Skip to content

About

Combine datasets to dataarrays, save to zipped numpy files and plot variables.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Repository files navigation

cesm-helper-scripts

pre-commit Code style: black

Install

Clone with git and install with poetry into a virtual environment:

git clone https://github.com/engeir/cesm-helper-scripts.git
cd cesm-helper-scripts
poetry install

This give two entry points / executable packages provided the virtual environment is activated:

  • cplt: create attribute plots using an aggregated .nc file as input, i.e. output of the gen_agg script. See cplt --help.
  • nc2np: generate an .npz file from an aggregated .nc file, output of the gen_agg script. See nc2np --help.

There is also a Makefile present, that can be used to install the gen_agg script. make install will simply copy it to ~/.local/bin/, while make autoinstall will copy it to ~/.local/bin/ and replace the shebang with the currently activated python executable. Therefore, when using autoinstall, you need to make sure the dependencies are installed in the current python environment.

Usage

gen_agg

Say you are in the location of your output files for the atmosphere module. It will list files with name <simulation_name>.cam.h0.YYYY.MM.nc for the month resolution. Check out what variables it contains by running

ncdump -c <file.nc> | sed 5000q | grep -i <search-string>

For example, we may search for forcing:

$ ncdump -c e_slab_custom_frc.cam.h0.0001-01.nc | sed 5000q | grep -i forcing
                H2O_CLXF:long_name = "vertically intergrated external forcing for H2O" ;
                H2O_CMXF:long_name = "vertically intergrated external forcing for H2O" ;
                LWCF:long_name = "Longwave cloud forcing" ;
                SO2_CLXF:long_name = "vertically intergrated external forcing for SO2" ;
                SO2_CMXF:long_name = "vertically intergrated external forcing for SO2" ;
                SWCF:long_name = "Shortwave cloud forcing" ;
                bc_a4_CLXF:long_name = "vertically intergrated external forcing for bc_a4" ;
                bc_a4_CMXF:long_name = "vertically intergrated external forcing for bc_a4" ;
                num_a1_CLXF:long_name = "vertically intergrated external forcing for num_a1" ;
                num_a1_CMXF:long_name = "vertically intergrated external forcing for num_a1" ;
                num_a2_CLXF:long_name = "vertically intergrated external forcing for num_a2" ;
                num_a2_CMXF:long_name = "vertically intergrated external forcing for num_a2" ;
                num_a4_CLXF:long_name = "vertically intergrated external forcing for num_a4" ;
                num_a4_CMXF:long_name = "vertically intergrated external forcing for num_a4" ;
                pom_a4_CLXF:long_name = "vertically intergrated external forcing for pom_a4" ;
                pom_a4_CMXF:long_name = "vertically intergrated external forcing for pom_a4" ;
                so4_a1_CLXF:long_name = "vertically intergrated external forcing for so4_a1" ;
                so4_a1_CMXF:long_name = "vertically intergrated external forcing for so4_a1" ;
                so4_a2_CLXF:long_name = "vertically intergrated external forcing for so4_a2" ;
                so4_a2_CMXF:long_name = "vertically intergrated external forcing for so4_a2" ;

and we see that the variable LWCF includes information about forcing, specifically the long wave radiation. So, lets generate a new file for all months with just that one variable, i.e., we grab LWCF from all files

$ ls -1 | sort
e_slab_custom_frc.cam.h0.0001-01.nc
e_slab_custom_frc.cam.h0.0001-02.nc
e_slab_custom_frc.cam.h0.0001-03.nc
e_slab_custom_frc.cam.h0.0001-04.nc
e_slab_custom_frc.cam.h0.0001-05.nc
e_slab_custom_frc.cam.h0.0001-06.nc
...

To accomplish this we first create an interactive bigmem job:

srun --nodes=1 --time=00:10:00 --partition=bigmem --mem=10G --account=nn9817k --pty bash -i

The needed allocated time and memory will depend on the number of files used. Load all needed modules and activate the virtual environment:

module load Python/3.8.2-GCCcore-9.3.0;
module load matplotlib/3.2.1-intel-2020a-Python-3.8.2;
module load GEOS/3.8.1-iccifort-2020.1.217-Python-3.8.2;
module load PROJ/7.0.0-GCCcore-9.3.0;
module load FFmpeg/4.2.2-GCCcore-9.3.0;
source ~/.virtualenvs/p3/bin/activate

We then run

gen_agg -i "e_slab_custom_frc.cam.h0.000*" -a LWCF SWCF

if we want two files; one for the variable LWCF and one for SWCF. We probably want the reference height temperature TREFHT as well, and maybe also the total aerosol optical depth in the visible band AEROD_v, in which case we run

gen_agg -i "e_slab_custom_frc.cam.h0.000*" -a LWCF SWCF TREFHT AEROD_v

(See the full list of attributes here.)

Mimic cycle with interp_missing_month

This is not really part of the project, but kept here just for convenience.

A file that in the CESM2 model is used in cycle mode on the year 1850 can be made into produce the same input to CESM2, but in "interp_missing_month" mode.

Assuming you are in the directry of c2imp.sh and set_date_var_nc.py, and the file you want to change is in.nc, do

sh c2imp.sh in.nc

You probably also have to run the ncks command as specified in c2imp.sh.

About

Combine datasets to dataarrays, save to zipped numpy files and plot variables.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Used by

Contributors

Languages