ALBERT

Description: We investigated turbulence and vertical transport at the “Tommeliten” site in the Norwegian sector of the central North Sea during the R/V Celtic Explorer (CE0913) cruise from 8 - 11 August 2009. The sediments at this site are rather flat, sandy and non-permeable, with the presence methane seeps, as well bacterial mats and seep-related fauna. The hydrography of the ~70 m deep water column was characterized by a mixed surface layer extending to about 20 m depth and a well-mixed ~30 m thick bottom layer that was separated by a stratified interior layer (Figure 1). Amplitudes of tidal velocities were as large as 0.3 m s-1 in the bottom boundary layer. Dissipation rates of turbulent kinetic energy (ε) etermined from microstructure shear profiles was weak (~10-9 W kg-1 - the detection limit of the profiler) in the thermocline but increased to 10-7-10-6 W kg-1 approaching the sea floor and the surface (Figure 1). Vertical turbulent eddy diffusivities (KZ, Figure 2) ranged from 10-6 m2s-1 in the stratified interior to 10-3 m2s-1 and 10-4 m2s-1 in surface and bottom boundary layers respectively; the pseudo-velocity, defined as t =L2 / 2Kz with L=1m (Figure 2) was on the order of hours to several weeks/months in the stratified interior. High-resolution dissolved oxygen (DO) profiles were measured with a fast galvanic AMT oxygen sensor (response time 0.2 s) mounted on the microstructure probe. The sensor is capable of resolving oxygen fine structures (1 cm scale), i.e. the structures in the stratified interior, that are completely overlooked by standard slow DO sensors (Figure 1). Vertical turbulent DO fluxes were calculated using the gradient method with locally-measured dissipation rates of turbulent kinetic energy. The average downward turbulent DO flux from the thermocline to the bottom water was estimated to be 4.4 ± 1.4 mmol m-2 d-1. The AMT sensor now allows us to resolve the before unrealized steep gradient in DO, and to properly characterize the downward fluxes. With benthic DO fluxes from chambers on the order of ~7 mmol m-2 d-1, the water column depletion should therefore be about 3-4 mmol m-2 d-1. This agrees with the observed DO concentrations of about 200 μmol L-1 (67% sat) and points to the thermocline being a significant source of DO. Previously, fluxes would have been grossly underestimated due to the inadequate response time of the traditional membrane sensors. The results of the study show that the acquisition of high-resolution constituent profiles together with local microstructure measurements are necessary to characterized the dynamics of a system with regard to constituent fluxes and to set proper boundary conditions for modeling applications.