later phase.
Others/Outreach .
Date of next phone meeting According to the schedule the
next one should be on:
2003, March 11th 16:00 UT
