This directory holds data and code underlying the figures of the JGR-Atmospheres
paper by T. A. Mooring and T. A. Shaw, "Atmospheric diffusivity: a new energetic
framework for understanding the midlatitude circulation response to climate change".
All of the figures were initially rendered in MATLAB as .ps files--therefore all the
code here is .m files and the data is stored in .mat files.  The .ps figures were
converted to .pdf using the epstopdf tool prior to submission to JGR-Atmospheres.

The main paper contains 9 figures and its supporting information contains 4 more.
This directory contains
1) this readme
2) a single MATLAB script, top_level_script.m (described below) and
3) 9 subdirectories

The subdirectories are named according to the figure(s) for which they contain
data and code.  Each subdirectory contains
1) A script whose name ends in "_archive.m".  This script reads in data and
     actually renders the figure(s) as .ps files, with names ending in 
     "_archive.ps".  This script implicitly assumes that the necessary data
     files are located in the current working directory.  
2) One or more .mat files containing the necessary data--sometimes the same
     data file is included in more than one directory, for clarity and
     convenience in scripting.
3) The figure(s) as generated at the University of Chicago for submission to
     JGR-Atmospheres.  They are included as .ps files, with names ending in
     "_vX.ps" where X is some number.
4) An additional .m function needed by the main "_archive.m" script.  The one
     exception to this rule is figure_C1, for which the main script does not
     need an additional user-provided function.

The purpose of top_level_script.m is to conveniently regenerate all 13 figures.
It does this by sequentially descending into each subdirectory and invoking the
the relevant "_archive.m" script.

The "_archive.ps" files generated using the code and data found here should
agree well with the "_vX.ps" files used for the paper itself.  Agreement will
not necessarily be perfect, perhaps because of variations in graphics rendering
behavior across MATLAB versions and/or machines.  The "_vX.ps" files were made
using MATLAB R2014b at the University of Chicago, and the ability of
top_level_script.m to successfully regenerate all of the figures was confirmed
using MATLAB R2018b on a different machine at Harvard.  Both machines run
Unix-like operating systems.  The scripts have not been tested on Windows or
with other MATLAB releases.
