## Turbulence in Coastal Currents

When observed currents are compared to modeled currents (courtesy Peter Oke and John Allen), it is usually found that modeled currents are faster (top 2 panels).

Comparison of modeled and observed turbulence (lower 2 panels) indicates that, although the model does include surface-forced turbulence and bottom boundary layer turbulence, its structure is quite different from observed surface layer and bottom turbulence. Perhaps even more importantly, the model is devoid of turbulence in the interior.

By isolating the shear due to internal gravity waves from the thermal wind shear (which is due to the coastal current), we can see that thermal wind shear alone is insufficient to reduce the gradient Richardson number (Ri) enough to trigger model turbulence. But the additive effect of the internal gravity wave field reduces Ri substantially.

Since models do not explicitly (and usually not implicitly either) include an internal gravity wave field they can never properly reproduce the low Richardson number condition for turbulence. More appropriate parameterizations that include an internal gravity wave field must be explored.

The lower panel shows the overlap of low Ri with high turbulence.