The Climate Laboratory at Age 10: Reflections on a Decade of Python-Based Open Climate Physics in Teaching and Research

Ten years ago I had the opportunity to develop a new course in climate science and modeling using an open, active learning approach. This aligned well with my research interest in developing more expressive and reproducible ways to define well-organized hierarchies of climate model complexity. I wanted to teach students process-oriented thinking about the physics of climate while putting computational tools in their hands that empower them to follow their curiosity. This requires providing recipes for results rather than just pictures of results, which is a core value of Open Science.

What began with simple Python scripts distributed by email later blossomed into a modular software library for interactive climate modeling and an open-access online textbook emphasizing physical reasoning, computational reproducibility, and interactive exploration. A decade later, both software and book are in use by researchers, students, and instructors worldwide. I would like to celebrate the tenth birthday of the Climate Laboratory project by reflecting on its broad goals, hits and misses, and future outlook.

The landscape for open science has changed dramatically in the past decade, with major advances in the Python ecosystem, open publishing tools, and use of the cloud, as well as social and institutional practices. I will put my evolving work with the Climate Laboratory in context of all these changes, offer some thoughts on where we are going next, including how we can more effectively reduce barriers to Open Science by weaving FAIR principles and research tools into university curricula.