lpjg2nc

A new conversion tool for classical LPJ-GUESS output (.out-files) to netCDF4. As opposed to out2nc it will create a lat-lon-grid which will be quick-viewable with standard viewers like ncview or panoply. It currently reads all classical LPJ-GUESS outputs (daily, yearly, and both, column and row monthly formats.

Available Scripts

• lpjg2nc.py: This will simply convert an LPJ-GUESS out-file into a netCDF4-file, rechunked for quick viewing.
• nodejob_lpjg2nc.sh: A script to run lpjg2nc.py on a node. Take it as a template.

Reguirements

• Python 3
• netCDF4

Running it

Activate your environment

conda activate py36 (or "source activate py36" on some machines, e.g. Tetralith)

Run jobs:

python lpjg2nc.py <in-file.out> <out-file.nc>

Run job on compute-nodes

sbatch nodejob_lpjg2nc.sh <in-file.out> <out-file.nc>

When you're done

conda deactivate

Installation guide

Download a copy:

git clone https://github.com/nierad/lpjg2nc.git
cd lpjg2nc

Load Anaconda: e.g.

module load Anaconda (use "module spider Anaconda" to search for installations)

...and create an environment (py36 is the name here , but can be anything...) Note: Conda-environments will use the folder .conda in your HOME-dir. You might want to link it to some other place, as over time it can become quite large. Any conda-environment will be stored there.

conda create --name py36 python=3.6

Activate environment:

conda activate py36 (on some machines it is $ source activate py36)

Now install missing packages:

conda install netCDF4

Deactivate environment:

conda deactivate

That should do!