Anatomy of an Ocean Model
You now know what ocean models are, the types that exist, and why we need them. In this lecture, Fernand Assene opens up the model and looks inside – at the structural choices and physical ingredients that go into building one from the ground up.
This is the most detailed lecture of the week, so take it steadily. By the end, you’ll have a clear picture of what a modeller actually constructs when setting up an ocean simulation. It covers six building blocks:
- The grid – how the continuous ocean is divided into a finite mesh of points. Structured grids (regular lon-lat, or curvilinear grids that follow coastlines) versus unstructured grids (triangles of any size, refining from kilometres offshore down to metres in a harbour)
- Vertical coordinates – how the third dimension is represented: z-levels (fixed depths), terrain-following sigma coordinates (good near the seabed), and isopycnic (density-following) coordinates, often combined as hybrid systems
- The two-step construction – first the physics (the primitive equations), then the numerics (solving them on a grid, step by step in time). This is where the CFL condition appears: finer grids demand shorter time steps, driving up computational cost
- Forcing – the external drivers that make the modelled ocean move: wind stress, heat fluxes, freshwater and rivers, tides. Their quality is a primary control on accuracy – even a perfect model produces poor results if forced with poor winds
- Boundary conditions – how the domain is closed: initial conditions (the ocean state at time zero), surface and bottom conditions, and lateral boundaries, including the special treatments that let signals leave and enter an open boundary without spurious reflections
- Turbulence closure – how the model handles the processes too small for its grid to resolve (the subject of a later lecture on parameterization)
A guiding thread: every one of these is a choice, and each choice is a trade-off between realism, the processes you want to capture, and computational cost. Understanding these building blocks is what lets you read a model’s output and know what it can – and cannot – represent.
The reading and the interactive activity that follow let you explore the grids and components in more depth.
