SIParCS 2025 - Andrew Kwong

Enhancing Scientific Reproducibility and Software Correctness in CrocoDash for Regional Ocean Modeling

Recorded Talk

Regional ocean models provide powerful tools for studying the impacts of climate change at finer spatial scales. By simulating a subset of the global ocean, they offer insights into physical and biogeochemical processes that are often unresolved in global climate models. This work incorporates novel scientific reproducibility and software correctness features into CrocoDash, which is a Python package for rapidly prototyping regional Modular Ocean Model 6 (MOM6) cases within CESM. The new features include enhanced configurators for grid generation and CESM case creation with built-in support for Git-based version control, snapshotting, and edit history tracking. These capabilities not only streamline case setup but also lay the foundation for systematic scientific validation, enabling more transparent, reproducible, and hypothesis-driven Earth system modeling.

Mentors: Dan Amrhein, Manish Venumuddula, Alper Altuntas

Slides and poster