FAIR Skies Ahead: Citable and Reproducible Geoscience Educational and Research Workflows with Project Pythia

The American Meteorological Society, as with many other scientific societies and organizations, has embraced a culture of open science. In December 2022, the AMS published its Data and Software Policy Guidelines for its publications. This aligns with the Enabling FAIR Data Initiative (Stall et al., 2018). That initiative advocates and adheres to the principles of FAIR (Findable, Accessible, Interoperable, and Reusable) open data and software (Wilkinson et al., 2016).

In this presentation, we demonstrate how Project Pythia, now in its fifth year of NSF funding, follows and advances FAIR principles with regard to software, data, and citable publishing. Project Pythia provides educational and training resources for geoscientists interested in leveraging the scientific Python ecosystem via its Foundations series of executable JupyterBooks. Our current efforts, also in the form of JupyterBooks, focus on Pythia Cookbooks, which deliver advanced geoscience-specific tutorials and workflows for both educational and research purposes.

Project Pythia’s cloud-hosted set of JupyterBooks and curated third-party content enable easy and fast discovery of software methods, libraries, and publicly-hosted datasets (thus Findable). Our content is web-friendly and interactive (Accessible). We achieve Interoperability through use of open standards and a template that encourages consistent style and rich cross-referencing. Users can download and/or clone all of our material to run on any device that can run Python, or alternatively launch an interactive Jupyterlab session on Project Pythia’s Jetstream2 BinderHub instance. Our automated health-checking ensures that Project Pythia’s Jupyter notebooks do not suffer from obsolescence, making them Reusable now and in the future. Finally, Pythia Cookbooks are Citable; a Zenodo DOI is minted once an author finalizes their contribution.