Diagnostics of ocean models with mom6-tools

We assume that forced ocean simulations are completed and the output is located in $SCRATCH. Next, we need to analyze the results with mom6-tools.

Conda environment

Feel free to use my environment:

module load conda
conda activate /glade/work/pavelp/conda-envs/env-from-npl-2024a

Otherwise, create your own environment as shown below.

We start by creating a conda environment by cloning one of the standard environments in NCAR:

module load conda
conda create --name env-from-npl-2024a --clone npl-2024a

The new environment can be found here:

/glade/work/pavelp/conda-envs/env-from-npl-2024a

Activate the environment and install the local version of mom6-tools:

conda activate env-from-npl-2024a
git clone https://github.com/NCAR/mom6-tools.git
cd mom6-tools
pip install -e .

Computing diagnostics

Provide the path to the experiment in experiment.yaml, and run the script:

./run_diagnostics.sh experiment.yaml

Feel free to uncomment the necessary diagnostics.

Note: short command to send computation to a compute node interactively:

qcmd -- ./run_diagnostics.sh experiment.yaml 

or as a standalone job:

qsub -v YAML_FILE=experiment.yaml pbs_script.pbs

See the results in figures:

tree PNG

Some diagnostics are stored as netcdf files:

ls ncfiles