CRIRES/Molecfit Stack V1.0

Code to automatically process and fit CRIRES spectra of telluric standard stars. When provided with a set of raw spectra, the code:

• Automatically finds matching calibration frames from the ESO Archive
• Produces input files for the CRIRES pipeline, and calls esorex to produce the frames
• Removes bad pixels and wavelength-calibrates the data
• Creates input files for and runs molecfit to determine column densities of greenhouse gases.

Dependencies:

molecfit

Install from https://www.eso.org/sci/software/pipelines/skytools/molecfit - GUI isn't required for this code. For the scripting to work properly, add the INSTALL_DIR/bin directory containing the molecfit executable to PATH.

CRIRES Pipeline

Install from http://www.eso.org/sci/software/pipelines/ - installing without esoreflex and gasgano is fine. Add the INSTALL_DIR/bin directory containing the esorex executable to PATH also.

Python packages

astroquery, skycalc-cli, tqdm, and emcee - install with your preferred method.

pigz

Currently the code uses pigz for marginal improvements in speed when decompressing. If you don't have it/want to use it, see cal_selector.py for code to swap to gzip.

Install:

• The setup script included will grab the necessary calfiles, as well as set up the directory structure, dropping an example of how to format your input in /raw.
• In the code for cal_selector.py, set your username. On first run, the keychain will request to store your password. After first run, you shouldn't be prompted again.

Running the code:

• Under raw/, create a folder for each observation block - each block should be same exposure time, wavelength
• Inside each observation block directory, put the raw frames to be processed in a subdirectory obj/
• From the top-level directory run python process_frames.py
• Currently, the calibration frame grabber, pipeline bootstrapper, and wavelength calibration routines run in sequence on each folder, then molecfit runs across N-1 cores.
• Note that each stage of the pipeline requires the previous one to have run successfully, so if the code crashes/fails to run, the next stage of the pipeline likely cannot find what it is looking for. Check the full documentation to find what is missing.

Known issues:

• In calibrate_frames.py, errors thrown by esorex will cause the code to crash. To fix: remove all sets that have already processed and remove the troublesome set, then re-run the code.
• molecfit has some sporadic segfaults for certain wavelength ranges. It's unclear why this happens for some datasets and not for others at the same wavelength range, further work needed. For now, tweak the wavelength ranges in molecfit_bootstrapper_parallel.py.