Turbulent Diffusion in the Sea

The diffusion equations of Goldstein-Michelson and Joseph-Sendner are examined in the light of experimental evidence derived from a Stommel-Richardson-type experiment. It is concluded that the Goldstein-Michelson theory is approximately valid for the small-scale experiment performed, but that modifications may be necessary for a proper description of diffusion from an initially very small to a terminally very large scale. The velocity autocorrelation function implicit in the Goldstein-Michelson theory, if appropriately modified, could be incorporated into a diffusion equation which would describe the empirical increase in diffusivity with increasing scale.