ASAC members visit the future site of the center of the ALMA
10 September 2001.
11-12 September 2001
Cerro Calan, Santiago Chile
Draft Minutes vers 4 Oct 2001
E. van Dishoeck welcomed all participants to this meeting, particularly those who are not members of the ASAC.
Participants: J. Baars, R. Bachiller, A. Benz, G. Blake,
R. Booth, L. Bronfman, R. Brown, Y. Chikada, P. Cox, R. Crutcher, N. Evans,
Y. Fukui, B. Glendenning, P. Gray, S. Guilloteau, M. Gurwell, E. Hardy,
T. Hasegawa, D. Hofstadt, R. Kurz, R. Lucas, J. Mangum, D. Mardonis, H.
Matsuo, S. Meyers, K. Morita, N. Nakai, L.-A. Nyman, J. Payne, S. Radford,
M. Rafal, J. Richer, S. Sakamoto, P. Schilke, P. Shaver, K. Tatematsu,
E. van Dishoeck, E. Vera (part-time), M. Walmsley, J. Welch, C. Wilson,
A. Wootten, S. Yamamoto
1. Project Update
R. Brown gave an overview of progress since the last face-to-face
meeting. He described the evolution from a two-way to a three-way project.
It is hoped that construction can begin in 2002, and that completion can
be achieved in 2010. Issues include the implications of delayed Japanese
involvement in construction, permission for use of the site, project schedule
and resources and the establishment of a centralized management structure.
ASAC issues include the project scope, which affects its cost and schedule,
and the possibility of incentives for technical performance during construction,
and for scientific staffing in the early operations phase.
2. Towards a Three-Way Partnership
T. Hasegawa described the funding situation in Japan. The MEXT budget request to the Ministry of Finance includes 929 million Yen (about $7.7 million) for ALMA R & D, but the main construction budget was not included. Very positive comments about the project have been made, and it is still hoped that the government may "express its commitment" to the project. In a letter to the E-ACC in August, N. Kaifu said that a major step is a budget that includes a 12m prototype, and that the decision by MEXT practically means the start of the project, but approval of the overall construction budget could be significantly delayed. NAOJ is spending about 200 million Yen per year on the project, and plans to continue this. Planning for the 3-way project scope and task division continues as before.
R. Kurz gave current cost estimates for the 3-way project.
The total resources in the –10% model including Phase 1 amount to $870
million (2000). Counting just the amounts from September 2001 onwards,
the resources are $816 million, and the cost estimate including enhancements
is $815 million. Thus, the project is feasible with the projected resources.
This confirms the conclusion reached in June 2001.
3. AMAC Report; Management and Structure of the Project in Chile
M. Rafal gave a brief summary of the history and charge
of the AMAC. It presently includes representatives from the US and Europe,
and will be augmented to include representatives from Japan. At its first
meeting in June 2001, the AMAC expressed its positive impression about
the project, recommended the establishment as soon as possible of an international
project office with a director with well defined authority, stressed the
need for a rapid definition of a legal organization in Chile, accepted
the IPT structure, pointed out that the prototype antenna difficulties
have project-wide implications, and said that the procurement model for
the antennas needs additional study.
ASAC members visit the site characterization facilities on Chajnantor 10 September 2001.
4. Site issues; Discussion of Visit to Chajnantor
S. Radford gave a summary of the US, European and Japanese site testing. Concerning the comparison between Pampa la Bola and Chajnantor, he said that there is a small but significant difference in favor of Chajnantor. L.-A. Nyman discussed the water vapor radiometer results along with the interferometric phase measurements. He noted that the ionosphere can affect the latter but not the former, complicating interpretation. He commented on the long-term climatological changes from 1948 to 2000 (memo 333); the annual Bolivian winter effect is clear, but there is no long-term trend. Satellite studies indicate that the Chajnantor area has the lowest water vapour content, in a comparison of several peaks.
S. Sakamoto discussed the comparison of Pampa la Bola with Chajnantor. The opacity and seeing are better at Chajnantor, and the wind is similar. His conclusion was a preference for Chajnantor, especially for the high frequencies. N. Evans asked about cloud cover, and referred to the experiences related by A. Readhead. R. Kurz said that D.A. Erasmus will present a proposal for study. E. van Dishoeck asked for comments about the site, following the visit. Overall the views were quite positive. J. Richer asked about future plans. S. Radford and L.-A. Nyman said that the studies would continue. R. Brown commented that initial facilities should be there by January 2005; hence they should be started by January 2003, with initial design and engineering studies starting in January 2002. Formal agreement with Chile for the site would be required by January 2003. D. Hofstadt made reference to the possibility of a "direct link" (ie, a dedicated road to the site from the Operations Support Facility (OSF) to be located some tens of km outside of San Pedro de Atacama).
Concerning the configuration concept, A. Wootten said
that the "self-similar" concept had been agreed some months ago, and that
this concept would extend out to the 14 km array, to provide a more elegant
way of making the transition between the compact and extended configurations.
These concepts will be finalized over the next several months, along with
the general configuration study. M. Gurwell asked about still longer baselines.
A. Wootten replied that there was not much possibility beyond the 17 km.
J. Richer suggested that some antennas could be located at the OSF, at
least for low frequencies (eg. below 350 GHz). This was considered a possibility,
although there may be limitations due to the fiber connections. S. Radford
and L.-A. Nyman asked what site issues the ASAC would like to be studied.
G. Blake suggested conditions for the best possible phase correction, and
C. Wilson suggested the cloud cover issue. A 10 micron camera could be
installed, and S. Radford said that the monitor data had already been assessed.
5. The Atacama Compact Array: Design and Simulations
S. Guilloteau reported that, in response to the "emergency actions" decided in June 2001, a "Tiger Team" had been established, and had simulated calibration errors and studied deconvolution methods. At IRAM over 700 CLEAN simulations had been performed, and the detailed results are on the web. MEM results have been obtained by M. Holdaway and K. Morita, and they are very similar. The main conclusion is that the ACA does give a significant improvement, and in particular it improves the robustness of the results, especially for images with extended structure. The image fidelity is improved by 30-100%. While ALMA + single dish fails on some images, the ACA gives robustness with respect to image type, pointing errors, primary beam errors etc. Importantly, it does not significantly increase the computing time. It brings an insurance on the quality of the result. It also provides a good high frequency array, which can be very complementary to ALMA. The 12 x 7m ACA plus 4 x 12m single dish model is sufficient for the purpose. Remaining work includes the science analysis and some other technical issues.
The assumptions and implications are as follows: the ALMA array has only 60 12m antennas; two correlators are required (60-antenna ALMA, 16-antenna ACA); the ACA and 4 x 12m antennas operate at the same frequency and are cross-correlated together; ALMA observes 25% in mosaic mode; and the ACA 12 x 7m antennas observe 100% of the time. The case presented was regarded as very convincing by the ASAC.
K. Morita presented his results including pointing errors and thermal noise, using the MEM technique. The ACA gives a clear advantage in the noise-free case. This was not so clear in the noisy case – the average S/N of the ACA visibilities had to be > 5 to obtain better results. However, MEM is weak in dealing with thermal noise, compared with CLEAN. S. Guilloteau said that the thermal noise case would be checked with CLEAN very soon. A. Wootten said that M. Holdaway's results are qualitatively similar to those from IRAM, and details will be available soon. He also repeated T. Hasegawa’s comment made at the Florence meeting that it is important for the wider (non-expert) community to have high fidelity images from ALMA. E. van Dishoeck asked whether there were further requests for the Tiger Team. R. Crutcher said that the turbulence case is very important. J. Richer mentioned S-Z images, and S. Guilloteau replied that this case was already effectively done with one of the images. On behalf of the ASAC, E. van Dishoeck thanked the Tiger Team for its great efforts.
The possibility of using the 8 x 6m SMA instead of the
ACA was briefly discussed, as a possible backup proposal. M. Holdaway showed
in his memo of September 6, 2001 that a large part of the improvement achieved
using the ACA could also be achieved using the SMA. M. Gurwell summarized
the status of the SMA, and said that there was interest amongst the SMA
group in the possibility of moving the SMA to Chajnantor, to provide a
long-term future. Pros include the high surface accuracy of the antennas,
the good polarization purity, the fact that the SMA already exists, and
its expected lower cost. Cons include the small secondary and primary beam
hole, the complex optics, the different antenna design, the different dewar
package, and the different polarization properties. Also, there may not
be enough baselines and sensitivity. R. Crutcher asked about the political
implications, and what the SMA partners would want in return. It was agreed
that this remains at the moment just an informal possibility.
6. Discussion of ACA, Science
A. Wootten gave an overview of the various science cases considered: M51, the M31 HII region, an extended protostellar envelope, a turbulent cloud, and a debris disk. J. Welch mentioned the example of HL Tau, using BIMA and a 12m dish, and a top-hat model. The ACA would make significant improvements. ALMA would be fine for a 10:1 fidelity index map, but several science problems require better. R. Crutcher commented that good fidelity is essential for polarization. K. Tatematsu discussed a cluster model with 11 protostellar cores. M. Gurwell agreed to help M. Yun in his analysis of M51. P. Cox said that some of the science cases are not as compelling as the statistical results shown by S. Guilloteau. It was suggested that the final report would give a list of the science cases, highlighting those requiring the ACA, and giving a few dramatic images to illustrate the improvement.
7. Report from Nobeyama Correlator Meeting
Y. Chikada gave a report from the correlator meeting of 5-7 August. The 2G correlator would provide high sensitivity, flexibility and resolution, and it would be able to accommodate the ACA. The increase in sensitivity by 9% would pay for the 2G correlator. The groups are working towards a "unified design", and the areas of common interest have been listed. Discussions are taking place with industry, with quotations expected within a year. A unified design is expected in 2002, and a prototype in 2004. The best final correlator architecture will be selected and the best technical options will be adopted – the groups are working in close cooperation, not competition. A joint review is planned in mid-2002. R. Kurz said that the baseline correlator may be converted into the ACA correlator. P. Shaver asked whether the baseline and 2G correlator timelines have been optimized relative to each other, as at the moment they seem rather close. They have not been optimized, but this could be considered. A. Wootten compared the 2G concepts with the EVLA correlator (WIDAR) and noted that both of them will be on the frontier of radio spectroscopy at the end of the current decade.
R. Bachiller gave a summary of the revised correlator
guidelines: ability to accommodate the ACA, number of channels > 8000 (2
MHz resolution over the full band, with options for binning channels for
further processing), 3-bit digitized format, resolution to 5 kHz or less,
16 sub-bands, 4 independent sub-arrays, and shortest integration time 1
msec (continuum). S. Guilloteau said that the ACA should have its own independent
correlator, and questioned whether 4 sub-arrays was sufficient. R. Bachiller
said that the project scientists should now work on these guidelines to
produce actual specifications. E. van Dishoeck concluded that the proposed
guidelines are now accepted by the ASAC.
8. Recapitulation of Second Generation Correlator Science
As preparation for the prioritization process, the science cases for the different enhancements were presented and reiterated. T. Hasegawa presented the scientific case for the second generation correlator. The second generation (2G) correlator would give higher sensitivity, simultaneously provide high spectral resolution and a wide wavelength coverage, and give the possibility of variable frequency resolution within a single band. The gain in sensitivity compared with the baseline correlator would be equivalent to the cost of the new correlator. Scientific applications include QSO absorption lines, the forests of emission lines in galactic sources, and the simultaneous observation of broad and narrow lines in AGN spectra (a very wideband coverage is necessary to obtain an accurate baseline for the observation of the broad AGN lines).
9. Recapitulation Bands 1, 4, 8, and 10 Science Case
E. van Dishoeck reviewed the science drivers for the frequency
bands. General issues for the various bands include continuous redshift
coverage for CO and [CII], the dust emission spectrum, a wide range of
physical diagnostics, and astrochemistry for both the cold and warm gas.
Bands 3, 6, 7, and 9 are in the baseline project, bands 1, 4, 8 and 10
are proposed enhancements, and bands 2 and 5 have the lowest priority.
P. Schilke suggested that some of the existing bands could be widened,
but there are technical trade-offs, and E. van Dishoeck said that it would
not be a simple matter to change this now, and several requests are still
pending. The science cases for the individual proposed bands were summarized
by J. Richer (band 1), P. Cox, (band 4), and S. Yamamoto (bands 8 and 10).
The value of photometric redshifts was discussed in this context, and some
doubts were expressed about this as a major science driver. The science
case for the Sunyaev-Zeldovich effect implies that Band 1 receivers need
to be present on the ACA. S. Sakamoto made the comment that, while the
site was excellent, if there was no receiver in a given band the effective
opacity in that band was infinite. J. Baars commented that all of these
bands were in the project from a long time ago, and so should not be presented
as "enhancements", but as "deferred capability".
10. Closed Session
A closed session was held in which the ASAC considered the E-ACC requests for (1) the science cases for the enhancements, and (2) a final prioritized list of enhancements. The discussion of possible guaranteed time in return for technical contributions was postponed to a later meeting. In this session it was agreed that T. Hasegawa would be able to vote as an ASAC member, but that in future he should become ex-officio and Japan should have one more ASAC member.
The conclusions of the closed session were announced later in the meeting. The ASAC had reached unanimous recommendations on the priorities, based on scientific merit, as follows (in groups of two):
The beginning of the open session on the second day started
with a period of silence to remember the victims of the terrorist attacks
that had taken place in the U.S. on the previous day.
11. Receivers
J. Payne presented the report on receivers. A draft of
the full report is presently with the E-AEC; it will be revised, and then
given to the ASAC. The report contains an overview of the evaluation receivers,
the front end D & D progress, and the status of the various bands and
components. An overview was presented of the overall concept and various
possibilities. The very promising developments on the photonics LO were
mentioned, although the overall LO development is still not very advanced,
especially at the higher frequencies. Other areas of concern included the
total power stability, the schedule and need for more manpower, and the
shortage of qualified receiver engineers. The plans for mass production
and the role of integration centers still need to be developed. J. Payne
announced that an ALMA Tucson Electronics Division (ATED) has recently
been formed. R. Booth pointed out that there are at the moment parallel
developments, for example at IRAM and Onsala on band 7, so better definition
and control are required. In response to a question from R. Crutcher, J.
Payne said that there would be ALMA production receivers on the test antenna.
T. Hasegawa asked about the DSB/SSB issue. Bands 8-10 will be DSB, but
the plans for SSB at bands 7 and below will be an issue for the receiver
group to consider, following the ASAC recommendations.
12. Calibration Issues
Wootten gave a report on calibration issues. Recommendations from the Cambridge meeting include: the need for an end to end description of the calibration system, the need to prototype as much as possible, and the need for a coordinated calibration group, in collaboration with the SSR group. The goal of 1% calibration accuracy in the mm bands was endorsed, but 3% may be more realistic at the highest frequencies. S. Guilloteau commented that 3% would be adequate at these frequencies, but that 10%, for example, would be very bad for imaging. S. Yamamoto stressed that these numbers (1% and 3%) are for the very best atmospheric conditions. The ASAC adopted the recommendation to relax calibration accuracy to 3% at submillimeter frequencies under good conditions. It emphasized that the calibration goal should be that the accuracy is limited by atmospheric effects, rather than by instrumental issues. Further studies and issues include the standard gain horn approach, the semi-transparent vane, simulations to guide the studies, demonstration of the photonic system, atmospheric modelling, and guide star pointing. The issue of polarization calibration was deferred. A report is due to the E-AEC, including a list of action items. A calibration group will be formed.
C. Wilson reported on the 183 GHz water vapour radiometer
systems installed on the SMA. There is a good correlation of the two systems.
It is intended to conduct a long campaign, to acquire experience on real
telescopes. The next step with IRMA is to test the system at Chajnantor
in Feb/Mar 2002; previous
tests compared to JCMT sky dips have given good correlations in water
vapor emissions and inferred water vapor column abundance. J. Welch commented
that the issue of anomalous refraction should be left for later study.
Matsuo noted that simultaneous 220/650 GHz monitoring of the atmosphere
provides good sensitivity to frequency-dependent effects of particulates.
One or more Fourier Transform Spectrometers may also serve this purpose,
Schilke noted.
13. High Speed Data Transmission in Chile
E. Vera (U. of Chile School of Engineering) gave a presentation
on high speed communications developments in Chile. High speed networks
are developing rapidly: a national university network, broadband networks
for astronomy in Chile, and high speed connections to Japan, the U.S.,
and Europe. He discussed possible solutions for ALMA connectivity. Following
his presentation it was reported that, in the near term, it is planned
to operate ASTE remotely from Tokyo.
14. Report Operations WG
N. Evans summarized the report of the Operations Working Group. The assumptions made are the following: (1) non experts should be easily able to use ALMA, (2) adequate information should be available on previous observations, (3) dynamic scheduling should match the observing conditions to the program, (4) calibration must be system-wide and "automatic", (5) there should be maximum flexibility consistent with smooth operations, (6) Regional Data Centers (RDCs) may supply data, computing power, and advice, and (7) the data should be made public in a timely fashion.
The following recommendations were made:
Several of the proposed recommendations were discussed.
Several ASAC members disagreed with a "standard" longer proprietary period
for students, as it depends on the individual case. A more flexible arrangement
will be recommended. The issue of one or several TACs requires further
study, but will likely be decided by the partners represented on the E-ACC.
The flexibility to be allowed for approved programs in phase II was discussed,
as was the balance of the dynamic scheduling priorities. Eavesdropping
and breakpoints were considered to be important; an open question was whether
or not to allow breakpoints in real time to be inserted. The concept of
the single SOC was agreed, and its location in Chile as a logical choice;
L. Bronfman suggested that, in that case, it could possibly be linked to
the Chilean RDC.
An extensive discussion took place on the RDCs. It was questioned whether we need 3 RDCs. S. Sakamoto suggested that they could be optional, and not inside the project. On the other hand, P. Schilke said that, if they are to be involved with software development, they should be under project control. R. Booth said that they should be part of a unified project, rather than too independent, and expressed concern about software development being taken away from individual institutes. G. Blake suggested that the project could support a minimum necessary core level of activity in the RDCs, and the individual regions could then enhance their local RDC as they wished. T. Hasegawa said that it was important to have some variety provided by the different RDCs. It was proposed that the name be changed to "Regional Support Facility".
E. van Dishoeck thanked the operations working group for
producing such a comprehensive discussion document on short notice.
15. Software
B. Glendenning discussed AIPS++ and ALMA. The SSR group is preparing the offline package requirements, and an "audit" of how well AIPS++ meets these requirements will be done. This will form the basis of a plan for AIPS++ to meet the requirements. An ALMA representative will join the AIPS++ consortium. A. Benz remarked that the project was supposed to look at a wider range of possibilities.
R. Lucas mentioned two documents: a pipeline requirements document, and an off-line analysis requirements document; feedback will be requested on both. There was an overview of simulators and their applications. Concerning the ALMA/AIPS++ test, an end-to-end analysis of Plateau de Bure data inside AIPS++ is taking place. Its duration is 9 months, and will end in April 2002. A phase 3 is also planned, in which simulated ALMA data will be used. The pipeline and off-line analysis documents will be ready in about one month. The audit will take place by spring 2002.
E. van Dishoeck asked for comments on this overall approach,
and whether adequate management and manpower will be available. A. Benz
agreed that it was reasonable to start with the requirements and then address
solutions. He also argued for the definition of a smaller core program
for the initial development of the software and user feedback. B. Glendenning
said that the off-line software is already being supplied from outside
the ALMA project; the pipeline software would be supplied inside. C. Wilson
commented that the pipeline and off-line packages may not be independent.
R. Crutcher asked about science software, and B. Glendenning replied that
image analysis is not included in the ALMA tasks. Concerning the simulator,
there were differences of opinion. N. Evans would like the simulator to
do as much as possible (e.g. running the model through the same
simulator as the data), but J. Richer felt it was dangerous to take this
too far. In any case there was general agreement in going to level 3 in
the definition of the simulator, but that the extent of the options at
level 3 need further study.
16. Reports on Prototype Antennas
J. Baars. J. Mangum and Y. Chikada described the status
of the three prototype antennas. The U.S. prototype was expected in the
first half of 2002. J. Mangum said that the foundations at the VLA site
were now complete. Erection of the prototype will start in January, and
the completed antenna will be turned over to the project on 23 April 2002.
In Europe, modifications and changes have caused the expected delivery
time to slip to the end of 2002. Y. Chikada reported that the Japanese
12m prototype would be delivered to the project in April 2003. It will
have a surface accuracy better than 20 micron rms, pointing 2" (global)
and 0.6" (local) (in a 10 m/s wind), and a total weight of less than 80
tons.
17. Status Test Interferometer
J. Mangum reported on the test plans and status. Following
acceptance of the Vertex antenna in April 2002, there would be holography
(July), optical tracking and pointing tests, installation of the evaluation
receiver (Nov/Dec), and radiometry tests in the winter. There will then
be time to do the same tests on the European and Japanese prototypes. The
antenna technical evaluation group will be comprised of 8 persons. There
will probably also be an evaluation committee to make the final decisions.
18. Status Polarization Issues
J. Mangum said that there will probably be four tests
for polarization: single dish polarization beam maps, interferometer polarization
beam maps, holography transmitter beam maps, and off-axis feed studies.
R. Crutcher commented on the need to evaluate polarization stability. E.
van Dishoeck said that someone was needed to concentrate on polarization,
perhaps within the calibration group.
19. The ASTE Project
S. Yamamoto described the ASTE project as a precursor
to ALMA, and a test bench. R. Kawabe is the project director. A description
was given of the instrumentation, the proposed infrastructure (ASTE will
be surrounded by containers, like the CBI), and the timeline: shipment
in October, erection beginning January 2002 and completed Mar/Apr 2002,
first photons April 2002, and the high frequency receivers installed by
July 2002. Holography will be done using a transmitter on one of the nearby
mountains.
20. The APEX Project
P. Schilke summarized the APEX project. The antenna will
be a modified copy of the Vertex prototype, which will include nasmyth
foci and a nutating subreflector to be provided by Vertex. The surface
accuracy goal is 18 micron rms. It should be completed on site by March
2003, and then there will be 6 months of commissioning. The suite of instruments
includes two state-of-the-art bolometer arrays, one at 850 micron (1000
channels) and one at 350 micron (117 channels), a suite of single pixel
line receivers, and a heterodyne array receiver CHAMP+ at high frequencies.
21. Preparation of Reports
Following meetings of the sub-groups, the group leaders presented outlines of their draft reports for comments and discussion.
E. van Dishoeck then gave the following schedule for the completion of the ASAC report:
As soon as possible: copies of viewgraphs, powerpoint presentations to E. van Dishoeck
Sept. 24: First drafts sections to E. van Dishoeck
Oct. 3: First complete draft distributed to ASAC
Oct. 9: Comments from ASAC members to E. van Dishoeck
Oct. 15: Final report to E-ACC
The schedule for the Science Cases for enhancements is
as follows:
Sept. 29: ACA science case to E. van Dishoeck
Oct. 8: Complete draft distributed to ASAC
Oct. 11: Comments from ASAC members to E. van Dishoeck
Oct. 15: Final report to E-ACC
The report sections are as follows:
Project status, management: from minutes
Site issues, configuration: Booth, Bronfman, Sakamoto
ACA simulations, science: Cox, Crutcher, Hasegawa, Welch
Correlator: Bachiller, Scoville, Yamamoto
Operations: Evans, Fukui, Wilson
Software: Benz, Gurwell, Tatematsu
Calibration, WVR: Matsuo, Richer, Wilson
Polarization: Crutcher
Receivers: Blake, van Dishoeck, Nakai
Antennae, test interferometer: Fukui, Walmsley, Welch
ASTE, APEX: Kawabe/Yamamoto, Schilke
22. Election of new Vice-Chairperson
E. van Dishoeck announced that the European members of
the ASAC have elected P. Cox as the next European vice-chairperson of the
ASAC. G. Blake becomes chairperson of the ASAC, and Y. Fukui remains Japanese
vice-chairperson and next chairperson-designate.
23. AOB
E. van Dishoeck asked that the liaisons give feedback
on this meeting to the technical team leaders when they return home, to
provide greater communication within the project.
24. Next Teleconference and Meeting
The next teleconference is on WEDNESDAY October 10, at the usual time.
The next face-to-face meeting is tentatively scheduled
for 18-22 March 2002 in Japan.