Previous: EVN/VLBA Joint Operations
Up: EVN TWG/Friends Meeting
Next: The Role/Format of the TWG
Previous Page: Line and Polarization Specials; Offline Software
Next Page: Polarization Residuals
Previous: EVN/VLBA Joint Operations
Up: EVN TWG/Friends Meeting
Next: The Role/Format of the TWG
Previous Page: Line and Polarization Specials; Offline Software
Next Page: Polarization Residuals
Spencer described the way in which calibration was done with the MERLIN array. He presented maps with 1000:1 dynamic range and tried to identify the reasons why MERLIN's calibration was better than the EVN's. He suspected that the main reasons could be attributed to the following features:
Spencer also presented some joint VLBI/MERLIN calibration data which suggested that the original gain and TSYS info made available by the friends was different from the numbers required to obtain a consistent calibration table. Spencer felt that there were lessons to be learned from MERLIN. In particular the method of calibration should be standardized across the EVN, more use of point source calibrators, more network time set aside for test and calibration data, calibration of closure offsets & professional scheduling. These last points would require a full time network operations management. Schilizzi asked how close the maps were to thermal noise. Spencer explained that the maps were all thermally limited and that the polarization cross talk was approx 0.5%. Porcas highlighted the fact that unresolved point sources were not readily available for VLBI. Spencer noted that for phase calibration the sources need not be point like - although they should not be completely resolved and they must be strong. Porcas was worried about reducing the freedom of astronomers as regards scheduling. It was important that astronomers should be able to supply input to the system. Massi asked what the dynamic range was without correcting for the closure offsets. Spencer estimated that it was 1000:1 before applying the corrections and 30000:1 afterwards. Schilizzi noted that better calibration would require better communication between the correlator, telescopes and institutes. Graham suggested that it would be nice to see a written report of the calibration info presented by Spencer.
Burgess gave an overview of the different ways in which Calibration data were collected at each of the stations. Both Jodrell and Westerbork use noise adding diodes. In Jodrell's case only a 2MHz wide channel is used. Burgess asked the group if they thought this was adequate and whether it was repeatable. Foley felt that calibration observations were repeatable at the 1% level. Garrett reported that at Jodrell the calibration runs were not always repeatable though the main cause of this was due to interference at 18cm. Spencer called for a uniform system to be implemented across the network. Schilizzi noted that it would certainly have to be uniform as seen from the correlator. Porcas highlighted the fact that it was no good having TSYS values accurate to 1% if the gain curve was out by 20%! Porcas went on to present some calibration questions that each station might like to consider. The presentation could be summarized by the three main questions (see appendix 2 for more details):
Porcas suggested that the friends should try and feed in their own calibration data into the AIPS task ANCAL. Bregman noted that all these steps were different at each station. Porcas countered that we could not afford to be complacent on this matter. Burgess called for each station to provide updated gain curves. Schilizzi requested polynomial coefficients as well.
There was some discussion about problems of set up time. It was generally felt that these were local problems and they should be dealt with at each site. Schilizzi hoped that JIVE support scientists will be able to start calibration checks after the February session. In the meantime Graham volunteered to organize a MkIII dual polarization experiment. Bregman wanted to know what we might achieve from this experiment. Alef felt that if nothing else it would establish the condition of the write heads at each station.
ACTION ITEM: Friends to supply gain curves and polynomials to ASTBO1.
ACTION ITEM: D.A. Graham to organize calibration experiment. R.W. Porcas to liaise with Schwartz regarding scheduling.