ice-eddy

Ice-Eddy interaction project in collaboration with Josue

The github contains the directories to copy on the home directory

Architecture of the project

on the home dir,

put there the 3 folders: INIT, RUN, DIAGS.

in the INIT folder,

create initial conditions using the Initiate_vortex.ipynb notebook: Choose parameters and name of the simulation

--> It creates a folder on datawork, in the ...../init_cond/ repository, that contains init_julia.nc, the init file, and parameters.txt with all parameters of the simulation (eddy, grid, etc.)

in the RUN folder,

launch the julia simulation

• adjust path_root in run.jl, and other parameterizations if needed : /!\ the drag and the closure scheme are changed in this file, put a if condition with the name of the experiment to save the choice !! and be able to re-run it quickly
• adjust name of simulation in submit2GPU.pbs
• qsub submit2GPU.pbs

--> it will run the simulation on GPU, with nothing more needed because everything is written in the parameters.txt file

--> all will be stored on the datawork dir ....../RUN/ in the correct folder

--> you can follow the simulation step in run.log file

note that there is a readme with some tricks, update it with new infos

in the DIAGS folder,

put notebooks to do diagnostics/plots or whatever wanted

on the datawork dir,

create 2 directories, init_cond and RUN. All the stuff will be stored there by the aforementioned actions.