Dissipation in the Baltic proper during winter stratification

Abstract

Profiles of dissipation rates and stratification between 10 and 120 m depth were measured with a loosely tethered profiler over a 9‐day winter period in the Gotland Basin of the Baltic Sea. Supplementary measurements of current profiles were made with moored ADCPs. Temporal and spatial patterns of the stratification were observed by means of towed CTD. Shallow freshwater lenses in the surface mixed layer, mesoscale eddies, inertial oscillations, and inertial waves as part of the internal wave spectrum provided the marine physical environment for the small‐scale turbulence. Two well‐separated turbulence regimes were detected. The turbulence in the surface mixed layer was well correlated with the wind. The majority of the energy flux from the wind to the turbulent kinetic energy was dissipated within the surface mixed layer. A minor part of this flux was consumed by changes of the potential energy of the fresh water lenses. The penetration depth Hpen of the wind‐driven turbulence into the weakly stratified surface mixed layer depended on the local wind speed (W10) as Hpen = cW103/2 Active erosion of the Baltic halocline by wind‐driven turbulence is expected for wind speeds greater than 14 m/s. The turbulence in the strongly stratified interior of the water column was quite independent of the meteorological forcing at the sea surface. The integrated production of turbulent kinetic energy exceeded the energy loss of inertial oscillations in the surface layer suggesting additional energy sources which might have been provided by inertial wave radiation during geostrophic adjustment of coastal jets and mesoscale eddies. The averaged dissipation rate profile in the stratified part of the water column, best fitted by ε ∝ EN, was different from the scaling of the dissipation in the thermocline of the ocean [Henyey et al., 1986]. The diapycnical mixing coefficient (Kv) was best fit by Kv = a0/N according to Stigebrandt [1987] with a0 ≈ 0.87 × 10−7 m2/s2. The diapycnal diffusivity estimated from the dissipation rate was lower than those estimated by the bulk method.