Time period: from at least July 2014 until Feb 2015 (WVR replaced March 2015).
Description: If DA59 WVR is noisy, one is probably better off by interpolating the wvr from nearby antennas.
Diagnostic: compare the blue (before) and green (after) plots in the "wvr_plot.png" file in the qa2 directory, and also by looking at the tables at the end of the uid*.ms.wvrgcal file in the calibration directory. You will see that DA59 looks phase variations are worse after WVR applied
Manual or Pipeline: We think Manual only, b/c PL has heuristics to interpolate noisy WVRs. But should confirm.
Solution: Add the line "wvrflag=['DA59']." to the wvrgcal call of the calibration script in mystep=4.
PRTSIR-6362: DV01 shows inconsistent phases but only on one baseband (BB3, counting from 1) (Todd)
Seen in multiple executions on 19-May2015. This particular example was a TDM spw.