simple_xml

This repo is used as a starting point for tutorials but feel free to use it as a template for ocean_only or ice_ocean configurations.

Installation

quick and dirty way:

git clone https://github.com/raphaeldussin/simple_xml.git

If you're confortable with git, you can also create your own repo by clicking use as template and populate it with your own xmls.

Slides

Slides from the 2022 MOM6 tutorial are in the TutoMOM6_FRE pdf file. Companion video is available on the MOM6 youtube channel

Compile and run the examples:

Phillips 2 layers

module load fre/bronx-21

fremake -x ocean_only_experiments.xml -p ncrc5.intel23 -t prod MOM6_compile
frerun -t prod -p ncrc5.intel23 -x ocean_only_experiments.xml Phillips_2layers_example

Baltic

module load fre/bronx-21

fremake -x ice_ocean_experiments.xml -p ncrc5.intel23 -t prod MOM6_SIS2_compile
frerun -x ice_ocean_experiments.xml -t prod -p ncrc5.intel23 Baltic_025_example

Exercices:

create a new Phillips experiment that inherit from the existing example, copy the MOM_override into the new of the xml and make it a f-plane experiment (i.e. BETA = 0.)
edit the Baltic example and replace the current diag_table by your own, in which you will save daily output for SST (tos), SSS (sos)
the Baltic sea has a maximum depth of 500 meters, finish the code snippet to build a custom diagnostic vertical coordinate with 5 meters resolution from the surface to 100 meters and 25 meters below 100 meters.

import numpy as np
import xarray as xr


interfaces =

thicknesses =

# define netcdf variables and write to file
vertcoord = xr.Dataset()

vertcoord["z_i"] = xr.DataArray(interfaces,
                                dims=("z_i"),
                                attrs={"long_name": "Interface target depth",
                                       "units": "m"})
vertcoord["dz"] = xr.DataArray(thicknesses,
                               dims=("z_l"),
                               attrs={"long_name": "z* coordinate level thickness",
                                       "units": "m"})

vertcoord.to_netcdf("diag_z_baltic.nc", encoding={"z_i": {"_FillValue": 1e+20},
                                                  "dz": {"_FillValue": 1e+20}})

Add this new coordinate to the DIAG_COORDS and add thetao and so to the diag_table.

advice: you can use my python install

alias ipython=/lustre/f2/dev/Raphael.Dussin/miniconda3/envs/repro/bin/ipython

if you're having issues creating the file (or we're running out of time), use the one I already created:

/lustre/f2/dev/Raphael.Dussin/input/MOM6tutorial/Baltic/diag_z_baltic.nc