Due to the abundance of bottom feeding organisms, the relatively thin 

 bottom boundary layer in the ocean is often called the benthic boundary layer. 

 Dynamics within the benthic boundary layer affect the animal community, the 

 sediment movement, and the diffusion of chemical species and as such have 

 received great attention from all disciplines of oceanography. Over the 

 shallow coastal waters, wave-induced oscillatory flow interacts with the 

 slowly varying current within a thin layer above the bottom to cause movement 

 of sediment and other materials. Models capable of predicting the benthic 

 boundary layer dynamics in shallow water are needed to accurately predict 

 bottom flow and resulting sediment movement. 



In the interior of a relatively large lake or ocean, the lateral 

 turbulent diffusion is generally smaller than the vertical turbulent 

 diffusion. Adjacent to the coastlines, thin lateral boundary layers exist 

 within which lateral turbulent mixing is also important. Tee (1976) studied 

 the tidal ly-induced residual currents in the Bay of Fundy and suggested the 

 importance of the lateral boundary layer in generating the large residual 

 eddies. However, most large-scale circulation models use numerical grid 

 spacings which are larger than the lateral boundary layer thickness and hence 

 do not adequately resolve the detailed boundary layer dynamics. 



Turbulence Parameterization 



One of the most important features in numerical hydrodynamic models for 

 coastal currents is the parameterization of turbulence. Most existing models 

 utilize the concept of an eddy viscosity. Eddy-viscosity models are 

 relatively easy to use and can give reasonable results if sufficient data are 

 available to establish the validity of the model parameters. Once the model 

 has been sufficiently calibrated for a given site, it is then suitable for 

 performing parametric studies or long-term simulations at that site. However, 

 the proper eddy viscosity formulation depends on both the process and the 

 environment of interest. Great discrepancy exists among the various empirical 

 formulations of eddy coefficients. There is a great need to reduce the 

 empirical "tuning" of the eddy coefficients which is often required to achieve 

 good agreement between model output and data. 



18 



