Action item 19 from ALMA ObsMode Cycle 2 workshop
See spreadsheet at Cycle2ObservingModesWorkshop
with Todd's response:
I recommend keeping the limit of 150 pointings per mosaic. Basis: Using 8 GB as the largest image (45000x45000 pixels), and 5 pixels per synthesized beam, this sets a limit of 5500 pointings in a 12m mosaic undertaken in a 3km configuration. This result is independent of frequency! Therefore, the practical considerations of observing will set the restriction. Most mosaics will be done in FDM mode, for which the minimum integration time in Cycle 2 is still 6 seconds. We would like to be able to finish the whole mosaic in one execution of an SB, or in one session if sessions are possible. Initial calibration requires 15 minutes. If the length of the scheduling block (or session) is x hours, then the time available on source is (x-0.25)*3600*e, where e is the observing efficiency with respect to phase calibration and subscan overhead. If we spend only n integrations per pointing, then the maximum number of pointings is: Pmax = (x-0.25)*3600*e/(6*n). With x=2hr, e=0.60, and n=2, Pmax=315. This is an absolute maximum (300 pointings) which is based on an optimistic efficiency, so I recommend sticking with 150.