Sensitivity of internal-tide generation to stratification and its implication for deep overturning circulations

The simple scaling relation for internal-tide generation proposed by Jayne and St. Laurent is widely used for parameterizing turbulent mixing induced by breaking of internal tides. Based on the internal-tide generation derived from a 0.1° ocean general circulation model, we show that depending on which stratification is used, this relation produces different vertical distributions of internal-tide generation. When using the buoyancy frequency at the seafloor, which is a common practice, the scaling relation produces, relative to the model, too strong internal-tide generation in the upper 2000 m and too weak internal-tide generation in the lower 2000 m. Moreover, the different vertical distributions in the different ocean basins, characterized by a generally decreasing internal tide generation with increasing depth in the Indo-Pacific but not-decreasing or even increasing internal tide generation with increasing depth in the upper 3000 m of the Atlantic, cannot be captured when using bottom stratification. These unsatisfactory features can be easily removed by replacing the buoyancy frequency at the seafloor by a buoyancy frequency averaged over a large part of the water column. To our knowledge, this sensitivity to stratification has not been explicitly quantified for the global ocean. Because of this sensitivity, the scaling relation of Jayne and St. Laurent should be used with an averaged stratification to ensure a more adequate representation of turbulent diffusivity due to tidal mixing and water mass transformation in the deep oceans.