New Oceananigans paper!

[glwagner]

Hello Oceananigans community,

I am very happy to announce that we have just submitted "High-level, high-resolution ocean modeling at all scales with Oceananigans" for review in the Journal of Advances in Modeling Earth Systems (JAMES). Check it out!

One goal of this paper is to describe the governing equations and numerical methods that underpin Oceananigans, and to showcase how simple scripts can direct a wide range of numerical experiments across all scales of oceanic motion. In this way, it serves as a long, detailed update to the shortform JOSS paper we published years ago about the nonhydrostatic model.

A second goal, however, which I think is more important than describing Oceananigans current set of features, is to argue that Oceananigans implements a vision for ocean modeling (and climate modeling more broadly) that combines three cooperative elements:

1. Simple numerical methods (we think structured grids help too);
2. Implemented efficiently on GPUs, and
3. Exposed behind an innovative and productive user interface.

This vision, which is broadly applicable, navigates the trade-offs between accessibility, achieved with simple numerical methods and a high-level interface, and accuracy, which is addressed by brute force high resolution via GPUs. Because while accuracy is essential for doing science, we must also prioritize how fast we can develop the code. Put another way: even more so than what Oceananigans can do today, it matters more what we can do tomorrow. The future of science is wide open --- hybrid AI/physics modeling, new parameterizations, new numerical methods, new hardware --- and we need to be ready.

Obviously Oceananigans is not perfect now and there is work to do not just to implement new features, but also to keep refactoring the internal code to continue to make it faster to develop. That's a crucial point for new developers --- don't just add, make the existing code better, too.

Let us know what you think!

[NoraLoose]

This is exciting! Is the GitHub repo with the simulations and figures in this paper open? In Appendix B I can see that it should live under https://github.com/glwagner/OceananigansPaper, but I get a Page Not Found error.

[glwagner]

You caught me trying to be all secretive! 🕵️‍♂️

Well, I just made it public. It's a little messy actually. I'm hoping to clean it up and make a README. I have so much free time so that should be easy.