Interpretable forecasting of the Madden-Julian Oscillation

Notebooks / example scripts

00-preproc -- Preprocessing examples for the ERA5 reanalysis data repository on NERSC Cori and Perlmutter systems:
- 00-nc-intro.ipynb: Basics of ERA5's netCDF4 data model.
- 01-cartopy-viz.ipynb: Visualization of ERA5 data using cartopy.
- 02-data-preprocessing.ipynb: ERA5 data preprocessing examples.
- 03-webdataset.ipynb: Examples for creating an ERA5 WebDataset.
- 04-preproc-fourcastnet.py: Script to create full resolution (0.25) training, validation, and test ERA5 WebDatasets.
- env.yml: Specification to create mjonet-preproc conda environment for running these examples.

Installation

Local development

python setup.py develop

From GitHub

python -m pip install git+git://github.com/Yu-Group/mjo@main

conda

# create conda environment and Jupyter kernel
ENV=mjonet-preproc
ENV_PATH=notebooks/00-preproc/env.yml

conda create -f $ENV_PATH
conda activate $ENV
python setup.py develop
python -m ipykernel install --user --name=$ENV --display-name="Python [conda:$ENV]"