Parameterization of small‐scale processes

Oceanic general circulation models must parameterize the effect of subgrid‐scale motions and generally do so with diffusion terms. The following questions then arise:• do oceanic observations allow inferences about mixing and diffusion coefficients?—if so, • how do “observed” coefficients compare with those used in numerical models? • can mixing rates be predicted from an understanding of the processes involved? and finally • does it matter? Are the results of numerical models sensitive to the choice of the diffusion coefficients?