Neural Operators as Surrogates for Coastal Ocean Modeling with MIKE 21 HD

Type of project

M.Sc.

Contact

Jesper Mariegaard (jem@dhigroup.com)

Topic

Neural operators are a new class of machine learning models that learn mappings between function spaces (e.g. inputs and solutions to PDEs), making them well-suited as surrogates for physics-based models. This project explores whether neural operators like the Fourier Neural Operator (FNO) or U-NO can emulate MIKE 21 HD simulations on flexible meshes. Two test cases, Southern North Sea and Øresund, will be used for model training and validation.

Relevance to DHI

Flexible mesh models like MIKE 21 HD are powerful but computationally expensive. Neural operators offer a promising way to emulate these models with fast inference while preserving physical consistency. This could unlock new use cases in ensemble forecasting, optimization, and real-time decision support.

What is the student's contribution

Implement and train one or more neural operator models on MIKE 21 HD data. Evaluate surrogate accuracy, generalization, and runtime. Compare performance and interpretability to an existing simpler surrogate framework FM-ROM. Deliverables: trained models, reproducible code, and a feasibility report.