examples

Examples

This section explains the reasones for the scripts and notebooks provided in this directory.

Running the examples may require adjustment to the wrkDir variable to a directory. This will allow the example scripts can write sample output files. Within that directory, you will need an INPUT directory to simulate a MOM6 model runs input file directory.

Some examples utilize the GEBCO 2020 bathymetry. The bathymetry will needed to be downloaded and scripts edited for the storage location before they can be used. See the Bathymetry in the resources portion of the gridtools manual.

mkGridInteractive.ipynb

This is the interactive grid generator. It was created as a way to generate model grids without needing to touch the command line or scripts.

mkGridIterative.ipynb

This script demonstrates how to create a model grid using the programming language. The script also demonstrates loading and displaying existing model grids.

mkGridsExample01.py

This script demonstrates creation of a model grid over California. It is Lambert Conformal Conic (LCC) and 20x30. This also demonstrates a high level of logging at the DEBUG level and storing of messages to LCC_20x30.log. This will create graphical plots of the grid in JPG and PDF formats (if supported on your system).

For grid creation and plotting, parameters have been defined. Most paramters have a default value. If a parameter is not defined, a warning or informational message may be emitted indicating the default value that the library is using.

mkGridsExample02.py

This script demonstrates some of the logging and debugging methods integrated into the gridtools library.

mkGridsExample03.py

This example generates a Mercator grid off the west coast of California. A similar grid is constructed in FRE-NCtools, in example 03FRE, for comparison.

This also demonstrates the use of an extended grid to overcome the outer boundary artifacts generated by the regridTopo() routine.

Please see the API manual for a guide that compares gridtools and FRE-NCtools grid generation methods.

mkGridsExample03FRE.py

This script generates a Mercator grid using FRE-NCtools.

To successfully create a topography file, these things are needed for make_topog:

• Input topography/bathymetry with variable of type NC_DOUBLE, NC_FLOAT or NC_INT.
• Input topography/bathymetry with an attribute named missing_value with type NC_DOUBLE, NC_FLOAT, NC_INT, NC_SHORT or NC_CHAR.
• Ocean vertical grid

Please see the API manual for a guide that compares gridtools and FRE-NCtools grid generation methods.

mkGridsExample04.ipynb

This example demonstrates the manual reconstruction of the IBCAO grid and the use of gridtools to construct the same grid.

mkGridsExample04a.ipynb

(INCOMPLETE): This example is the same program from Example 4 except there is an attempt to use dask.

mkGridsExample05.py

This example demonstrates the manual reconstruction of the IBCAO grid and the use of gridtools to construct the same grid.

mkGridsExmaple05a.py

This example is the same as Example 5 except a slightly different radius of the Earth is chosen to show how it impacts model grid creation.

mkGridsExample06.py

This creates a mini IBCAO grid that will fit on smaller memory systems. The plot of the grid shows the grid cells.

mkGridsExample07.py

This is currently a full fledged example of creating a grid, creating a topography, initial ocean/land masks and FMS coupling and mosaic files that should allow basic MOM6 modeling operation. See the manual for an expanded description of this example.

mkGridsExample07a.ipynb

This example is just Example 7 but in a jupyter notebook form.

mkGridsExample08.py

This example exercises the gridtools.regridTopo() function that provides a bathymetry and ocean mask that is a fraction instead of 0 and 1 bits.

NOTE: This example requires a fair bit of memory to run. This example will NOT run on a machine with 8 GB of RAM.

mkGridsExample09.py

This example requires the use of ipython. The interpreter must be launched with pylab: ipython --pylab. Code copy and pasted into ipython.

After editing, the function to write out a ROMS grid does function.

In [2]: romsObj.write_ROMS_grid(romsGrd, filename='grid_py.nc')
 ... wrote  theta_s
 ... wrote  theta_b
 ... wrote  Tcline
 ...

matplotlib pylab

The backend must be able to open an interactive window. The agg backend is not sufficient:

$ ipython --pylab
Python 3.8.10 (default, Jun  2 2021, 10:49:15)
Type 'copyright', 'credits' or 'license' for more information
IPython 7.24.0 -- An enhanced Interactive Python. Type '?' for help.
Using matplotlib backend: agg

aarch64

The python venv does not support installation of pyqt at the moment. The module exists, but it fails to install via pip. The conda package manager does seem to support ipython pylab:

$ ipython --pylab
Python 3.9.4 | packaged by conda-forge | (default, May 10 2021, 22:03:40)
Type 'copyright', 'credits' or 'license' for more information
IPython 7.24.0 -- An enhanced Interactive Python. Type '?' for help.
Using matplotlib backend: Qt5Agg

mkGridsExample09a.ipynb

This example is a jupyter notebook that opens a ROMS model grid for ocean mask editing.

mkGridsExample09b.ipynb

This example is a jupyter notebook that opens a MOM6 model grid for ocean mask editing.

mkGridsExample10a.ipynb

NOTE: INCOMPLETE EXAMPLE

This example requires the use of ipython --pylab. The ipython interpreter should be started and commands from the script copy and pasted to start the editor.

Once editing is complete, there are more commands required to save the edited grid.

mkGridsExample11.ipynb

NOTE: INCOMPLETE EXAMPLE

This example might be replaced by example 12.

mkGridsExample12.ipynb

This shows the grid extension capabilities for various grid projections. It shows how model grid can be grown in any direction.

mkGridsShowLoggers.py

This scripts demonstrates how to access logging from other python modules should it become necessary for debugging problems.

NewGridMOM6.ipynb

This is all the code from a fully written tutorial about constructing a MOM6 model grid.

See: Build and Edit a MOM6 Grid in Jupyter

DIR: ./bokeh

This provides a quick demo of the software package used to create the interactive ocean mask editor for jupyter.

clickablePlot.ipynb

This demonstrates a quick editor using a simple 1D latitude and longitude grid.

clickablePlotCurvilinear.ipynb

This demonstrates a an editor using a a curvilinear grid in a 2D latitude and longitude grid. To find the grid point for each click, a great circle calculation is required.

NOTE: It may take some time to update between mouse clicks due to the large grid. The implementation of the ocean mask editor uses a subset of the grid to increase responsiveness.