Data for figures (and code for generating both data and figures) in "Bridging the Prandtl number gap: 3D simulations of thermohaline convection in astrophysical regimes"

Overall summary This repository primarily hosts files that contain the data necessary for generating the figures in "Bridging the Prandtl number gap: 3D simulations of thermohaline convection in astrophysical regimes," a manuscript accepted for publication in The Astrophysical Journal Letters. The code in this GitHub repository, which also includes the code for generating this data, can then be used to generate the figures from this data. The contents of the GitHub repository are included here for convenience. File formats To preserve the directory structures assumed by scripts that generate figures from data, all data files are in datafiles.tar.gz, which must first be unpacked. Most of the data is in .h5 files (using the HDF5 file format), which contain the outputs of simulations run for the manuscript. As described below, the data behind Figure 1 includes only simulation inputs, i.e., the parameters chosen for each simulation -- these quantities are contained in .cfg files, which can be read as plain text. Each individual simulation/run for this manuscript has a name, with inputs given in config_files/RUN_NAME.cfg (plain text) or config_files/hydro_LPN/RUN_NAME.cfg for the Pr=0 cases, and .h5 (HDF5) outputs given in various subdirectories of runs/RUN_NAME/ or runs/hydro_LPN/RUN_NAME/ for the Pr=0 case. For all .h5 files, the relevant data is found in the tasks/ group, while scales/ contains independent variables (t, x, y, z) and related quantities (e.g., timestep number, timestep size). These .h5 files were generated by Dedalus v2, as described in the Dedalus tutorial here. 1D data (timetraces) for each run is found in .../RUN_NAME/scalars/scalars_s*.h5 (where scalars_s1.h5 contains the first several timesteps of data, scalars_s2.h5 -- if present -- contains the next several timesteps, etc.), while 2D data (snapshots/slices) for a small selection of runs is found in .../RUN_NAME/slices/slices_s*_w_ywall.h5. If you want to work with this data "from scratch", i.e., with your own scripts rather than using/manipulating the scripts provided here, you will need to use a package that can read HDF5 files, such as h5py (link) if you are using python. The .cfg files are all plain text, where input parameters (e.g., domain sizes, resolutions) are specified by INPUT_PARAMETER = VALUE separated by newlines. Note that all data is unitless, having been nondimensionalized as described in Section 3 of the manuscript. Also note that some .cfg files and/or runs/RUN_NAME/ subdirectories exist for runs for which no data is included. These "ghost simulations" were run in earlier stages of this work but ultimately not included in the manuscript. To get a list of the runs that are included, run the script check_BGS13.py, which will print the full list of simulations included in the figure. Users who look closely at the hydro_DDC_... .py Dedalus scripts will notice that more data was generated by these simulations than eventually made it into the manuscript (including 1D profiles and checkpoints). These other data files, being unnecessary for the manuscript, are not being maintained long term and are not included in this repository, but may be procurable upon reasonable request to the author. File descriptions The description below goes through each figure and lists the data/files associated with the figure and the files that are used to create the figure from the data. The files for generating the data are hydro_DDC_IVP.py and hydro_DDC_IVP_LPN.py, whose uses are demonstrated in the corresponding .pbs scripts (which are meant to be run on NASA's Pleiades cluster, or a computing cluster with similar job submission/queueing software). Note that these two hydro_DDC_ ... .py scripts require v2 of the Dedalus python package installed, and all its associated dependencies. You can find the GitHub repo for Dedalus v2 here, and the documentation and installation instructions here. Note that if you want to use the scripts in this repository to regenerate the data, you need to ensure you are installing and using Dedalus v2, not the latest v3. Figure 1 Data for Figure 1 The data necessary for Figure 1 is contained in config_files/RUN_NAME.cfg, where RUN_NAME is the assigned name of the various simulations that were run to generate the data for this manuscript. Files for Figure 1 From these files, Figure 1 can be recreated by running R0_tau_diagram.py, a python script that requires python 3.*, numpy, and matplotlib, in addition to various standard python modules. Note that this script assumes that the data is in the same file structure as in this repository. Figure 2 Data for Figure 2 The data necessary for Figure 2 is contained in the following files: runs/eps0p1_tau5e-7_Pr1e-6_2z/slices/slices_s11_w_ywall.h5 runs/eps10_tau5e-7_Pr1e-6_hres/slices/slices_s4_w_ywall.h5 runs/eps300_tau5e-7_Pr1e-6_vvhresx1p5/slices/slices_s5_w_ywall.h5 Files for Figure 2 From these files, Figure 2 can be recreated by running plot_w_ywall_comparison.py, a python script that requires python 3.*, numpy, matplotlib, and h5py, in addition to standard python modules. Note that this script assumes that the data is in the same file structure as in this repository. Figure 3 Data for Figure 3 The data necessary for Figure 3 is contained in runs/RUN_NAME/scalars/scalars_s*.h5, for Pr > 0 runs, and runs/hydro_LPN/RUN_NAME/scalars/scalars_s*.h5, for Pr = 0 runs Files for Figure 3 From these files, Figure 3 can be recreated by running check_BGS13.py, a python script that requires python 3.*, numpy, matplotlib, and h5py, in addition to standard python modules. Note that this script assumes that the data is in the same file structure as in this repository.