Jupyter / Julia notebooks that illustrate some of the JuliaClimate packages working in concert.

For example, MeshArrays.jl is used to analyze global ocean transports derived accurately from gridded model output. An important requirement in climate science is to derive transports using native model grid output to allow for maximum precision e.g. in closing energy budgets.

IndividualDisplacements.jl extends this approach by providing a Lagrangian tracking framework that readily operates on any model C-grid supported by MeshArrays.jl. In our examples, model output from the MITgcm are loaded into MeshArray using functions provided by MITgcmTools.jl.

Table of content

• 1. Ocean Transports
• 0. Data Structures

The following notebooks demonstrate various standard computations related to ocean transports.

• 04_transports.ipynb uses TransportThrough() and LatCircles() to compute seawater transports between latitude bands. It plots interpolated results over the Global Ocean.
• 05_streamfunction.ipynb uses ScalarPotential() and VectorPotential() to compute horizontal streamfunction along with the divergent transport component.
• 06_overturning.ipynb computes meridional overturning streamfunctions (the MOC).
• 07_particles.ipynb computes particle trajectories that follow a gridded flow field.

• 01_MeshArrays.ipynb illustrates the main data structures defined in MeshArrays.jl (documented here), the exchange of data between neighboring arrays in a MeshArray, and demo's these concepts in a simulation of diffusive transport.
• 02_climatetools.ipynb illustrates the use of ClimateTools.jl and ClimatePlots.jl
• 03_nctiles.ipynb converts binary data into meta-data-rich NetCDF files using NCTiles.jl for (1) a simple rectangular grid written to single file; (2) a tiled model domain on a less simple grid written to multiple files.

• MeshArrays.jl is a Julia package that handles gridded earth variables. It was introduced in this JuliaCon-2018 presentation which corresponds to 01_MeshArrays.ipynb in this folder.
• Each .ipynb notebook is paired with a .jl file via jupytext
• An interactive version can readily be spun up via the launch binder badge
• Rerunning the examples can involve data downloads into the inputs/ folder that can safely be removed afterwards
• For now, this is serial. Efficiency can be improved in various places or simply through parallelization.
• Please use the repository issue tracker for queries, bug reports, new contributions, etc.