Atmospheric lines for wavelength calibration at longer wavelengths.

Graham Harper asked me about what depth atmospheric line we need for wavelength calibration.  His project at 26.00 microns will have a few very strong lines and a 4-5% line.

I looked at this by fitting Gaussian absorption lines using CLICKFIT.PRO to black-sky from the 17 micron H2 setting on Jupiter during flight 1.  I should probably do other wavelengths and actually use the pipeline, but this was how I started.

The black-sky looks like this:

I used CLICKFIT to test how well we can fit a Gaussian to various lines.  The lines that I used (and the derived centroids) can be seen in this version with crosses on the lines.

The model atmosphere for this region (not ATRAN, but my AT version) is below.  The white is for 10 microns PWV and the red is for 40 microns PWV.  The green line is at 95% transmission.  The CO2 band near the bottom of the images above is near the right edge of the atmo plot.  The two strong lines are near the middle order and in the second order.  In the images above, wavenumber increases to the right and down, like a book.  In the atmo plot, wavenumber increases to the right.

The upshot is that most of the lines I used are ~10% and one of them is around 5% (just to right of strong line at 585.7 cm-1 in the second order from top).  Since the pipeline will sum along the slit and can combine multiple black-sky images, I think we are probably safe with 5% features at wavelengths where the sky emits well.  This probably includes 13 microns, although should check to be sure.  At 7 microns and shorter, I'm pretty certain we need stronger features.  Have to see how bright they should be....

Lens wheel optics for CD18

Layout of lens wheel for CD018

Angle     Optical Component                          Voltage
-120 deg  mirror                                     4.9V
-60 deg   1% neutral density filter                  5.9V
0 deg     25mm KRS5 lens                             6.9V
60 deg    Open                                       7.9,1.9V
120 deg   ZnSe lens                                  8.9,2.9V
180 deg   38mm KRS5 lens + 1% neutral density filter 3.9V

i_KRS5=39.9 mm, installed at 40.6 mm from detector
i_ZnSe=28.8 mm, installed at 29.0 mm from detector
i_KRS5_25mm=52.6, installed at ~50mm from detector

38mm KRS5 lens + 38mm ND filter:
In CD017, we were unable to image the pupil with the previous arrangement of the 38mm KRS lens and 25mm 10% ND filter. This was probably due to multiple reflections between the lens and filter.
In CD017, the arrangement had been detector-ND-lens.
For CD018, the 38mm 1% ND filter has been placed outside the lens:  detector-lens-ND.


We have a shortage of 38mm retaining rings for the barrels and thus couldn't hold any blocker windows in the remaining free slot. I chose to put in another pupil lens, purposely installed closer to the detector than would be ideal for imaging the echelle grating. Maybe it will do a better job on imaging the TAAS hot plates?

I will ask Emmett to make more retaining rings for us in time for CD19.

PSF sizes from Ceres maps at 1305cm-1, 731cm-1, 587cm-1 and 354cm-1

Reporting on the PSFs during flight using the Ceres mapping data at different wavelengths.

Ceres was the only pseudo-point-source target observed in mapping mode during flights. According to the JPL HORIZONS website, Ceres had an angular diameter of 0".815-0".817. 

Subtracting this value in quadrature from our typical 3" image size reduces the value by about 0".1. I made this adjustment in the FWHM values in the table.

Processing steps:
proceded through rectification stage of pipeline reduction. Selected "skip coadd", which produces one rectified 2-D spectrum for each map position.
Outside of the pipeline, I lined up the spectral orders end-to-end, so that the 2D spec was just wavenumber x height on slit per each map position. In the process I masked low signal areas at the end of the orders, and then summed along rows (wavenumbers).
At this point I had a 1-D slit profile for each map position. Placed side-by-side, it makes an image with pixels 0".206 x (map step size).
I resampled the images to have square "map_step_size" pixels using the interpol.pro function
I used a clear patch of sky on the images to get a median sky value and subtracted it from each row (corresponding to a single map position). This was only important for the 1305cm-1 which didn't have much continuum flux to begin with and was effected by small changes in the sky line brightness.

IDL .jou file is on dropbox at:
everything EXES/analysis/cd17/ceres_maps/ceres_maps.jou

The data summary is as follows:
filenumbers      wn0  npts  map_step_size  fwhm1(")  fwhm2(")
30039,30040 1305cm-1  15       0".95       3.22      3.06
40059,40060  731cm-1  15       0".95       2.98      2.27
30046,30047  587cm-1  11       1".6        3.17      3.25
30051,30048  354cm-1  11       1".6        3.49      3.19

fwhm1 is along the slit direction
fwhm2 is along the mapping direction

Map step size should be verified with the telescopes housekeeping data, especially for unusually small FWHM2 for 731cm-1.

The figures show the re-sampled, skysubtracted maps for each setting.

1305cm-1 scan. X is slit direction. +Y is mapping direction. Pixels are 0.95" x 0.95"

731cm-1 scan. X is slit direction. +Y is mapping direction. Pixels are 0.95" x 0.95"

587cm-1 scan. X is slit direction. +Y is mapping direction. Pixels are 1".6 x 1".6

354cm-1 scan. X is slit direction. +Y is mapping direction. Pixels are 1".6x1".6