Examine variation in demodulation matrices over mission

[jburkepile]

Steve to provide all computed demodulation matrices for the different emission lines over the mission in order to examine their variation with sky conditions, time-of-day, seasonally, over mission. This info can help assess the stability of the instrument and to assess uncertainty in the derived Stokes polarization. Pipeline currently uses an average (created awhile back) of the matrices for each emission line.

[StevenTomczyk]

All demodulation matrices are located in subversion at:

HAO-IG\UCOMP\Integration and Testing\Polarimetric Calibration\Results

The average containing the temperature dependence is located at:

HAO-IG\UCOMP\Integration and Testing\Polarimetric Calibration

It is called Dmx_Temp_Coefs.sav. It was created on 20220721.

I think this issue can be closed.

[jburkepile]

Thanks for the location information.

This ticket needs to stay open.

We will use the files that are in subversion as per Steve's info.

[jburkepile]

This info is from Ben from Sept 6:

I started digging into the UCoMP Demodulation matrixes using the .sav files in Subversion as a source. As always with UCoMP I am only looking at the 1074 data and assuming whatever we find will work for the other lines.

Plotting a time series for the 16 demodulation values, we quickly noticed the demodulation calculation was noisier at the start of the mission. And there were a few lousy demodulation computations. See plot for the first demodulation value vs time.

We installed the modulator header 2022-11-16 between cal# 138 and 139. The cals 139 and 140 appear to be noisy due to clouds (139) and haze (140) but by 2022-11-19 #141, the plot seems to settle down to the flatter curve.
The bad calls are associated with the following data files:

'20210823.191904.59.ucomp.1074.l0.fts', '20210824.200341.83.ucomp.1074.l0.fts', '20211014.002418.35.ucomp.1074.l0.fts', '20211117.210220.64.ucomp.1074.l0.fts', '20220419.185223.69.ucomp.1074.l0.fts', '20220419.191236.99.ucomp.1074.l0.fts', '20220703.214052.86.ucomp.1074.l0.fts', '20220806.205312.60.ucomp.1074.l0.fts', '20220827.214407.38.ucomp.1074.l0.fts'

The 2022-04-19 files were actually focus scans and not calls. And on spot checking a few ohter days I see clouds. But whatever the source these measruements are clearly out of spec and should be excluded from the averageing.

After kicking out the bad calibrations and data taken before the modulator heaters were installed, I have plotted the timeseries for all 16 demodulaton measruments. Within the noise I think I see some structure to these measruemtns inlcude a slight step change in values around the x=60 which is late Januaary 2022 when we change the field lens position.

Still digging

-Ben

[jburkepile]

From Steve:

It would be good to plot the modulator temperature over the same time period and see if it correlates with changes in the Dmx elements.