MIRI/MRS destriping tool to remove background systematics

This repository provides a simple fix to background systematics in MIRI/MRS data cubes (*s3d.fits), as described in Matthews et al. (submitted; pre-print on request), Appendix B.

Installation

First clone the github repository:

git clone [email protected]:ecmatthews/mirimrs_destripe.git

Then, navigate to the cloned directory, and install with pip:

cd mirimrs_destripe
pip install -e .

Note that this has not been beta-tested, and there may be issues, e.g. missing dependencies! Please reach out with any problems.

Examples

While the package is not fully documented, we provide a brief guide to using the code here.

First, define the location of the s3d.fits file you wish to correct, and optionally an x1d file with the wavelength calibration (the x1d file is not used for the systematics removal, but only for a diagnostic plot showing the extent of the systematics).

file = "data_ifualign/Level3_ch1-long_s3d.fits"
file_wavs = "data_ifulaign/Level3_ch1-long_x1d.fits"

Then, run the analysis. There are several key steps:

• Mask the source, and dithered copies of the source. Here we use the brightest and faintest points in a collapsed cube to mask sources, but a custom approach may be needed especially for other astrophysical sources.
• Run the destriping algorithm on the s3d.fits cubes
• Re-run the stage3 extraction from the normal JWST pipeline (please update with your preferred keywords!)
• Make diagnostic plots, showing the spectrum with and without the systematics correction term.

cube_destripe = CubeDestripe(file, file_wavelengths=file_wavs, outfolder="example_outputs")
cube_destripe.mask_source()
cube_destripe.destripe()
cube_destripe.run_stage3_extract(fringe=False, run=True, ifu_rscale=2.)
cube_destripe.plot_spectra_with_flattening()

When running the code, several files are created in subdirectories of the defined outfolder.

• flat_cubesfiles contains s3d.fits files, with the background systematics removed.
• flat_extract_1d contains extracted spectra (x1d.fits) from these systematics-corrected cubes.
• basic_extract_1d contains identically extracted spectra, but without the systematics correction, for comparison.

Several diagnostic plots are provided, in the following subdirectories of the defined outfolder.

• diagnostic_plots contains images of the cube (collapsed across wavelength), as well as the average pixel value in each row and column. The diagnostics are created both with and without the correction applied; in the case with correction these traces should be close to 0. The average pixel traces are also created for sub-sets of the wavelength range, to indicate how systematics are changing through the cube. We recommend to check in these cubes that the source and any negative copies of the source are suitable masked out in these diagnostic images.
• plots_speccompare shows a comparison between spectra extracted with and without the systematics corrections.

Improvements

List proposed improvements:

• Beta testing: does installation work correctly? does the code work on other datasets?
• Better/additional masking algorithms to remove sources from data. Use astrophysical information for this?
• Output plot with all wavelengths plotted (as well as current wavelength-by-wavelength plots).
• Better documentation?
• Explore data-driven methods in 2-D detector images (*cal.fits) rather than 3-D cubes (*s3d.fits, as used here).

This package is still being actively developed. Please reach out if you're interested in contributing!

Attribution

If you make use of this code in your research, please cite Matthews et al. (submitted).