WHOI PO

Abstract: 

"Numerical ocean models play an essential role in understanding ocean circulation and Earth's climate. Most operational models rely on numerical methods that have been refined over decades and are known to be robust and computationally efficient. However, mathematical and computational advances have made it possible to explore alternative approaches based on high-order numerical methods, which promise greater accuracy for comparable grid resolution. But do these methods actually provide an advantage for realistic ocean simulations?
In this talk, I will compare two ocean models with fundamentally different numerical approaches: HYCOM, a widely used operational ocean model, and h-NUMO, a prototype model based on high-order discontinuous Galerkin methods. Rather than focusing on the mathematical details, I will discuss the practical questions that motivate this comparison: How do different numerical methods model the same equations? When is higher accuracy worth the additional computational complexity? And what trade-offs arise between accuracy, computational efficiency, and scalability on modern supercomputers?"