klein 8640369edb Fix small bug | 2 years ago | |
---|---|---|
.. | ||
generate_new_bdy_dta.py | 2 years ago | |
generate_new_bdy_dta_onefile.py | 3 years ago | |
generate_new_bdy_geom.py | 3 years ago | |
nn_interp.py | 3 years ago | |
prepare_ecearth_hindcasts_paraso.sh | 2 years ago | |
prepare_ecearth_paraso.sh | 3 years ago | |
prepare_oras5_paraso.sh | 3 years ago | |
readme.md | 3 years ago |
Scripts for generating lateral ocean boundary condition from EC-Earth ORCA1 outputs.
Steps:
python generate_new_bdy_geom.py
. All inputs are defined on top of the script with explicit comments.prepare_ecearth_paraso.sh
or prepare_ecearth_hindcasts_paraso.sh
. All inputs are defined on top of the script with explicit comments. Be sure to use the "geometry" file that the script in 1. created.Then you should be settled, i.e. have boundary geometry and data files. Both scripts defined above are commented.
Step 2. is quite long, because one intermediate step is to remap the EC-Earth data to the full NEMO grid. It's an overkill (in the end, we're only interested in the boundary), but it's not straightforwartd to make a simple (in terms of code readability) implementation which would work without that. For example, on pelican, generating one year of monthly boundary data for PARASO will take about several hours. If you want to have daily data for PARASO (I haven't done that), you'll thus need about more than a day per year, and you'll generate intermediate files (which will be cleant off) of about 150GB. In terms of CPU, the bottleneck is mostly drowning the full 3D fields (cdo setmisstonn
in prepare_ecearth_paraso.sh
), so if you want to speed things up, start with that.
It may be worth saying that N. Jourdain (IGE, Grenoble) has a more general tool for building configs which is called BUILD_CONFIG_NEMO.