Report of the External Advisory Committee for the Australia Telescope MNRF Upgrade Project
S. Guilloteau, IRAM, France.
P. Napier, NRAO, USA.
R. Padman, Cambridge University, UK.
A. Young, CSIRO CTIP, Australia.
8 October, 1997
1. INTRODUCTION
The committee met during the period 1-5 September, 1997. During a visit to the Paul Wild Observatory we inspected the Australia Telescope Compact Array. Included in the tour were the areas of the railway track where the new pads and N-S spur will be added, the transporter section of the antenna which will be modified to allow transition to the N-S spur and the antenna receiver turret where the new high frequency receiver will be located. Oral presentations were given by AT staff on all aspects of the Upgrade Project and detailed discussions were held on a number of key technical topics including feed array technology, interference excision, radiometric phase calibration and receivers.
We wish to congratulate the ATNF on successfully proposing and receiving funding for the Upgrade Project. The project represents a wise investment to increase the scientific return from a major national instrument. The Australia Telescope is already very productive, and the upgrade will greatly extend the range of scientific problems that can be addressed with this instrument from its valuable southern hemisphere location.
Much preparatory work has already been done, and on the basis of what we were shown we believe that the project is very well-founded and has every chance of achieving its technical and scientific goals. While we have noted some minor reservations and queries in the remainder of this report, we stress that for the most part these have been drawn to our attention by the project members themselves. None of them appear to be "show-stoppers", and we are certain that the relevant expertise is available to the project to overcome any problems.
2. MANAGEMENT
The project has a good management plan with clear statements defining the scope of the project and the responsibilities of key personnel. The only concerns that we have are to do with the amount of manpower available to accomplish the project. It is particularly unfortunate that the beginning of the project coincided with the need for a reduction-in-force for the ATNF as a whole. This means that the ATNF will have to carry out the upgrade project and continue its normal operations and maintenance functions with little increase in total observatory manpower. We think that it would be wise to try to identify areas where operations or maintenance support could be temporarily reduced during the year or two of peak upgrade activity. One example of such an area could be, for example, the frequency with which the compact array configuration is changed. An area where we are particularly concerned about having key people available to concentrate on the upgrade is the effort to resurface and thoroughly test the Mopra antenna. This activity is particularly important to the upgrade project because it must be finished before it is known with complete certainty that the solid surface will be installed on all antennas.
3. BUDGET
The project has a clearly defined budget but we did not discuss it in enough detail to be able to comment on the adequacy of the funds allocated for the various parts of the project. We are concerned that there is no specific allocation for contingency in the budget. At this early stage in the project it would be wise to have some contingency funds and urge that the various parts of the project be prioritized so that some parts with lower priority can be left to later in the project. In this way, if more funds for high priority items are needed, they can be found with the minimal possible impact on science capabilities. For example, something that would have minimal impact on science capabilities would be the outfitting of only two, rather than all five, antennas in the compact array with the transporter modifications needed for operation on the N-S spur.
4. EXTRA STATIONS AND THE NORTH-SOUTH SPUR
Improving the UV coverage is essential for high frequency observations, but also for wide field imaging at lower frequencies. The N-S spur is mandatory to provide good UV coverage at high frequency where observations at low elevations are not possible. However, it is also justified by projects requiring very compact configurations, such as studies of the Sunyaev-Zeldovich effect at 22 or 43 GHz. The location chosen for the N-S spur appears suitable and has the advantage that space exists for a longer spur in the future.
We consider that observations with the antenna on the rails rather than on station pads would be problematic in terms of pointing and baseline length stability, and may also cause serious safety issues. Accordingly, we recommend instead the addition of 4 stations on the E-W track, and to proceed with the construction of the N-S spur as civil work scheduling allows.
5. SOLID SURFACE UPGRADE
Whilst we do not think it very likely that the replacement of the perforated panels with solid panels will cause undesirable wind performance degradation in the antennas, the absence of any careful aerodynamic, structural or servosystem analysis of the change is a concern. Apparently the perforated panels were thought to be necessary when the antennas were originally designed and it was not made clear to us why this has changed. Of concern would be the survival wind performance and the antenna pointing and phase stability performance in moderate winds. Note that if there is any
significant increase in wind loading as a result of removing the perforations, it is at the edge of the dish where the area and resulting increase in turning moment is largest. Of most concern is probably the pointing performance of the antenna in moderate winds. Although 3 mm observations will only be made in calm conditions, good pointing is also essential for 21 cm mosaic observations which are often done in windier conditions. We recommend that, as soon as possible, an effort be made to do
the necessary calculations to give high confidence that there will be no wind performance problems. An additional need for the calculations is to confirm that it is valid to test the modification on the Mopra antenna, which has a different design to the compact array antennas. The modifications at Mopra should be made as soon as possible and should be accompanied by careful before-and-after tests. Additionally, although some measurements have already been made to confirm that gravitationally driven changes in subreflector position and surface shape are acceptable, these measurements should be repeated at 3mm wavelength with the solid surface in place.
6. LO UPGRADE/BACKEND
The only aspect of this area where we have any reservations is the conversion of the LO reference distribution system to work over optical fibre. This is the first application of fibres for reference distribution in a mm radiotelescope where there is the need to disconnect frequently. Whilst we support this work, we wish to advise against underestimating the effort in proving such a system. We suggest that the ATNF try to benefit as far as possible from the experience already gained by JPL, BIMA, OVRO and SMA. The upcoming URSI meeting in the US could also be quite useful.
7. RECEIVER SYSTEMS
The MMIC approach to the construction of the 3-mm frontend seemed reasonable to us, and potentially very exciting -- if successful the technology would be of value to the entire radioastronomy community, with major implications for the next generation millimetre-wave arrays for example. This may warrant greater investment than is available just through the Upgrade Project, perhaps in collaboration with CTIP. There is an obvious risk of high costs with this approach but in the event that MMICs are not available on the required timescale it is likely that a backup plan would exist in the form of discrete InP HEMT amplifier designs developed at other observatories.
The design of the receiver should NOT be compromised to provide coverage of the upper end of the 3-mm band (100 -115 GHz). In particular we do not feel it is worthwhile to provide a second (high frequency) channel for this band. This will provide considerable simplifications and cost savings over the initial design discussed during our meeting. It is reasonable to expect that usable performance in the 100 -115 GHz band will be obtainable in the future by broadbanding the 85 - 100 GHz band.
The outline optical design for the high-frequency receiver package appeared to show possible blockage of one or more feeds, and we would encourage the receiver group to check the blockage with a gaussian-optics calculation before proceeding much further.
It is currently proposed to implement rapid frequency changes (between the 12mm water radiometer channel and the 3mm/7mm channel) by translating the entire feed package. We were somewhat nervous about this, and ask the receiver group to look again at whether this switching can be achieved in the receiver fore-optics or by rotating the receiver package (or by any other means). Again, this problem will be simplified by abandoning the second 3-mm receiver.
Finally, we note that use of the 12-mm receiver for continuous water vapour monitoring implies that the sounding beam will be off-axis from the observing beam. We endorse the ATNF's plan for a near field analysis of this off-axis beam to confirm that it will be useful for water-vapour monitoring.
8. ATMOSPHERIC PHASE CORRECTION
On-going experience with other mm arrays has shown the value of atmospheric phase correction systems. Given the low altitude site and the expected water vapor content, we recommend that the AT concentrate on the development of a 22 GHz water line monitoring system. The cooled astronomy receiver should be used for this purpose if possible, which implies good stability and more detailed investigations of the spillover properties, but also offers the possibility of using narrower filters.
Similar systems would be valuable for VLBI experiments, although these have not been costed in the current development plans.
The atmospheric correction will require dedicated software to be used in real time, either in the correlator or LO system. The ATNF should also explore the possibility of using the IRAM scheme, in which corrected and uncorrected data are both produced and saved for off-line processing.
9. VLBI UPGRADE
The committee is happy with all aspects of the VLBI upgrade at this time. We do however wish to commend the ATNF for two very important decisions. The ATNF appears to have made a good choice for the purchase of very cost effective hydrogen maser clocks and it is pleasing to finally see an observatory implement digital filtering technology. As an additional comment we believe that it would be desirable to look for ways that may allow 22GHz atmospheric phase correction to be implemented at all VLBI antennas.
10. SOFTWARE/ACC
We were concerned by the parlous state of spares for the current Antenna Control Computers (ACC). Replacement of this system with modern maintainable hardware and software must have very high priority. We emphasize the importance of using standard hardware and programming languages (as far as possible) to minimize the effort required when this next generation system needs replacing in its turn. We note that a project group of 2-3 people has been formed with the defined goal of producing replacement ACCs by late 1998. We stress the importance of this project and recommend that the ACC aspect of the project be, in general, given priority over other developments and operational responsibilities with which the group is charged
In general the plans for software upgrades seem to be reasonable, although we were left with the feeling that there was rather little manpower available for the work that needs doing (and to some extent that manpower is already committed to other aspects of the upgrades project, or to day-to-day operations and maintenance). It is important to prioritize the various projects to cover the possibility that not all of them can be carried out.
11. ARRAY TECHNOLOGY
Although it is clearly too early to apply any of the new ideas in array technology to the MNRF upgrade at this time, we encourage continued research within the ATNF. Both focal plane and aperture plane technologies are likely to be very important for the future and the ATNF is well placed to work in these areas, especially if it could benefit from the connection to CTIP.
12. INTERFERENCE EXCISION
Interference excision is likely to be very important for the future of radioastronomy as the number of LEO satellites increase. The Parkes multibeam system provides the ATNF with a unique capability to develop this new technology and we recommend that research be continued in this area. We also encourage continuing involvement in frequency management issues and a greater AT involvement in the IRIDIUM interference tests.
13. 43 GHz RECEIVER
The 43 GHz receiver is being designed into the new short-wavelength dewar, but there are not adequate funds to include its implementation in the upgrade project. There is a lot of excellent science to be done in this band and we encourage the ATNF to vigorously seek funding for this additional enhancement. This would be an excellent project for funding by a potential foreign partner.
14. OBSERVING AND SCHEDULING
Flexible (dynamic) scheduling will be necessary for high frequency observations, and should be implemented with the array upgrade. This is an important issue, which represents a significant change from current practice. We recommend the study of the possible operating modes with the ATNF Users committee.
The upgrade itself will have impact on observations, during the civil work for the additional stations and N-S spur, during the antenna resurfacing, and during the local oscillator system modifications. The influence of these items on the observing schedule should be investigated.
15. SCIENCE STRATEGY
Several aspects of the upgrade seem to be predicated on successfully resurfacing the full 22-m apertures of the ATCA elements for operation at 3-mm. At this point it is not absolutely obvious to us that the resurfacing will achieve high enough surface accuracy and/or pointing performance for the telescope to be viable at the highest frequencies. The proof of this capability must await the resurfacing and testing of the Mopra Antenna. If only the present solid surface (15-m diameter) of each dish can be made to operate at 3-mm the range of science that can be done will be somewhat restricted. It is not clear to us at this point whether the science that can be done with an array of 15-m dishes with a minimum 30-m baseline is sufficient to warrant the effort of equipping the telescope with 3-mm receivers.
In the event of NOT being able to utilize the full dish diameter at 3-mm, it may well be better to concentrate on the 7-mm system, where there would still be 5 (or even 6) dishes operating at the full aperture and sensitivity (and a good deal of very useful science to be done). There are thus two possibilities: if there is sufficient 3-mm science to warrant operation with the 15-m useable apertures, then the project can proceed as currently planned; if there is NOT sufficient science, then construction of the 3-mm receivers should wait until the full dish performance at 3-mm can be verified. At present there are different opinions within the ATNF staff and the ATNF user community with respect to the value of this "reduced" 3-mm science: further consideration is clearly required.
Clearly, the upgrading and testing of the Mopra antenna should be brought forward as much as is practical since so much of the further work depends on the results of these tests.