Key points are not available for this paper at this time.
A one‐dimensional numerical ocean model of the southern Baltic Sea is used to investigate suitable parameterizations of unresolved turbulence and compare with available observations. The turbulence model is a k ‐ε model that includes extra source terms P IW and P LC of turbulent kinetic energy (TKE) due to unresolved, breaking internal waves and Langmuir circulations, respectively. As tides are negligible in the Baltic Sea, topographic generation of internal wave energy (IWE) is neglected. Instead, the energy for deepwater mixing in the Baltic Sea is provided by the wind. At each level the source term P IW is assumed to be related to a vertically integrated pool of IWE, E 0 , and the buoyancy frequency N at the same level, according to P IW ( z ) ∝ E 0 N δ ( z ). This results in vertical profiles of ε (the dissipation rate of TKE) and K h (the eddy diffusivity) according to ε ∝ N δ and K h ∝ N δ−2 below the main pycnocline. Earlier observations are inconclusive as to the proper value of δ, and here a range of values of δ is tested in hundreds of 10‐year simulations of the southern Baltic Sea. It is concluded that δ = 1.0 ± 0.3 and that a mean energy flux density to the internal wave field of about (0.9 ± 0.3) × 10 −3 W m −2 is needed to explain the observed salinity field. In addition, a simple wind‐dependent formulation of the energy flux to the internal wave field is tested, which has some success in describing the short‐ and long‐term variability of the deepwater turbulence. The model suggests that ∼16% of the energy supplied to the surface layer by the wind is used for deepwater mixing. Finally, it is also shown that Langmuir circulations are important to include when modeling the oceanic boundary layer. A simple parameterization of Langmuir circulations is tuned against large‐eddy simulation data and verified for the Baltic Sea.
Lars Axell (Fri,) studied this question.