ALMA Memo #403
Fast Switching Phase Correction Revisited for 64 12 m Antennas

M.A. Holdaway (NRAO/Tucson)

2001/12/17

Fast switching phase calibration has not been investigated for the ALMA telescope since ALMA has been defined as 64 12 m antennas. Furthermore, the logic chain which picked the optimal calibrator in past investigations was approximate. In order to better understand the requirements which are placed on the current ALMA design by fast switching, we have rewritten the fast switching simulation code in AIPS++, including a more complete optimization with fewer assumptions, using updated sensitivity, antenna slewing, and atmospheric information.

We find that when the observing frequency is matched to the phase stability (ie, high frequency observations are always carried out during the most stable phase conditions), the Chajnantor site is good enough to permit fast switching observations of the expected frequency range (ie 30 to 950 GHz) to succeed with high efficiency. Typical observing efficiencies, including both time lost to the phase calibration cycle and decorrelation losses, range between 0.80 and 0.90 for sources above 45 deg elevation angles. The observing efficiency decays very gently at lower elevation angles, with a typical efficiency of 0.70 at 20 deg elevation.

The extra sensitivity provided by 64 12 m antennas does not help as much as might be expected with fast switching because the time spent integrating on the calibrator is very small compared to the entire cycle and is a moderately small portion of the calibration phase of the cycle. The 1.5 s delay due to changing frequencies is pretty well matched to the slew times for typical objects. The slew profiles provided by Vertex are sufficient.

The residual phase errors resulting from fast switching will cause baseline-dependent decorrelation. Some minor algorithmic work should proceed on fixing this decorrelation. It seems likely that the phase information gleaned from observing the calibrator will be sufficient to accurately estimate the decorrelation correction on a per baseline basis.

View a pdf version of ALMA Memo 403.

Download a postscript version of ALMA Memo 403.