Plots for trends in metadata for occulter radius and plate scale

[detoma]

Make a plot of IMAGESCL over the mission to test the nominal plate scale. Compute median, mean, and sigma.

Make a plot of RSTWVL over the mission and consider how to fit data with a spline. It may require adjustment for the wavelength shift when the filter was retuned to put data on the same scale. Joan and I notice that RSTWVL seems to become smaller on cloudy days or late in the day when clouds gets in and would like to correct for that.

Tasks

• color code points in IMAGESCL plot by occulter ID in ucomp_mission_image_scale_plot.pro
• create text file corresponding to IMAGESCL plot with columns date/time, image scale, plate scale, occulter ID
• tighten scale on TCAM/RCAM radius on the centering plot to about 320 to 360
• get new occulter sizes from Dennis. He thinks he can measure with an accuracy of 0.0002" with a micrometer in the CG1 lab.

[mgalloy]

Below is the IMAGESCL plot. It is in the engineering directory:

It plots the plate scale, by epoch, in the red with an error band.

[detoma]

We discussed this plot and changes happen when the observers change occulter. This point to an error in the measured occulter radius in the lab. The plan is to bring back the occult's and have them measured again with high accuracy. We should measure the diameter at different position angle, i.e. 0, 45,90,135deg, to verify that they are round but all the data we have from CoMP and UCoMP point to the ocular being nearly perfectly round. It seems the problem is that we do not know the diameter accurately enough.

[detoma]

It would be useful to do the following:

• Plot eccentricity data over the mission (my recollection is that eccentricity is negligible and this was also true for CoMP except for one time period when we thought the occulter was inserted with a tilt). Eccentricity over a certain value could be a flag to check that the occulter in the correct position

• Plot the average radius for all wavelengths. I suspect this get noisier for the fainter lines. If so, and if we convince ourselves we know the platescale well for 1074, can we derive a platescale for those wavelengths from the 1074 one?

• Plot the radius of the occulter for camera 0 and camera 1 over the mission. It seems there are times when the difference between the two cameras is quite large, e.g. the difference is over a pixel for 1074 in March 2022. This should not happen if the images are properly corrected for distortion and our algorithm to find the radius is robust.

[mgalloy]

I have plots the contain the more recent data, but they look weird because of the big gap.

[mgalloy]

The other wave regions look fairly similar:

[mgalloy]

Eccentricity of the mission for 1074 nm (I have plots for the other wave regions if you want them):

[mgalloy]

I am not sure what you mean by "Plot the average radius for all wavelengths.", but I do have the radius per camera by wave region, e.g., for 1074 nm:

[detoma]

Thank you Mike! For average radius I meant the average of camera 0 and camera 1 that we use for the plate scale, but what you plotted is perfectly fine. Can you please make additional plots where you change the scale so we can see better how stable it is? let's try 330 to 360.

[detoma]

When looking at the difference between the two cameras, is larger early on. RCAM is systematically smaller at all wavelength. I found some notes from Steve saying this difference could be a focus problem.

[mgalloy]

The code to compute the image scale is:

  if (occulter_id eq 'NONE') then return, !values.f_nan

  ; occulter physical diameter [mm] measured in the lab
  occulter_diameter = run->epoch('OC-' + occulter_id + '-mm', datetime=run.date)

  ; magnification of optical system (occulter image radius/occulter radius,
  ; 10 um pixels)
  magnification = radius * 0.01 / (occulter_diameter / 2.0)

  ; focal length at this wavelength [mm]
  focal_length = run->line(wave_region, 'focal_length')

  ; image scale in [arcsec/pixel]
  ; 206264.8062471 = 360 * 60 * 60 / (2 * pi)
  platescale = 206264.8 * 0.01 / magnification / focal_length

[detoma]

According to Ben's log, this is when the occulter where replaced:

Date	Topic	Occulter	Comment
2021-08-04	Hardware	Occulter	Occulter changed to #28 (was #32)
2021-09-02	Hardware	Occulter	Spacer ring added to the occulter stage to move the occulter closer to the sun. Original design assumed occulters installed upside down.
2021-09-14	Config	Occulter	Occulter swapped #31 installed
2021-10-27	Config	Occulter	Occulter changed to #34,  *note: the logs says #34 but the header says was #35
2022-03-05	Hardware	Occulter	New occulter post installed. This removed the "field stop" round ring that originally shipped with UCoMP and replaced with with a single vertical post that extends to a horizontal bracket outside the UCoMP fov. This change may have removed the bright blob seen at the end of the lyot stop pupil
2022-03-17	Config	Occulter	Occulter #31 installed
2022-04-29	Config	Occulter	Occulter Changed to #27
2022-09-15	Config	Occulter	Occulting disk changed to #31
2022-10-24	Config	Occulter	Occulting disk changed to #35

More info can be found in this document.

There may be an error in the logs. On 2021 1027 the logs say occulter #34 was installed but the occulter in the header is listed as #34. Ben is trying to read the occulter number in the data to confirm which one is the right one.

[detoma]

from Mike:

Here are the occulter sizes the pipeline uses:

OC-1-mm                        : type=float, default=20.599
OC-2-mm                        : type=float, default=20.625
OC-3-mm                        : type=float, default=20.676
OC-4-mm                        : type=float, default=20.752
OC-5-mm                        : type=float, default=20.803
OC-6-mm                        : type=float, default=20.828
OC-7-mm                        : type=float, default=20.879
OC-8-mm                        : type=float, default=20.904
OC-9-mm                        : type=float, default=20.955
OC-10-mm                       : type=float, default=20.955
OC-11-mm                       : type=float, default=21.006
OC-12-mm                       : type=float, default=21.031
OC-13-mm                       : type=float, default=21.057
OC-14-mm                       : type=float, default=21.082
OC-15-mm                       : type=float, default=21.158
OC-16-mm                       : type=float, default=21.209
OC-17-mm                       : type=float, default=21.234
OC-18-mm                       : type=float, default=21.285
OC-19-mm                       : type=float, default=21.336
OC-20-mm                       : type=float, default=21.387
OC-21-mm                       : type=float, default=21.438
OC-22-mm                       : type=float, default=21.463
OC-23-mm                       : type=float, default=21.488
OC-24-mm                       : type=float, default=21.615
OC-25-mm                       : type=float, default=21.692
OC-26-mm                       : type=float, default=21.692
OC-27-mm                       : type=float, default=21.717
OC-28-mm                       : type=float, default=21.793

OC-29-mm                       : type=float, default=21.869
OC-30-mm                       : type=float, default=21.946
OC-31-mm                       : type=float, default=22.022

OC-32-mm                       : type=float, default=22.073
OC-33-mm                       : type=float, default=22.123
OC-34-mm                       : type=float, default=22.174
OC-35-mm                       : type=float, default=22.225

OC-36-mm                       : type=float, default=22.301
OC-37-mm                       : type=float, default=22.377
OC-38-mm                       : type=float, default=22.454
OC-39-mm                       : type=float, default=22.530
OC-40-mm                       : type=float, default=22.606
OC-41-mm                       : type=float, default=22.682

[detoma]

There is no information in the headers about the occulter used before 2021 08 24. The headers after 2021 08 04 report occulter #28 but is unclear when this occulter was changed and what was used before. I do not see a note saying there was an occulter change on 2021 08 04 in the logs for that day. There was an occulter swap on 2021 08 05 but language is unclear. Maybe Ben can clarify what his notes mean.

The information in the logs about the occulter installed on 2021 10 27 and used until 2022 03 16 is inconsistent with the one in the headers.

It seems there is some uncertainty in the occulter used during these two periods:
2021 07 21 - 2021 08 03
and
2021 10 27 -2022 03 16

The second period correspond to the jump in the plate scale. Assuming the occulter used was #34 instead of #35 reduces the plate scale slightly but not sufficiently to explain the large difference seen in the measured occulter.

[detoma]

It seems we used six different occulters so far: #27, #28, #31, #32, #34, and #35.

We updated the request to measure UCOMP occulter so it includes all occulter with numbers from 27 to 35.

[detoma]

The large jump in plate scale in early 2022 is not caused by the occulter but is a real change in platescale.

The jump did not occur at a time of an occulter change. Moreover, no reasonable inaccuracy in the measured occulter size, nor a mistake in the occulter installed could explain a 4% jump in plate scale. The difference between occulter #27 and #35 is ~2.3% and we would see obvious spillage if the observers had installed a significantly smaller occulter than the planned one.

After examining the UCOMP data around the time we see the large and sudden increase in plate scale, I could identify that the jump occurs between February 1st and 2nd 2022 when the new field lens was installed (see observer log)
February 2nd 2022 is when the observed occulter radius in pixels suddenly changed from ~356pixels to ~342 pixels in 1074 (a similar decrease of 12 pixels in the apparent size of the occulter is also seen in 789). This seems to be a real change in the plate scale due to changes in the instrument that currently we are not accounting for.

There are smaller discrepancies in the plate scale at the times when an occulter was changed that hopefully can be resolved by the new and more accurate physical measurements of the UCOMP occulters that we have requested.

[bberkeyU]

Before the field lens change in Feb 2022, UCOMP was out of optical perception. The field lens was too close to the modulator/lyot stop reducing the effectiveness of the Lyot stop.

The fix brought the field lens back to the location expected in the zeemax model. Moving the field lens required changes to the O1 positions (command positions changed in the config file) and camera focus position (changed by manually moving the camera focus stage hardware).

The (camera/O1) focus changes took place on Feb 07, 5 days after the field lens change; this may lead to a minor jump in the plate scale during this period.

[detoma]

We need to take flats with the occulter in to properly determine the radius of the occulter and derive the plate scale accurately, like we do for KCor.

A new diffuser with higher signal at the short wavelength will help with this.

When we reopen, we need to take these flats daily at least for a while until we have some statistics to compute the plate scale. Then we can make the measurements once or twice a week. I know this is time consuming but accurate plate scales are critical for science. To minimize the impact, we will determine the minimum number of repeats to get good S/N. We do not need multiple wavelengths. Data at line center are sufficient for this. We should be able to do this in 5-10m.

We like these data near zenith (+/- 2h), unless there is a CME going on. We can plan them after the wave program.

[bberkeyU]

We have a cookbook that will do this. measure-occulter.cbk. We can put this into daily.menu after the Waves program, along with all the other changes we plan to make to the observation program when we get back online.

[detoma]

Was this program created by Steve?

[bberkeyU]

No, I made it an hour ago. I put a dark in there but am not 100% sure if it is needed.

[detoma]

These new flats with the occulter in will. to be used by the pipeline to flat-field science data. The pipeline is already ignoring the flats with the occulter in.

However, they will be saved into the new flats directory.