Configuration per Database

Configuration per Database#

pyROX supports calculations from several online databases. Molecular and atomic line opacities can be computed from the ExoMol, HITRAN, HITEMP, and Kurucz line lists. Collision-Induced Absorption (CIA) can be computed from the HITRAN and Borysow databases.

The different formats of these databases sometimes require or allow different configuration parameters to be given to pyROX. Here, we describe some notable differences between the databases. Examples of the configuration files are available on the GitHub repository.

Downloading#

pyROX can automatically download (most) required data files from the online databases. The urls list in the configuration file should contain the following urls:

  • ExoMol: The line list’s definition file (*.def.json) should be given in the urls list. The .def.json-file contains information to download the line list’s states, transitions, and partition function files (.states, .trans, .pf). Adding the broadening files (*.broad) to the urls list is optional, but recommended for ExoMol line lists (if available).

  • HITRAN and HITEMP: The partition function file (q*.txt) should be given in the urls list. You can find the url of the respective isotopologue on the HITRAN website under “Documentation” > “Isotopologues”. We also recommend adding ExoMol’s .broad-files to the urls list. The .par-file contains all transitions and needs to be downloaded as well. However, it requires a login, so we recommend downloading this manually from the HITRAN/HITEMP website using your account login.

  • Kurucz: If the species name is provided, pyROX can automatically download the transitions (gf*.pos or gf*.all) and states (*.tsv, from NIST) files.

  • CIA_HITRAN and CIA_Borysow: For the CIA calculations, you should specify the exact urls from the HITRAN website or Aleksandra Borysow’s website.

Input-data files#

pyROX should be run with a files dictionary pointing to the downloaded input data in the input_data_dir directory. The following files are required for each database:

  • ExoMol:

    • transitions (list or str): Line transitions in *.trans-file (or compressed as bz2).

    • states (str): State energies in *.states-file (or compressed as bz2).

    • partition_function (str): Partition function in *.pf-file.

  • HITRAN and HITEMP:

    • transitions (list or str): Line transitions in *.par-file (or compressed as bz2).

    • partition_function (str): Partition function in q*.txt-file.

  • Kurucz:

    • transitions (str): Line transitions in gf*.pos-file (or gf*.all).

    • states (str): *.tsv-file (from NIST), which is used for calculating the partition function.

  • CIA_HITRAN and CIA_Borysow (see also “Collision-Induced Absorption”):

    • cia (list of tuples or list of strings): If list of tuples, the expected structure is (cia_file, temperature_mask, wavenumber_mask). If list of strings, only filenames are expected.

Pressure-broadening description#

The line list databases can use different pressure-broadening descriptions (see also “Pressure Broadening”).

  • ExoMol: The broadening coefficients can be made to depend on the rotational quantum numbers J. This is handled for the “a0” and “m0” broadening diets if .broad-files are given (see “Coefficients from a file”). Alternatively, you can provide a callable function that takes the lower state J-quantum number as input and returns the broadening coefficients gamma and n (see “Parameterised coefficients”).

  • HITRAN and HITEMP: Currently only handles constant broadening coefficients, that is the same for each transition. If a .broad-file is given, pyROX will use the average values.

  • Kurucz: Uses the Van der Waals broadening coefficient from the gf*.pos-file (or gf*.all). By default, the Lorentz-width is calculated using Eq. 23 of Sharp & Burrows (2007). For alkali atoms, however, the default width is calculated using Eqs. 2 and 3 of Schweitzer et al. (1996).