Pending problems in ALMA Calibration 1) Dual-load in the subreflector - Tests at BIMA have revealed a coupling coefficient varying by up to 30 % as a function of frequency, and by about 10 % as a function of time. It also depens on elevation. It is thought that these variations are due to standing waves between the receiver and the edges of the loads in the subreflector, which vary as a function of elevation because of the change in focal length of the antenna - No design exists for the ALMA antenna 2) Semi-Transparent vane - preliminary tests have been carried out at the 30-m. Differences in absorption coefficient of about 20 % have been found between several measurements. The only material tested so far is a dense polystyrene foam. Other materials have been received and will be tested later. The cause of the variations is unknown. 3) Photonic emitter for bandpass calibration - Initial laboratory tests shown amplitude (power) variations of up to 100 % as a function of frequency and time. 4) Standing waves - Aurore Bacmann has evaluated the level of standing waves expected with the ALMA antennas due to diffraction at the edges of the sub-reflector. This is around 1 % at 3 mm. It can be reduced perhaps by a factor 10 by an appropriately shape cone tangent to the subreflector. It cannot be reduced by any other device: loads or an inverted cone which would not be tangent to the subreflector surface would lead to similar (within a factor 2) levels of standing wave. The standing wave amplitude is roughly proportional to the wavelength. - These numbers have been confirmed by Richard Hills 5) Standard "Chopper Wheel" Calibration technique - J.Mangum showed in memo 418 that a two-load calibration technique is more accurate than a single-load technique at sub-mm wavelengths. - Although the exact numbers used J.Mangum document can be debated, this is generally confirmed by an independent work (yet unpublished) by S.Guilloteau - The difference between the two methods (single load vs dual load) can be reduced by having more accurate atmospheric models. Such models can be improved by - better monitoring of the temperature structure of the atmosphere - use the water vapor content determined by the WVR - direct measurement of sideband opacity differences leading to improved knowledge of the atmospheric transmission as function of frequency (contribution of O2, O3) - However, the improvements which can be obtained by such techniques remain unknown. 6) Bandpass calibration - Current work on bandpass calibration suggest the required accuracy can be obtained using astronomical sources in most circumstances. - Difficulties may occur for some projects - strong extended spectral lines - strong planets Comments: - We have NO device with a proven stability of 1 % - We have NO clear understanding of the cause of the large effects seen. - The expected level of standing waves for the ALMA antennas is around 1 % at 3 mm, similar to (or even larger than) the expected coupling with loads in the subreflector. Hence, obtaining a 1 % precision on the effective load values would require cancellation of these standing waves between the possible load configurations better than 1 %... - Standing wave problems are also expected in the semi-transparent vane systems. Conclusions: - Development is not sufficiently advanced to guarantee a working solution - The goal of 1 % precision on these calibration devices is extremely challenging Timescales: - The design of the subreflector must be frozen soon in order to prepare the documents for a possible Call for Tenders. The current deadline for the documents is July 2003. - The design of the cryostat and receiver optics must be frozen by the end of the year. Given the required work to develop this design, specifications of the required widgets must be frozen Possible decision paths: 1) Relax the amplitude calibration specification to 3 % Saturation effects in the receivers are most likely negligible at this level. 2) Ask for the implementation of either 2.a) a cold load for Bands 7 and above 2.b) a hot load for the same bands 2.c) a hot load for ALL bands (except Band 1) Preliminary enquiry with the receiver people indicate that at least option 2.a and 2.b should be technically possible. 3) Abandon the dual-load in the subreflector 3.a) definitely, i.e. make no reservation in the subreflector design to retrofit such a system 3.b) with reservations in the subreflector design for a possible implementation at a later date. This implies the design of a remove cone to reduce standing waves. 4) Abandon the semi-transparent vane - this would impact ABSOLUTE calibration for bands which have possible saturation problems. - the effect on RELATIVE amplitude calibration remains to be studied in a more detailed way Cost implications: - 2.a) and 2.b) / 2.c) have significant cost implications - however savings resulting from 3.a) or 3.b) may be of similar order. 3.a) will result in larger savings than 3.b) - 4) would result in negligible savings. Process: - We should submit a Change Request to the CCB (Change Control Board) in order to have the Receiver IPT study the technical issues. - Timescale is quite short. Need to be checked further with Receiver IPT, but in any case not later than end April 2003.