Frequency-Based Correction of Finescale Parameterization of Turbulent Dissipation in the Deep Ocean

Among the existing finescale parameterizations of turbulent dissipation rates, the Gregg-Henyey-Polzin (GHP) parameterization is thought to produce the most accurate estimates of turbulent dissipation rates since it takes into account distortions from the Garrett-Munk (GM) spectrum using the shear/strain ratio [GRAPHICS] . The GHP parameterization, however, applies the single-wave approximation to infer turbulent dissipation rates in broadband internal wave spectra with a multiplication factor up to 3, so as to adjust the predicted value at [GRAPHICS] to the theoretical value for the GM spectrum. Because of this multiplication, the GHP parameterization overestimates the dissipation rates for [GRAPHICS] . This study explores the possibility of further improvements of the GHP parameterization and reformulates the parameterization to make it applicable to both 1) a narrowband frequency spectrum characterized by a prominent near-inertial peak ( [GRAPHICS] ) and 2) a broadband frequency spectrum like the GM ( [GRAPHICS] ). Furthermore, the performance of the modified parameterization is assessed in comparison with the GHP parameterization using the available microstructure data obtained near prominent topographic features in the North Pacific.