Eddies, fronts, and filaments with spatial scales of 10 m to 10 km and temporal scales of minutes to hours are believed to play an important role in the local energy transfer from the large-scale ocean circulation to turbulence, as well as many biogeochemical processes and phytoplankton production. Submesoscale features may therefore present a key element for local ocean dynamics with significant implications for the large scale ocean circulation. Submesoscale processes are, however, still poorly understood since traditional oceanographic and remote sensing techniques do not fully capture the required small spatial and temporal scales simultaneously. Only recently, new observations combining high-resolution satellite imagery with aerial and rapid in situ observations have delivered the first detailed measurements of small-scale spiral eddies and sharp fronts. The observations reveal highly non-geostrophic dynamics and a quick transition to balanced flow. Submesoscale features are most pronounced in the surface layer suggesting a strong seasonality and regional differences. The talk will introduce submesoscale ocean dynamics and highlight the recent progress in high-resolution observations.

We present results based on an unique data-set including both micro-structure transects and ADCP-data covering several tidal cycles. Those data are collected in a highly energetic tidal inlet in the Wadden Sea during the presence of a small along-channel density gradient. Most of the time the water column is well-mixed, but significant vertical stratification occurs at the end of flood towards high water depending on the position in the channel. Those peaks in stratification are clearly associated with the onset of lateral circulation once the instantaneous lateral Simpson number (based on instantaneous values of density gradient, water depth and bed stress) becomes large enough. An analysis of the evolution equation of the potential energy anomaly shows a predominant balance between vertical mixing and lateral straining for most of the tidal cycle. In contradiction to former studies, suggesting classical tidal straining to influence vertical stratification significantly in those regimes, our data show that lateral circulation is of much greater importance, indicating a situation where the along channel density gradient indirectly causes vertical stratification via the generation of lateral gradients rather then directly due to classical tidal straining. Those lateral dynamics have strong implications for residual estuarine circulation and thus need to be taken into account when modeling similar systems. 

The horizontal kinetic energy spectrum and its budget are analyzed on the basis of a mechanistic general circulation model that is run at high spatial resolution (spectral truncation at total wavenumber 330 and a level spacing of less than 250 m from the lower troposphere to the lower stratosphere). The mechanistic character of the model is due to simplistic parameterizations of radiative and latent heating. The only subgrid-scale parameterization is a Smagorinsky-type anisotropic diffusion scheme which is scaled by a Richardson criterion for dynamic instability and combined with a stress-tensor based hyperdiffusion that affects only the very smallest resolved scales. This setup allows to simulate the transition from the synoptic -3 to the mesoscale -5/3 slope of the upper tropospheric kinetic energy spectrum. We present indications that the -5/3 range can be interpreted as stratified macro-turbulence, as has been proposed in recent works of E. Lindborg and others. In particular, the model shows a forward horizontal energy cascade in the mesoscales around 300-150 hPa that is accompanied by an equally strong forward spectral flux due to adiabatic conversion. The latter results from a vertical energy exchange analogous to that by mesoscale gravity waves. Within the troposphere, the source of the associated vertical potential energy flux (assuming pressure as vertical coordinate) is located in the mid troposphere, where the enstrophy and energy cascades maintained by baroclinic Rossby waves are strongest. A second region of stratified turbulence is identified for the lower troposphere where mesoscale energy from the mid troposphere is deposited too.

In a sensitivity experiment with a conventional vertical resolution (level spacing larger than 1 km above the boundary layer) and identical model parameters otherwise, the mesoscale -5/3 slope develops even more clearly. It is dominated by the rotational flow and does not fulfill the scaling criterion for stratified turbulence with regard to the spectral fluxes due to horizontal advection and adiabatic conversion. Since the scaling criterion for stratified turbulence is also fulfilled if the forward cascades of kinetic energy and available potential energy are comparable, a complete interpretation of the macro-turbulence in a high-resolution circulation model would require to analyze both spectral budgets.

Hypoxia in the hypolimnion of Lake Erie has been examined by assessing (i) the spatial and temporal extent of the hypoxia (e.g., July to October in the central basin, >10⁴ km²) and (ii) linking the rate of oxygen (DO) depletion to the hypolimnion thickness. However, assessing the processes driving inter-annual variability in oxygen and the small-scale temporal and spatial patchiness in DO depletion (-0.7 to +0.3 mg/L/d) remain unknown.

Data from the summers of 2008 and 2009 in central Lake Erie included that from 13 moorings with high-frequency temperature loggers, acoustic Doppler current profilers (ADCP), dissolved oxygen loggers, and more than 600 SCAMP temperature microstructure profiles and enabled us to quantify how much of the DO variability is controlled by physical processes, relative to biological processes and the sediment oxygen demand (SOD). We found that the flux of oxygen through the thermocline to the hypolimnion was equivalent to ~18% of the total oxygen depletion in the hypolimnion over the stratified period and our results highlight the importance of turbulent mixing in hypolimnetic oxygen depletion.

The role of submesoscale flows in setting stratification in the upper ocean over the annual cycle is investigated.  It is found that
simulations which resolve submescales have a deeper mixed layer and sharper transition layer.  This destratification is primarily diffusive and happens despite the presence of submesoscale restratification.  The diffusive flux does not have a submeoscale spectral signature and so it suggests that the presence of submeoscales has changed the larger-scale flow field to generate the stronger diffusive changes in stratification.  The results imply that implementing submesoscale parameterisations which only act to restratify in global climate models could accentuate existing biases in the mixed layer depth compared to observations.  As such the role of submesoscales in large-scale flows needs to be parameterised.

Ageostrophic linear stability analysis is used to investigate submesoscale mixing on different

length and time scales. The linearized Navier-Stokes equations using an unstable background

flow as basic state yield unstable wave solutions over a wide range of important parameters,

i.e. Richardson number Ri  and aspect ratio delta.  The different regimes of the instability are

discussed and compared with numerical model solutions.

Furthermore, a closure for these processes are proposed: The fastest growing mode of the linear

solution is taken as representation for the vertical structure of the submesoscale eddy fluxes

-- i.e. in terms of a diapycnal diffusivity K and an eddy streamfunction B. Maximum growth rate

and corresponding wave number of the fastest growing mode yield the magnitude of K and B.

Model simulations are used to validate the parameterization by comparing the diagnosed

eddy fluxes with their parameterized counterparts. It turns out that vertical structure and

magnitude of the eddy fluxes are well captured over a wide range of Ri and delta.

Sea are presented for a time when the water column has a distinct two-layer structure. Bulk shear estimates, based on ADCP measurements, show a bulk shear vector that rotates in a clockwise direction at the local inertial period, with episodes of bulk shear spikes that have an approximately twice daily period, and occur in bursts that last for several days. To explain this observation, a simple two-layer model based on layer averaging of the one-dimensional momentum equation is developed, forced at the surface by wind stress and damped by (tidally dominated) sea bed friction. The two layers are then linked through an interfacial stress term. The model reproduces the observations, showing that the bulk shear spikes are a result of the alignment of the wind stress, tidal bed stress, and (clockwise rotating) bulk shear vectors. Velocity microstructure measurements are then used to confirm enhanced levels of mixing during a period of bulk shear spikes.Anumerical study demonstrates the sensitivity of the spike generation mechanism to the local tidal conditions and the phasing and duration of wind events.

Inspired by previous studies on upwelling current systems which have elaborated the importance of submesoscale flow in the oceanic mixed layer, we have studied the sensitivity of the upwelling system of the Caspian Sea to resolved scale of the flow. The Caspian Sea, a landlocked sea between Asia and Europe, has a meridional extension of more than 1000 km which results in great variability of climatic conditions over the sea. In summer, as a consequence of persistent northerly winds, upwelling occurs off the eastern coast of the Middle Caspian (MC). This phenomenon affects the entire area of MC, induces irregularity of the temperature field and compensates for downwelling in the western coast.

In this study, upwelling current system of the Caspian Sea is simulated using the General Estuarine Transport model (GETM) and atmospheric fluxes are computed from NCEP Climate Forecast System Reanalysis (CFSR) data. A set of numerical simulations is carried out using three models with different uniform horizontal grid spacing, i.e. 1.5, 3 and 6 km which we interpret as high resolution, eddy resolving, and eddy-permitting models, respectively. With these models, the effect of mesoscale and submesoscale structures on total transport, mean and instantaneous tracer structures, growth of meanders and filaments and velocity fields is studied. Furthermore, the role of submesoscales on upper-ocean re-stratification after the wind relaxation is investigated. Results show that increasing the resolution leads to appearance of convoluted patterns of sea surface temperature as well as increases in velocity and tracer variances. Moreover, resolving the submesoscales enhances the adaption of the upper ocean to wind conditions. This is more evident when the temporal scale of upwelling favorable forcing is in the order of a few days.

Observations of turbulence collected at 1 location in the main channel of the Marsdiep basin during spring and neap tidal conditions with a moored ADCP and with a shear velocity microprofiler are presented. Generally partially stratified estuaries, tidal straining results in well-mixed conditions during the flood phase and increasingly vertically stratified conditions during the ebb phase (Stacey et al. 1999; Peters & Bokhorst 2000; Chant et al. 2007). The bottom-generated turbulence is strong in the water column during flood and dampened during ebb due to the asymmetry in vertical stratification. In the German part of the Wadden Sea, these dynamics have also been observed (Becherer et al. 2011) and modeled (e.g. Hans Burchard & Hetland 2010). At the specific measurement location in the Marsdiep channel, we find the opposite. Vertical stratification is strongest during the late flood phase of spring tidal conditions resulting from lateral processes similar to the mechanism described in Lacy et al. (2003) for the Suisun Bay in California.

In the Marsdiep, the lateral density gradients have a similar magnitude as the horizontal density gradients (De Vries et al., in prep.). The inlet of the Marsdiep basin is relatively wide (5000 m) and deep (20-40 m). Flow velocities are up to 1.8-2.0 m/s during spring tidal conditions with large lateral shears and strong cross-stream velocities (0.4 m/s) which vary in magnitude and direction on an intra-tidal timescale. Besides the contrasting timing of vertical stratification with respect to other partially stratified estuaries in the Wadden Sea, the asymmetry in bottom-generated turbulence also deviates from the general trend. In addition, the shear stress and eddy viscosity have their greatest values during the ebb phase. A large variability in the dynamics of the tidal currents between neap and spring tidal conditions is observed (De Vries et al., in prep.), which has implications for the turbulence characteristics.

References     

Becherer, J. et al., 2011. Evidence of tidal straining in well-mixed channel flow from micro-structure observations. Geophysical Research Letters, 38(17), pp.2–6.

Burchard, Hans & Hetland, R.D., 2010. Quantifying the Contributions of Tidal Straining and Gravitational Circulation to Residual Circulation in Periodically Stratified Tidal Estuaries. Journal of Physical Oceanography, 40(6), pp.1243–1262.

Chant, R.J. et al., 2007. Estuarine Boundary Layer Mixing Processes: Insights from Dye Experiments. Journal of Physical Oceanography, 37(7), pp.1859–1877. Available at: http://journals.ametsoc.org/doi/abs/10.1175/JPO3088.1 [Accessed April 1, 2013].

Lacy, J.R. et al., 2003. Interaction of lateral baroclinic forcing and turbulence in an estuary. Journal of Geophysical Research, 108(C3), pp.1–15. Available at: http://www.agu.org/pubs/crossref/2003/2002JC001392.shtml [Accessed March 15, 2012].

Peters, H. & Bokhorst, R., 2000. Microstructure Observations of Turbulent Mixing in a Partially Mixed Estuary . Part I : Dissipation Rate. Journal of Physical Oceanography, 30, pp.1232–1244.

Stacey, M.T., Monismith, G. & Burau, J.R., 1999. of Reynolds stress profiles in unstratified. Journal of Geophysical Research, 104, pp.933–949.

De Vries, J.J. et al., On the spatial and temporal variability of currents and density in the Marsdiep basin.

The Dutch Wadden Sea, situated between continental Europe and the Dutch Wadden Islands, is a semi-enclosed basin composed mainly of tidal flats and sea gullies. Its dynamics are mainly governed by tides and the fresh water discharge from the sluices in the Afsluitdijk, which have a highly variable discharge rate. In order to study the overall circulation in the Dutch Wadden Sea,  three-dimensional high-resolution numerical simulations have been carried out using two different numerical models: the General Estuarine Transport Model (GETM) and Delft3D. The results of the two models are compared against each other for benchmarking. Two features of high ecological importance that must be reproduced faithfully are (1) the distribution of fresh water and (2) the flooding and drying of tidal flats. For this, an accurate bathymetric map has been constructed, and special care has been put in reproducing the discharge at the sluices as close to reality as possible. For validation, the results are compared against different observational data sets for sea surface height, salinity, temperature, and velocities.  Results on the residual circulation, the distribution of the fresh water, and their respective variability, will be presented. In addition, some first results on the validation of the flooding and drying of the tidal flats using remote sensing measurements will be discussed.

A framework for consistently connected parameterisation in ocean models is presented ranging

over the three principal dynamical regimes: small-scale turbulence, internal gravity waves, and

geostrophically balanced flow. implementation of the framework in a global general circulation

model and the effect of the different parameterisations on the large-scale meridional overturning,

ventilation rates and and watermass ages is discussed.

By using a Microstructure Turbulence (MSS) profiler, measurements of microstructure turbulence were carried out in the upper 250 m at 46 stations located in the tropical and subtropical Atlantic and Pacific oceans during the 2010 Malaspina expedition (December 2010-July 2011), Vertical temperature diffusivity (K_T) was estimated from turbulent kinetic energy and thermal variance dissipation rates measured by the MSS profiler, and also from hydrographic and meteorological data by using an adaptation of the K-profile parameterization (KPP). Empirical K_T derived from the profiler for the ocean interior (below the boundary layer) was higher in the Atlantic (0.6-3.0 · 10^-4 m² s^-1) compared to the Pacific (0.11-0.41 · 10^-4 m² s^-1), due to the higher mixing efficiency associated with salt-finger activity. Parameterized K_T showed a good agreement with empirical data in the boundary layer. In the ocean interior there was a tendency for parameterized K_T to underestimate diffusivity, except in those stations influenced by the equatorial undercurrent. The analysis of turbulence generation mechanisms in the ocean interior showed that salt-fingering was more relevant in the Atlantic compared to the Pacific, whereas shear induced and internal wave mixing were important in those stations influenced by the equatorial undercurrent. Estimates of salt diffusivity (K_S) together with nitrate gradients across the thermocline were used to compute the input of nitrate into the photic zone through turbulent diffusion. Nitrate diffusive fluxes were higher and more variable in the Atlantic (1.7-12.9 mmol m^-2 d^-1) compared to the Pacific (0.15-0.80 mmol m^-2 d^-1). The comparison of these figures with other potential sources of nitrogen into the euphotic zone showed that, at least for the time of sampling, and contrary to previous reports, biological nitrogen fixation represented a minor source (<10 %) of new nitrogen.

During the last decades, the Euler-scheme was the common ``workhorse'' in particle-tracking, although it is the lowest order approximation of the underlying stochastic differential equation. To convince the modelling community of the need for better methods, we have constructed a new test case that will show the shortcomings of the Euler-scheme. We use an idealised shallow-water diffusivity profile that mimics the presence of a sharp pycnocline and thus a quasi-impermeable barrier to vertical diffusion. In this context, we study the transport of passive particles with or without negative buoyancy. A semi-analytic solutions is used to assess the performance of various numerical particle-tracking schemes (first and second order accuracy), to treat the variations in the diffusivity profile properly. We show that the commonly used Euler-scheme exhibits a poor performance and that widely used particle-tracking codes shall be updated to either the Milstein-scheme or second order schemes. It is further seen that the order of convergence is not the only relevant factor, the absolute value of the error also is.

In summer, seasonal winds rotate from downcoast to onshore, and slightly upwelling, such that fresh water associated the Mississippi/Atchafalaya river plume pools on the Louisiana shelf and is pulled offshore.  During this time, evidence for an energetic eddy field that forms along the plume front can be seen in satellite images and hydrographic observations, and is reproduced in numerical simulations of shelf circulation.  Curiously, these eddies form when the surface mixed layer formed by the river plume is thin, while previous results indicate that submesoscale eddies are more likely to form with a larger source of potential energy associated with a thicker mixed layer.  There is evidence from numerical model simulations that these eddies enhance cross-shelf exchange, and are an important factor in exporting fresh water off of the shelf.  In this presentation, the characteristic space- and time-scales of submesoscale eddy field are estimated from available observation evidence and numerical model simulations.  Processes that control and affect the submesoscale eddy field are discussed.

One main result of the Baltic Sea Tracer Release Experiment (BaTRE) was that boundary mixing processes dominate the basin scale mixing rates within the deeper Gotland Basin. This finding was derived from a point injection of the tracer CF₃SF₃ into the Gotland Basin and its sampling over a time span of 2.5 years with a cruise roughly every six months. Due to the rather low sampling rate it was not possible to study the lateral dynamics of the tracer cloud, which is of general interest since it helps to understand the connection between boundary mixing processes and the subsequent isopycnal distribution of the mixed boundary waters. To address this question we have set up an eddy-resolving numerical model of the Central Baltic Sea. The model is validated using the BaTRE dataset and is capable to reproduce the differences between internal and boundary mixing rates. The limitations of internal wave modelling are shown by comparing the lateral resolution of the model with the wavelengths of internal waves modes in the Gotland Basin. Within the numerical model an ensemble of 39 passive tracer were released at locations slightly different from the original injection. Model results indicate that the injection location strongly influences the amount of tracer mixed out of the deeper part of the basin during the first part of the experiment. An analysis of the lateral dispersion of the tracer suggests that the lateral dispersion of the tracer patch increases dramatically when the patch grows to a size similar to that of the two gyres of the topographic wave.

Essential contribution of mesoscale processes to the vertical exchange of nutrients in the open ocean as well as in the Baltic Sea has been suggested and proved by a number of studies in the recent two decades. Latest results based on analysis of high resolution in-situ, numerical modeling and remote sensing data acquired in the Gulf of Finland showed that the submesoscale features (with lateral scales of a few km or less), such as upwelling filaments and intra-thermocline intrusions, significantly shape the distribution pattern of tracers. These processes could also substantially contribute to the vertical material exchange (create episodic nutrient pulses or subduct organic carbon) and re-stratification of the water column.

Regular measurements using ferrybox (flow-through) systems are conducted between Tallinn and Helsinki in the Gulf of Finland since 1997. Temperature, salinity and chlorophyll a fluorescence are recorded in every 20 seconds (corresponding approximately to a spatial resolution of 160 m) during two crossings of the Gulf of Finland a day. Water intake is located approximately 4 meters below the water level. Since 2009 an autonomous buoy profiler is installed close to the ferry line in the southern part of the gulf during summer months. Vertical profiles of temperature, salinity and chlorophyll a fluorescence are recorded from 3 to 50 meters with a time step of 3 hours.

The main aim of the present paper was to describe inter-related horizontal and vertical variability of temperature, salinity and chlorophyll a at the meso- and submesoscale by combining data acquired by the ferrybox system in the surface layer and by the buoy profiler through the water column. Main conclusions obtained are:  1) Wind forcing favourable for formation of upwelling events causes high spatial variability of temperature and salinity in the surface layer at the meso- and submesoscale; 2) During these events and their relaxation high variability of vertical stratification and layered vertical distribution of tracers was observed; 3) The shape of horizontal spectra for temperature and salinity seems to converge rather towards -2 than -3 slope at the spatial scales from 10 to 1 km; 4) The observed variability related to meso- and submesoscale dynamics creates high variability in Chl a distribution with patches of high Chl a both in the surface and subsurface layer.

An asymmetry between anticyclonic and cyclonic spiral eddies is observed in the submesoscale regime. The predominant cyclonic rotation, reported e.g. by  Munk et al. (2000), is not fully understood yet and contrasts with a more symmetric balance between cyclones and anticyclones observed in mesoscale dynamics. Furthermore meso- and submesoscale eddies differ in their structure,
while submesoscale eddies feature a spiral-like surface signature, mesoscale  eddies form a more closed circulation pattern, called vortexes.
In order to investigate the different dynamics, leading to different surface signatures and the observed asymmetry, numerical simulations with the idealised python Ocean Model (pyOM, Eden (2011))  were carried out. We find that ageostrophic flow components in submesoscale dynamics lead  to smaller sized cyclones, with enhanced horizontal velocity and pressure  gradients. These sharpened horizontal gradients,  which are not geostrophically balanced yield enhanced vertical velocities.  Spiral structures can only evolve on a horizontally non-divergent velocity  field and are thus also associated with enhanced vertical velocities, unlike
to nearly two-dimensional mesoscale dynamics. In contrary the submesoscale  anticyclonic regions have less enhanced gradients and vertical velocities  and thus do not favour spiral structured eddies, but closed, vortex-like structures similar to mesoscale dynamics.
An analysis of pressure perturbation, relative vorticity and vertical  velocities for different dynamical regions confirms our hypothesis
of the coherence of submesoscale cyclonic eddies and enhanced vertical  velocities, as well as the different sizes of cyclones and anticyclones  in the submesoscale  regime.

References:
 
Eden, C. (2011): A closure for meso-scale eddy fluxes based on linear instability theory. Ocean Modeling 39 (3-4) 362-369

Munk, W., L. Armi, K. Fischer, and F. Zachariasen (2000), Spirals on the sea, Proceedings
of the Royal Society of London. Series A: Mathematical, Physical and Engineering
Sciences, 456 (1997), 1217
 

follows

A new three-dimensional hydrodynamic model of the Black Sea has been set up using the adaptive vertical coordinate method implemented in the General Estuarine Transport Model (GETM). Spherical coordinates are used with a resolution of ~2x2’ and 40 vertical layers. The vertical coordinates are adapted to be sensitive to stratification to resolve the Cold Intermediate Layer (CIL) of the Black Sea proper. The model is forced with meteorological data from the European Centre for Medium-Range Weather Forecasts (ECMWF) with a temporal resolution of 6 h and with river data from the Global Runoff Data Centre (GRDC). Climatology data of the Black Sea is taken from the MEDATLAS, the Azov Sea climatology is derived from the Climatic Atlas of the Sea of Azov provided by the NOAA. Several multi-annual simulations were done for optimizing the model parameters and performing a sensitivity analysis. For model validation the monthly-averaged Sea-Surface-Temperature (SST) of the model is compared with satellite SST datasets. The generation and propagation of internal waves in the interior basin and on the shelf break is investigated in more detail.

Five years of continuous mooring data and CTD/LADCP measurements from five cruises are used to investigate the influence of the Deep Western Boundary Current (DWBC) on the internal wave field and associated vertical mixing at the continental slope at 16°N in the western tropical Atlantic. The data set resolves timescales ranging from the low frequency variability of the large scale DWBC that generates internal waves due to interactions with the topography, to high frequency vertical mixing. Diapycnal diffusivities obtained from finescale parameterizations show elevated mixing rates up to 10^-3m²/s in the bottommost 1500m during times of a strong DWBC at the moorings where velocities reach up to 50cm/s. During these periods spectra of horizontal velocity and internal wave available potential energy change substantially in depths below 1200m showing a strong increase particularly in the near inertial frequency band. The generation of low frequency, near inertial waves due to the interaction of the DWBC with the slope topography is proposed as mechanism generating the observed intensification of low frequency waves; ray paths agree well with the observed spectral changes in different depths. Variability in the high frequency range, considered as a proxy for turbulent mixing, is significantly correlated with the DWBC strength over the continental slope. Wave-topography interaction in combination with the enhanced background shear due to the high vertical shear inherent in near inertial waves could promote the breaking of internal waves, inducing the observed increase in vertical mixing.

Following Eady's model of baroclinic instability and Blumen's
fundamental work on interactions between thermal waves, the so-called
surface quasi-geostrophic (SQG) framework has evolved. This describes
the approximately two-dimensional turbulent behavior of temperature or
density anomalies present on a boundary. In recent years, SQG has
gained much attention as a possible description for submesoscale
processes in both the troposphere and near-surface ocean.  We will
discuss the SQG framework, then consider the application to both
systems. In the atmosphere, SQG produces a kinetic energy spectrum
like that observed from aircraft. In the ocean, SQG can be used (in
combination with standard baroclinic modes) to reconstruct ocean
currents and density down to 1 km below the surface, given satellite
measurements of surface temperature and surface height.

The TKE dissipation rate was measured at 38 stations of the northern South China Sea (SCS), between 18°N and 22.5°N, and from the Luzon Strait to the eastern shelf of China. The data indicates that in the summer season the average dissipation rate in the upper pycnocline <eps_p>  at the deep stations to the north of ~ 20°N is about twice as that at the stations to the south of ~ 20°N. The northwest propagation of internal waves from the Luzon Strait and the transformation of energy therein to turbulence are attributed to these observations. Approximately linear increase of <eps_p> with the estimates of available potential energy of internal waves  suggests the characteristic time scale of  dissipation close to 6 hrs.

On the shelf stations, the averaged  and depth-integrated  dissipation {eps} were analyzed for the pycnocline and for the bottom boundary layer (BBL), which was ~ 10 – 30 m in height. It appears that {eps_BBL} dominates the dissipation in the water column below the surface layer, where it reaches 17 – 19 mW/m². In the pycnocline, the integrated dissipation {eps_p} was ~ 10–30% of {eps_BBL}, but {eps_p} = {eps_BBL} was also observed at times, which could be linked to turbulence induced by propagating internal solitary waves.

A weak dependence of bin-averaged dissipation  on the Richardson number was noted, according to[eps]=eps₀+eps_m(1+Ri/Ri_cr)^-1/2, where eps₀ and eps_m are the background values of [eps] for very strong and weak stratifications, respectively, and Ri_cr = 0.25, pointing to the combined effects of shear instability of smallscale motions and the influence of a larger-scale low frequency internal waves. This broadly agrees with the MacKinnon-Gregg scaling for turbulence dissipation in the SCS pycnocline.

A host of fundamental mathematical researches have recognized discretization of advection terms as the main source of numerical errors. This effect is also investigated in a wide range of Geophysical Fluid Dynamics applications from engineering scale to large, synoptic scales. However, developing new effective advection schemes in Applied Mathematics is still an ongoing project and their roles in flow formation and evolution is not well verified. Furthermore, the accelerating progress of computing technology in addition to new suggestions in diagnosing numerically induced effects on basic properties of flow (e.g., mixing) motivated us to implement some of the most novel celebrated advection schemes in a single ocean model (General Estuarine Transport Model) and study their mechanisms in submesoscale flow regime. In these scale ranges, fast development of flow and shorter time scale of eddies are prominent in comparison to larger scale. These features highlight the essence of applying advection schemes that are more shape preserving and are also able to ignore the presence of discontinuities for flow simulations at the mentioned scales.

Herein, an analytical solution, Honeycomb, is derived as an initial value problem to verify the dissipation rate and shape preserving properties of advection schemes in barotropic mode of the model. A zonal re-entrant channel with linear vertical and meridional buoyancy gradient is also designed as a second test case to study the sensitivity of baroclinic instability in a spin-down process to different advection scheme scenarios.

We review the energy balance of internal gravity waves in the ocean main thermocline. Most of what is known about the spectral balance applies to the spectral model GM76. The evaluation of the scattering integral for resonant wave-wave-interactions acts as backbone in all concepts about the spectral balance. We give an overview of the processes and transfer rates for wave-wave-interactions in GM76. The idea is to fill the sinks induced by the scattering integral with external sources and the sources of the integral with dissipation by wave breaking. This concept is old but has gained new interest in the community in search for a energetically consistent parameterization of the diapycnal diffusivity induced by wave breaking.

A new model for the diapycnal diffusivity has recently been proposed by Olbers and Eden (JPO 2013) and tested in local and global settings. The model  (IDEMIX; Internal wave Dissipation, Energy and MIXing) is based on the spectral radiation balance of the wave field, reduced by integration over the wavenumber space, which yields a  set of balances for energy density variables in physical space, governed by advective differential equations (advection is by a mean group velocity). The flux of energy to high vertical wavenumbers is parameterized  by a functional derived from the wave-wave scattering integral of resonant wave triad interactions, which also forms the basis for estimates of dissipation rates and related diffusivities of ADCP and hydrography fine structure data. We review the construction and performance of IDEMIX 1.0, where a single energy compartment for the wave continuum is forced by wind-driven near-inertial motions  and baroclinic tides, radiating waves from the respective boundary layers at the surface and the bottom into the ocean interior. An extension is IDEMIX 2.0 which models the generation and propagation of low-mode waves (baroclinic tides and/or near-inertial waves) and interactions with the wave continuum. The models predict plausible magnitudes and three-dimensional structures of internal wave energy, dissipation rates and diapycnal diffusivities in rough agreement to observational estimates.

Eddies are a common feature in the oceans with diameters ranging from millimeters, to hundreds of kilometers. The smallest scale eddies, due to turbulence, may last for a few seconds, whilst the larger features may persist for several months. Those eddies which are over 50 km in diameter, and persist for periods of days to months are commonly referred to as mesoscale eddies and are a common feature of the Leeuwin Current System, the dominant oceanic boundary current in the region. Here, the Coriolis force is dominant and therefore the Rossby number, R_o << 1. Smaller scale eddies with less than ~40km diameter are defined as sub-mesoscale eddies, with R_o of order 1, and persists for periods of 2-3 days and are now believed to be important features for nutrient cycling in the upper ocean. Along the Rottnest continental shelf, data from the WERA HF Radar system, indicate the presence of these eddy structures in which the water depths influence the vertical scale and are defined here as peddies (‘petite’ eddies). In this presentation I will provide examples of selected peddies including those which were sampled with a ship and an ocean glider. In many cases satellite imagery have shown these features to have eddy centres and/or perimeters which had different water properties (temperature, chlorophyll) to those of the adjacent ocean. Shipborne ADCP and CTD transects across one of the peddies indicated max current speeds up to 0.5 ms^-1 with colder water upwelling at the centre of the peddie which was associated with high chlorophyll at the surface and sub-surface. It is shown that, along the south-west Australian shelf, majority of the peddies occur on the interface between the southward flowing Leeuwin and northward flowing Capes currents due to the horizontal shear between the two current systems.

Gravity waves at the ocean surface interact with the underlying flow to produce turbulence. It has been suggested that this Langmuir turbulence is an important process over the global ocean, affecting mixed layer structure and evolution, and must be included in turbulence parametrization schemes. One difficulty encountered when producing parametrization schemes is how to account for high order turbulence moments (such as pressure-strain and turbulent transport terms) which appear in the budget equations for low order turbulence statistics.

Using large eddy simulations (LES) of Langmuir turbulence for a range of forcing conditions we find that pressure-strain terms are the dominant sink within the second-moment turbulence budgets. To develop a model for the pressure-strain terms we separate them into mean shear ('fast'), wave ('Stokes') and turbulence ('slow') contributions. The LES 'fast' pressure-strain term is found to show good agreement with existing parametrizations, albeit with different coefficients, despite mean shear production of turbulence being smaller than the wave-driven production in Langmuir turbulence. The 'Stokes' pressure-strain term has not been investigated before to our knowledge. We find that by modifying the 'fast' parametrization to respond to changes in wave forcing rather than mean shear, the LES Stokes pressure-strain can be modelled reasonably well, particularly near the surface where the term is at its largest. The 'slow' pressure-strain term is commonly assumed to drive the turbulent flow towards isotropy. The most popular parametrization for this process is the Rotta return-to-isotropy model, which is a linear function of the local anisotropy. It is found that the Rotta model does not provide a good estimation of the 'slow' pressure-strain term in Langmuir turbulence. We propose a modified return-to-isotropy model which also takes into account the inhomogeneity of the turbulent flow. This new model is compared to LES results of the slow pressure-strain. In summary, a model for pressure-strain terms in Langmuir turbulence is developed using a combination of large eddy simulations and existing models, with additional terms accounting for wave effects and inhomogeneity.

Stratification and de-stratification processes in tidally energetic, weakly stratified regimes in the Sylt-R∅m∅ channel in the South-Eastern North Sea have been investigated in this study. In order to figure out the complexity of this system, a three-dimensional circulation model (General Estuarine Transport Model), with high resolution was used to simulate field observation of stratification during 3-day campaign in April 2008. A comparison of observation and model results shows that the model is able to reproduce the processes of stratification and de-stratification. Analysis of potential energy anomaly is investigated to find out the major terms and their roles in the generation of vertical stratification in the tidal inlet. Despite the classical estuarine circulation, model results show onset of stratification during early and late flood which is in good agreement with observed data. It could be shown that the diurnal tidal components lead to enhanced vertical mixing in each second tidal cycle in such a way that early flood stratification is prohibited.

Sea coastal flow is challenging for numerical models especially for vertical mixing parametrization. These regions are in general shallow and characterized by complex geometry. The presence of the coastline, rapid varying bathymetry and anthropic structures introduce complexity in the flow field, making it essentially three-dimensional.

The LES-COAST is a numerical model developed to work on coastal area (Roman et al. 2010). It solves the unsteady, non-hydrostatic, three-dimensional form of Navier-Stokes equations under Boussinesq approximation. The model works on curvilinear structured grid with second order accuracy both in space and time. Complex geometry is handled by using the technique of the Immersed Boundary Method, see for details Mittal and Iaccaino 2005. For turbulence closure we make use of Large-Eddy-Simulations (LES), it means that the large energy carrying turbulence structures are directly resolved, while only the small ones, more isotropic, are parametrized. Specifically we use a Smagorinsky two eddy viscosity model, one for the horizontal directions the other for the vertical one, in order to face with grid anisotropy. The model can be nested at the boundaries with large circulation models to provide realistic boundary conditions, using a technique based on turbulent kinetic energy. Wall effects are resolved using wall function, see Roman et al. 2009. The model takes into account buoyancy effect due to active scalars like salinity and temperature.

Due to the computational cost, the model works on horizontal length scale of the order of some kilometers, with a grid resolution spanning from 1 to 10 meters, the vertical resolution is of order 0.50 m. Because of these features, LES-COAST can be an efficient tool to study sea coastal circulation and environmental impact of pollutants. Pollutants, depending on their chemical and physical properties, are treated with an Eulerian approach (as scalars), or using Lagrangian particles. For example, the model includes a module to study oil spill problem in the sea. The adopted approach is hybrid, it means oil film treatment for light oil and Lagrangian particles for the analysis of tar trajectories.

Numerical results are in good agreement with observational data and with drifter trajectories. Two applications of the model in semi-closed basin will be shown: the Barcelona Harbour (Spain) and the Muggia Bay (Italy).

- Roman, F., Stipcich, G., Armenio, V., Inghilesi, R., Corsini, S., Large eddy simulation of mixing in coastal areas. Int. J. of Heat and Fluid Flow, 2010, 31, 327-341.

- Mittal, R., Iaccarino, G., Immersed boundary method. Annu. Rev. Fluid Mech., 2005, 37, 239-261.

- Roman, F., Armenio, V., Froelich, J., A simple wall layer model for LES with IBM. Physics of Fluids, 2009, 21, 101701.

Turbulence is important for mixing processes in the ocean. Especially in shallow water regions with high current velocities, e.g. in a tidal channel, vertical mixing is strong. The origin and continuity of turbulence depends strongly on the Reynolds Stress τ and the velocity shear S.
This work presents data, on the interaction of the Reynolds Stress τ and the velocity shear S in a tidal channel and identifies the dominating element in turbulence production. Both variables were estimated using current measurements from a bottom-mounted high-frequency Acoustic Doppler Current Profiler (ADCP). The measurements were conducted at a Time Series Station in an East Frisian tidal channel in the southern North Sea. Data were collected during one tidal period in November 2010.
The comparison between the Reynolds Stress τ and the velocity shear S reveals a period, which is related to the quarter-diurnal pattern M₄. The maximum values were measured close to the the seabed, suggesting that turbulence is generated by sheared currents at the bottom of the sea. Thus, shear acts to increase turbulent mixing. Due to the shallow water depth turbulent mixing occurs across the entire water column.

The deep cycle of turbulence in the equatorial Pacific is related to the notion of self-organized criticality, wherein a forced-dissipative system spontaneously attains a state of near-criticality. This state is characterized by scale-invariant geometry (like the turbulent inertial subrange).

The origin of the Pacific deep cycle has been a mystery since its discovery 25 years ago. In a layer of thickness 20-60m, between the base of the surface mixed layer and the core of the Equatorial undercurrent, the mean value of Ri remains at ¼ (the critical state) for three months of the year, while local values fluctuate between <1/4 and >1/4.  The origin of the fluctuations has been controversial, but they display a distinct diurnal cycle despite being apparently far removed from surface influences. 

Ri<1/4 suggests shear instability, but to explain the growth of instabilities in this regime, we must allow instability growth to be affected by pre-existing turbulence. The classical (Taylor-Goldstein) theory for inviscid, laminar flow has recently been extended to include the effects of ambient turbulence. Here, the new theory is applied to 7 days of measured profiles of currents, temperature and salinity. The results show that instability displays a diurnal cycle, both near the surface and in the deep cycle layer. Moreover, its timing suggests that shear instability is the trigger for the deep cycle.

This leads us to a scenario in which, over thousands of kilometers of the equatorial Pacific, a wind-driven, sun-stabilized shear layer forms at the surface every day, then becomes unstable and begins to entrain the fluid below as the solar heat flux wanes in the afternoon. The shear layer continues to descend, reaching the deep cycle layer around midnight. At all depths, it adds to the ambient shear, driving Ri to values <1/4 and triggering strong instability and hence turbulence. The result is a turbulent layer, extending nearly to the undercurrent core, that lasts well into the next day.

At the submesocale a strong physical-biogeochemical coupling exists due to similar temporal and spatial scales of the physical and biogeochemical processes involved. A swarm experiment with seven gliders equiped with biophysical sensors measuring pressure, temperature, salinity, oxygen, chlorophyll fluorescence and turbidity was conducted in early 2013 in the upwelling region off Peru. The goal of the experiment was to sample a relatively small spatial area as synoptically as possible, in order to allow seperation between temporal and spatial variability. In particular, the role of submesoscale processes for the near coastal vertical oxygen distribution was assessed. Each glider carried out about one dive per hour measuring two multi-parameter profiles with a lateral  resolution of about 500 m. Alltogether more than 15.000 profiles were recorded during the two-months deployment.

The glider data shows small scale filamentary structures of different tracers such as salinity and oxygen at different depths. There are density compensated salinity filaments at about 150 m depth were lateral mesoscale eddy stirring of the lateral background salinity gradient is suggested as the main formation mechasim. Closer to the surface, however, non-density compensated salinity and oxygen intrusions reaching well below the mixed layer can be seen. These structures are found at strong lateral mixed layer density fronts, suggesting that submesoscale frontal processes are responsible for the observed structures. This findings suggest that beside diapycnal mixing vertical velocities due to submesoscale frontal processes lead to an additional oxygen supply from the oxygenated mixed layer into the oxygen minimum zone.

As part of an ongoing series of experiments, surface drifter experiments have been carried out during the summer months from 2010 to 2013, at locations in the Gulf of Finland near the city of Tallinn. In 2010 and 2011, a total of 34 passive surface drifters were deployed, 2 or 3 drifters at a time, providing a position record every 15 min for a duration of up to a month. Drifter trajectories were analyzed with respect to both absolute dispersion (i.e. the distance from the initial point as a function of time) and relative dispersion (pair separation). Results indicate that the drifter dispersion exceed the values estimated from a 2-nmi model simulation of currents in the Gulf of Finland. However, results for relative dispersion are very dependent on the initial separation of the drifter pairs. Further experiments are taking place during the summer months of 2013. These results provide valuable information about small scale, horizontal turbulent motion, which is an important parameter for ocean circulation and trajectory models.

High resolution optical and radar imagery display abundant mesoscale and submesoscale (hereafter (sub)mesoscale) structures(fronts,eddies,filaments,squirts,dipoles,mushroomstructures). Besides other physical phenomena, in the Gulf of Finland these structures also result from the instability of baroclinic upwelling and downwelling jets. In the narrow elongated Gulf of Finland upwelling along the one coast is accompanied by downwelling along the opposite coast, i.e. two longshore baroclinic jets and related fronts are developed simultaneously. Wind-driven coupled upwelling and downwelling events are frequent in the Gulf of Finland. During nutrient depleted summer period, these upwelling events ventilate the surface layer with nutrient rich waters from the deeper layers and are considerable sources of phosphorus for the cyanobacterial blooms. Upwelling related filaments and eddies transport nutrients from upwelling zone offshore and instability of simultaneous downwelling jet also contribute to the mixing of water masses and distribution of phytoplankton in the open gulf area.

The coupled upwelling and downwelling events (1999 and 2006) were reproduced with the Princeton Ocean Model (POM) using high horizontal resolution (0.5 nautical miles) within the gulf, thus partly covering the (sub)mesoscale. Simulations were supported by field measurements. Two equations describing passive tracer balance were added to the POM to simulate nutrient transport. Simulations showed clear increase of apparent lateral diffusivity (up to 500 m² s^-1) in the open gulf area. To characterize the intensity of water motions of different nature simulated velocity was decomposed into (sub)mesoscale fluctuations, inertial oscillations and mean current. (Sub)mesoscale fluctuations contained 66% of the total kinetic energy. Similarity between the horizontal distribution of nutrient concentration and kinetic energy of (sub)mesoscale processes was detected. A review of the obtained results and conclusions will be presented.