****************************************************************

Create: 2020-07-01

Last Modified: 2023-08-02

****************************************************************

diag_feedback_E3SM

• a developing version to diagnose feedbacks for E3SMv1 and developing versions

description:

• This package is used to run some cloud feedback analysis for E3SM developing versions.

Environments:

• python (anaconda)
• latest CDAT
• cartopy [conda install -c conda-forge cartopy]
• NCO [conda install -c conda-forge nco]
• sklearn [conda install -c conda-forge scikit-learn]
• statsmodels [conda install -c conda-forge statsmodels]
• global-land-mask [pip install global-land-mask]

Detailed commands:

• conda create -n diag_feedback -c conda-forge -c cdat/label/v8.2.1 python=3.7.12 cdtime cdms2 genutil cdutil psutil cartopy nco scikit-learn statsmodels
• conda activate diag_feedback
• pip install global_land_mask
• f2py -c tropo.f90 -m tropo [a *.so file will be generated]

Pre-needed data

• please download radiative kernel data (Huang et al., 2017) from: https://www.dropbox.com/sh/ngfb7bxhwcbxwu8/AAC6AIha5rLjsl3lUZPiLO6Ua/toa?dl=0&subfolder_nav_tracking=1 first. Then, set RadKernel_dir = datadir in main.py.

How to use it?

• main.py is the control script. Please modify it following all related settings in that script.

• main-plot.py is used to get several plots for diagnosis.

• additionally, tropo.f90 is a fortran file. so, use f2py to convert it into one .so file: f2py -c tropo.f90 -m tropo

  • If some errors occurs, please try to 'module load gcc' first and then execute the command again.]
  • If loading gcc does not help and you get some errors like , please run 'export CFLAGS=-std=c99' first and then execute the command.

• other cal_xxx.py files are defined functions used by main.py. Please don't modify them.

outputs?

• the directory "figure" includes some basic output figures.
• the directory "data" includes some NC files and CSV files generated by main.py and they will be used by main-plot.py for further analysis.

history:

• Aug 6, 2020: added plotting scripts as v1
• Sep 30, 2020: use different color for LW,SW,NET in CldRadKernel plot; include amipFuture as a comparison
• Nov 10, 2020: update input data for other CMIP models
• Feb 18, 2021: add plot_tas_latlon
• Mar 21, 2021: add plot_CldRadKernel_latlon
• Jul 20, 2021: re-organize
• Oct 10, 2021: clear-up and re-organize main.py and main-plot.py
• Dec 22, 2021: generate a web page to better organize and visualize all figures.