Example applications for the Ensemble and Assimilation Tool (EAT)

These examples can be run on Windows, Mac, or Linux. They require Anaconda or Miniconda.

All commands below should be executed from a terminal window (Windows: "Anaconda prompt" in the start menu).

Installation

First install EAT. The easiest way to do this is to create an isolated eat environment with the pre-compiled eatpy package:

conda create -n eat -c conda-forge eatpy

Then add the Python packages that we will use for post-processing:

conda install -n eat -c conda-forge jupyterlab ipympl netcdf4 cmocean scipy

Running the experiments

To get started with the example applications:

1. Activate the EAT environment by executing conda activate eat
2. Open JupyterLab by executing jupyter lab. The JupyterLab environment should now open in your browser.
3. Within JupyterLab, open the start.ipynb notebook.
4. Click one of the three examples. In the notebook that opens, execute the cells one by one.

Provided files

Each application directory contains the following:

• experiment.ipynb: the Jupyter Notebook that runs the experiment
• reference: a directory with the physical and biogeochemical configuration for the reference (free running) simulation
• forcing: a directory with forcing files (e.g., meteorology) used by all simulations
• observations: a directory with observations to assimilate
• da*: one or more directories with run scripts for data assimilation experiments