TestCaseData/2009-10-14_1934-638_commissioning_data
Description
Investigation of bandpass variations using the long observations of 1934-638 done during scientific commissioning.
This data set has an unusual fault - the time stamps on some records are incremented by 86400 seconds (instead of 0 or 1). This causes miriad to complain about the visibilities being out of time order; this is not the problem, the problem is the timestamps are off by 1 day.
To hack out the problem records, run uvflag
with these selections:
select=visibility(91993,93432) select=visibility(83353,84792) select=visibility(63361,64800) select=visibility(56161,57600) select=visibility(44581,44640) select=visibility(8581,8640)
Initial calibration was to take a small section of the data (2009-04-11_18:42 to 19:06) to compute the bandpass correction (with mfcal). This bandpass correction was applied to all the data on either side of this time range. No time-variable was bandpass applied.
6 cm Data
The first interesting feature (after taking out the median amplitude and phase of the spectrum at each time step) is that there is a clear systematic change in phase gradient (ie astronomical phase as a function of channel number) with time, that has larger amplitude and varies faster on longer baselines.
3 cm Data
The 3 cm data exhibits the same systematic change in phase gradient and is of similar magnitude.
Data Set(s)
2009-04-11_1103.CX168 2009-04-11_1115.CX168 2009-04-11_1219.CX168 2009-04-11_1340.CX168 2009-04-11_1500.CX168 2009-04-11_1621.CX168 2009-04-11_1741.CX168 2009-04-11_1901.CX168 2009-04-11_2022.CX168 2009-04-11_2142.CX168 2009-04-11_2303.CX168 2009-04-12_0018.CX168 2009-04-12_0139.CX168 2009-04-12_0300.CX168 2009-04-12_0421.CX168 2009-04-12_0442.CX168 2009-04-12_0602.CX168 2009-04-12_0721.CX168 2009-04-12_0824.CX168 2009-04-12_0825.CX168 2009-04-12_0835.CX168
Tickets
N/A (so far)