An empirical shear-based finescale parameterization of turbulent dissipation rate

An extensive microstructure data set along with hydrographic and shipboard ADCP is used to develop an empirical finescale parameterization for the dissipation rate of turbulent kinetic energy and to validate existing finescale parameterizations for the open ocean. The multi-cruise data set was collected in the tropical North Atlantic between 5 and 15 N east of 30 W, supplemented by data from midlatitudes from Polzin et al. (1995). The empirical parameterization was obtained by fitting measured turbulent dissipation rates to finescale velocity and hydrographic data. Two predictors were used: finescale shear power spectral density and internal wave slope. The parameterization is particularly appropriate for shipboard ADCP when underway, with occasional knowledge of buoyancy frequency needed, but no internal wave strain required. Exponents of the parameterization are found to be different from the Gregg-Henyey-Polzin parameterization (GHP), particularly 4/3 in shear power spectral density compared to the canonical 2. Despite the different appearance of the two parameterizations, the dissipation rates measured in the tropical Atlantic also agree well with GHP predictions. However, the data suggests a correlation between shear variance and frequency content (shear-to-strain ratio), both of them are used as predictors in the GHP parameterization. Whether this is a spurious relationship is unclear. Nevertheless, this correlation implies that when using GHP, internal wave shear and strain profiles need to be measured simultaneously. Estimates of turbulent dissipation rates using GHP, but based on shear or strain measurements exclusively, are subject to bias. For the study region in the tropical North Atlantic, the proposed parameterization allowed to enhance the dissipation rate data coverage by tenfold. The resulting estimated area-averaged dissipation rate agrees with the outcome of a simultaneous large-scale tracer release experiment