Issue with continuum emission and long wavelength variations (IDL PIPELINE)

[lucarizzi]

Dear Luca and Josh,

We are reaching out for some guidance. We reduced our recent KCWI data (medium slicer, no blue filter, 2x2 binning) and found that the continuum emission in the final reduced IFS data cubes has long wavelength waves in it as you move spatially across the source. My guess is that this has something to do with the atmospheric dispersion correction not working quite right.

I'm attaching a PDF with some diagnostic plots we made. The top plots are looking at the raw data, to see how much variation there is in the continuum shape for different adjacent columns - this is before any processing or reduction is done. We compared data from a previous 2020 run to our recent March 2021 run, for sources at different airmass. There are clear "color" variations in the continuum in the raw data for different columns, where the stellar continuum is redder or bluer in different columns. I assume this has to do with atmospheric dispersion? We checked both science targets and standard stars, and see it in both. We also checked a flat field image, and we don't see it there.

If you then go to the ~middle of the PDF we show the final reduced cubes (flat field, atmospheric dispersion corrected, flux calibrated), and you can see that our data for a source (J1205) from 2020 did not show differences in the "color" of the continuum. For a different source (J0944) from 2019 we see long wavelength "dips" in the continuum, which are unphysical for a star (ie, galaxy spectra don't have these long wavelength dips in the stellar continuum). Then in 2021 for the same original source (J1205) we see similar continuum dips and rises, and we see this again for 2021 data on J0944.

It seems to me that somehow the continuum is being over-corrected at the central wavelengths for some spaxels, and under-corrected for other spaxels.

Have you seen anything like this before? It "smells" to me like at atmospheric dispersion correction, but it could be something else. We have tried redoing the data reduction using external instead of internal flats, different standard stars, and even calibrations from a different run, and this issue persists.

Any ideas on what this might be, or how we might try to fix it?

Thank you! Alison Diagnostic Plots for Continuums_updated.pdf

[lucarizzi]

this is issue is related to the IDL pipeline, but chances are we will see it on the Python one too. I will download the data and try a reduction.

[lucarizzi]

Additional conversation and the night log You have my permission to access the archive for this data. Let’s do J0944 as a test - I’m attaching the log, the science exposures are #72-75, and you can use the standard taken in evening twilight (G191B2B).

log_2021-03-18.pdf

[lucarizzi]

https://drive.google.com/drive/folders/1K1oGwxX2-K_-imhKOvbd9PAJaJpR7UEL?usp=sharing

Data is reduced with Python pipeline, now awaiting reply from Observer.