Examples, Notebooks

There are several ways to use MITgcm.jl to e.g. analyze model results or run a new model simulation. The notebooks presented here focus on running MITgcm mostly.

Tip

Compiling MITgcm requires a fortran compiler. This is a requirement for all notebooks except MITgcm_configurations.jl.
The ECCO-Docker image has MITgcm.jl pre-installed, as well as gfortran, MPI, and NetCDF allowing to run any MITgcm configuration. The ECCO-Binder instance (free, but small) is available to try functionalities in the cloud.

Notebooks Listing

The following notebooks demonstrate core features of MITgcm.jl, for running MITgcm and analyzing results.

MITgcm_worklow.jl (code link): build, setup, run, and plot for a chosen standard MITgcm configuration.
MITgcm_run.jl (code link) : a detailed look into compiling and running the model.
MITgcm_configurations.jl (code link); explore MITgcm configurations and their parameters.
MITgcm_scan_output.jl (code link) : scan run directory, standard output, read grid files, and vizualize.

The HS94* series of examples need to be run in sequence, as they rely on output from one another. This tutorial runs the Held and Suarez 94 benchmark with MITgcm on a cube sphere grid, and illustrates particle tracking in the Atmosphere using MeshArrays.jl and IndividualDisplacements.jl.

HS94_animation.jl (code link) : run hs94.cs-32x32x5 configuration, read output, interpolate, and plot maps.
HS94_particles.jl (code link) : compute particle trajectories from hs94.cs-32x32x5 output generated earlier.
HS94_Makie.jl (code link) : using Makie.jl instead of Plots.jl

Note

The three examples in this series have to be run in sequence. HS94_animation.jl generates files that are then used in HS94_particles.jl, which generates output in turn used in HS94_Makie.jl.

The following example uses the Darwin3 package that models marine microbes and microbial communities. See MIT's Darwin Project for more information about this effort.

Darwin3_1D.jl (code link) : one-dimensional examples.

Standard Modeling Workflow

Here is the standard way to run a model configuration using MITgcm.jl.

using MITgcm
MC=MITgcm_config(configuration="advect_xy")
run(MC)
MC.inputs

OrderedCollections.OrderedDict{Any, Any} with 4 entries:
  :main   => OrderedDict{Any, Any}(:PARM01=>OrderedDict{Symbol, Any}(:rigidLid=…
  :pkg    => OrderedDict{Any, Any}(:PACKAGES=>OrderedDict{Symbol, Any}())
  :eedata => OrderedDict{Any, Any}(:EEPARMS=>OrderedDict{Symbol, Any}(:nTx=>1, …
  :setup  => OrderedDict{Any, Any}(:main=>OrderedDict(:category=>"verification"…

rundir=joinpath(MC,"run")
sc=scan_rundir(rundir)
sc.params_grid

(usingCurvilinearGrid = false, nSy = 2, usingCartesianGrid = true, nPx = 1, sNy = 10, usingCylindricalGrid = false, Nx = 20, sNx = 20, nSx = 1, Nr = 1, Ny = 10, usingSphericalPolarGrid = false, nPy = 1)

Pluto Notebooks

Below are links to static html versions of the examples which one can open with a web browser.

If instead you wanted to run the notebooks using Pluto.jl, then you might proceed as follows:

open julia in terminal window
type the commands shown below at the Julia prompt
in web-browser, open one of the notebooks' code link using the Pluto interface.

using Pluto
Pluto.run(notebook="examples/MITgcm_configurations.jl")

Alternatively, you can run an example at the command line as, e.g.,

julia -e 'using Pluto
notebook="examples/MITgcm_configurations.jl"
Pluto.activate_notebook_environment(notebook)
using Pkg; Pkg.instantiate()
include(notebook)
'