Abstracts and Presentations/Posters
Click on title to see the full abstract text.
Follow the link shown at the end of each abstract to view the presentation (not all presentations are available).
To understand and quantify the processes causing transport of suspended particulate matter (SPM) in tidally energetic channels, high-resolution water column observations have been carried out in the Wadden Sea of the German Bight. These observations obtained during a five day uninterrupted survey in an ebb-dominated tidal channel of 12 m average depth include more than 4000 temperature, salinity, micro-structure shear and turbidity profiles from a free-falling micro-structure probe, as well as current velocity profiles from a vessel-mounted Acoustic Doppler Current Profiler (ADCP) and SPM samples to calibrate the turbidity profiles. A landward buoyancy gradient was established by a landward temperature gradient built up during a warm and calm spring season. The horizontal salinity due to saltier water in the Wadden Sea as compared to the adjacent North Sea was slightly opposing the buoyancy gradient. Tidal averaging along sigma layers (relative depth) showed during the first two days of weak wind forcing classical estuarine circulation with net inflow near the bottom and outflow near the surface. Increased westerly (up-estuary) wind during the second part of the campaign caused inverse estuarine circulation and eastward net transport (probably across the watershed towards the next tidal basin). SPM concentrations showed a strong quarter-diurnal signal with maxima near full flood and full ebb and were generally lower during the calm period and increased during the windy period, mainly due to wave-related resuspension over nearby inter-tidal flats. The sediment flux analysis was based on a decomposition of the vertically integrated SPM flux into a barotropic advective component, an estuarine circulation component and a tidal pumping component representing the covariance between SPM concentration and current velocity. As a result, tidal pumping dominated the (due to the ebb-dominance weakly seaward) SPM flux, whereas barotropic advection dominated the strong landward SPM flux during the windy period.
The main role of vertically propagating waves is to transfer pseudo
momentum from the region of generation to the region of wave
breaking. The most prominent examples in atmospheric dynamics are
planetary Rossby waves forced in the troposphere, which drive a
poleward residual circulation in the winter stratosphere, and
mesoscale gravity waves with tropospheric origin, which drive a
summer-to-winter-pole circulation in the mesosphere. In addition,
the role of energy deposition by gravity waves has long been
recognized to contribute substantially to the energy budget above
In atmospheric circulation models, mesoscale gravity waves are
usually parameterized. Their energy deposition can be computed
along with the momentum deposition and the turbulent diffusivity
associated with wave breaking. In particular, the energy deposition
is expressed in terms of secondary moments of the parameterized
waves. Therefore, one is tempted to assume that the energy
deposition of waves that are resolved in circulation models, e.g.,
Rossby waves and thermal tides, is automatically taken into account
by the dynamical core. This assumption is, however, flawed. We show
that the energy deposition by resolved waves corresponds to the
shear production (frictional heating) of the subgrid-scale
turbulence model by which these waves are damped. Computational
results from an atmospheric circulation model with energetically
consistent treatment of momentum diffusion and frictional heating
indicate that the energy deposition of thermal tides is quite
substantial above the mesopause. This effect is ignored in
conventional atmospheric models that resolve the mesopause region.
An idealized sensitivity experiment furthermore shows that thermal
tides lead to a significant downward shift of gravity-wave breaking
in the upper mesosphere.
The internal wave equation for basin scale internal waves (boundaries of no vertical flow at surface and lake bed) represents a regular Sturm-Liouville problem. Hence, an infinite number of solutions exists. These solutions – traditionally called “modes” – form a complete set of independent solutions. Any current profile matching the boundary conditions (at surface and lake bed) can be represented as the sum of modes. The distinct coefficients can be evaluated from the orthogonality relation corresponding to the internal wave equation. – In this talk, we show measured profiles from the Sill of Mainau during four internal seiche events. We decompose the current profiles and show the temporal evaluation of each of the modes, after weak winds opposed to strong winds. At the measuring site, the first two modes suffice to represent the current profiles very well. We discuss possible reasons for this result and give some hints for vertical mixing from the current profiles.
The fluxes of momentum, heat and mass across the Ocean surface are important boundary conditions for coupled air-sea models. These exchanges are controlled by complex dynamics within the turbulent wave boundary layer. We present direct measurements of wind shear stress within the first 10 cm above waves, and investigate the structure of the turbulence within the airflow. Observations were achieved for several wind speeds, wave ages and slopes, in controlled conditions in the large (42-m long) wind-wave facility at University of Delaware's Air-Sea Interaction laboratory. The wave field and the airflow above were investigated using simultaneously laser-induced fluorescence and PIV. We were able to obtain high resolution velocity measurements in the airflow inside the wave boundary layer, within and above the viscous sublayer. We observe intermittent airflow separation events past the crest of the waves. These events are usually accompanied by a dramatic drop in the measured surface viscous stress, a flux of vorticity away from the surface, and an intensification of the turbulent shear stress. Despite the intermittent aspect of this phenomenon, these events may affect the average distribution of the wind stress between viscous, wave and turbulent contributions, which impacts wave growth and the air-water momentum balance. Our results hold for wind speeds that would normally be considered low to moderate. Drag coefficient estimates as a function of wind speed will be discussed, as well as implications for models of air-sea momentum flux. Similar measurements were also achieved above waves in the field, in the Delaware Bay, USA, using a unique novel experimental approach.
The physical oceanography community is embarked on a longlasting endeavor to understand and properly account for energy dissipation and mixing into the ocean. Difficulties are outstanding, particularly in the ocean interior (i.e. below the well mixed layer). Indeed a myriad of strongly intermittent processes, often interacting with each other, are responsible for mixing the ocean interior with intensity that poses great observability challenges. A related issue concerns the dissipation of the balanced (quasi-geostrophic) flows which is, at present, poorly constrained. The degree to which this dissipation contribute to ocean mixing is also not clear.
In my talk I will present and discuss several recent studies focused on the sources of mixing in the ocean with an emphasis on small-scale kinetic energy dissipation of the balanced circulation and near-inertial motions. The context will be the deep ocean with a particular emphasis on the Southern Ocean. Current limitations of the numerical approach will be underscored.
The oceans gain most of the energy at large scales, while the extraction of kinetic energy is known to occur only at small scales by molecular dissipation. The forward energy cascade from large to small scales is characteristic of small-scale turbulence, where the energy is finally dissipated at molecular scales (Kolmogorov, 1941). However, most of the ocean is in geostrophic balance and turbulent flows in quasi-geostrophic balance show a transfer of energy from smaller to larger scales (Charney, 1971). The exact mechanism of large-scale dissipation, however, is not well understood and hence not included in the ocean models. We aim to understand and parametrize this existing discrepancy in the representation of energy cycle in the ocean models. The methods employed to approach the problem include idealized and realistic setups of numerical models as well as observations, to understand and quantify the energy pathways from geostrophically balanced motions towards large scales and (or) towards unbalanced gravity waves and small scale turbulence.
Our initial results show the existence of an ageostrophic direct route to dissipation for a dynamical regime with large Rossby number and small Richardson number (which emulates nearly ageostrophic dynamics) (Brüggemann and Eden, 2015), in accordance with the results from Capet et al. (2008c) and Molemaker et al. (2010).The energy cycle was diagnosed in physical space to understand the differences in the energy dissipation under different dynamical conditions, and in wavenumber space to identify the spatial scales of the sources and sinks of energy. A similar spectral analysis of the energy cycle can be done in frequency space to identify temporal scales of potential energy sources and sinks. We hypothesize that the energy contained in the super-inertial frequencies is much higher for nearly ageostrophic motions than for quasi-geostrophic motions. The spectra of energy in Fourier space for different dynamical regimes characterized by a range of Richardson numbers confirms our hypothesis. Hence, gravity waves could be a potential candidate for the dissipation of mesoscale eddies. The next steps involve investigating the possibility of explaining the downscale of energy by triad wave interaction and a linear stability analysis to identify how much energy projects on which linear mode which will facilitate understanding how much energy dissipation occurs from the direct generation of ageostrophic instabilities.
The efficiency of a mixing event is related to the fraction of turbulent kinetic energy transferred into irreversible diapycnal exchange. It is becoming well recognized that the mixing efficiency is not constant but rather varies with the evolution of turbulence. This idea is explored here using a combination of temperature sensors moored in the deep ocean and idealized non-hydrostatic numerical simulations. The Rockall Bank in the Northeast Atlantic Ocean is known for hosting topographically-trapped diurnal tides and, at a particular depth range around 900 m, for extensive cold-water coral mounds. It is revealed that baroclinic cross-isobaths motions, i.e. internal tides, result from the interaction between the barotropic diurnal currents and the topography. It is also suggested that internal tides are breaking on the slopes of Rockall Bank, causing increased buoyancy fluxes, which in turn leads to high mixing efficiencies.
This talk questions the reality of widely accepted wave cooling during gravity wave breaking in the mesosphere. The concept of wave cooling seems to reside on the erroneous philosophy of mixing the potential temperature instead of the temperature, which is required by the second law of thermodynamics. Experiments comparing theta and temperature diffusion do not deliver wave cooling if temperature is diffused. This is independent of the shape and the magnitude of the turbulence coefficient.
There seems to be a link between the temperature diffusion philosophy and the supplement of downgradient theta diffusion with a countergradient term in case of the convective boundary layer. The traditional countergradient approach may not be employed in the mesosphere because measures like surface heat flux and boundary layer height are undefined. It seems that the countergradient term is just needed to drive the total flux proportional to a negative temperature gradient. It is also interesting to note that the countergradient term was firstly introduced into atmospheric modeling because the upper part of a convective boundary layer was simulated too cold. This points to a similarity of problems when simulating boundary layer turbulence and turbulence due to breaking waves in the mesosphere.
Many fjords, large inland seas and oceanic basins are vertically divided by a permanent pycnocline which weakens the vertical transport of oxygen. The reduced vertical flux leads to anoxic deep waters. Prominent examples are the Black Sea and the Baltic Sea. Laterally intruding water masses are a feature commonly observed in the Baltic Sea. Their strength ranges from so called Major Baltic Inflows, which are able to renew the brackish deep water masses of the Baltic Sea, to small intrusions of water masses with a vertical size of the order of meters interleaving into intermediate depths between 80 m and 150 m. In contrast to the Major Baltic Inflows, happening on the time scale of decades, small intrusions are observed regularly in the western and central Baltic Sea. While Major Baltic Inflows have an immediate effect on the Baltic Sea and have been studied for quite a while, the role of the small scale intrusions is rather unclear. We present measurements of a long term underwater profiling mooring with a mounted CTD, oxygen and turbidity sensors and at some deployments with temperature microstructure sensors. The mooring was deployed in two major basins of the Baltic Sea: the Gotland-Basin and the Bornholm-Basin. Results show the presence of water intrusions with a low but non-zero oxygen concentration within the otherwise anoxic water masses in virtually every deployment. Turbulent exchange between the intrusions and the ambient water masses is quantified by a turbulent diffusivity derived from the temperature variance decay of the temperature microstructure. The measured rates of mixing of the intrusions indicate that the minimum lifetime is several weeks, confirming that even small scale intrusions exist long enough to laterally transport oxygenised water over long distances.
Coupled GCM simulations are analyzed to quantify the dynamic effect of Southern
Ocean (SO) winds on transports in the ocean.
It is found that the closure for skew diffusivity
in the non-eddy-resolving ocean model does not allow
for a realistic eddy saturation of the zonal transports in the SO in response to the wind changes,
and that eddy compensation of the meridional transports in the SO is underestimated too.
Despite this underestimated eddy compensation in the SO, however,
and in contrast to previous suggestions,
the Atlantic Meridional Overturning Circulation (AMOC)
strength is almost insensitive to SO winds.
In the limit of weak SO winds the AMOC waters do not upwell in the SO but in the
Through their effect on sea-ice, weaker SO winds also lead
to less production of Antarctic Bottom Water and therefore a deeper and stronger
Olbers and Eden's simple analytical model for the ocean mixed layer in particular yields an energy balance which can be used for estimating the amount of energy which is (i) provided by the windstress and (ii) transferred to the deep ocean. Because it is in principle available for diapycnal mixing, this energy is of major importance for sustaining the global meridional overturning circulation. By transformation to reciprocal space, the model equations become algebraic and can be reduced to a single differential equation for vertical velocity. Spectral properties of several relevant quantities are studied by a numerical approach using the MIT general circulation model including e.g. Gaspard et al.'s mixed layer scheme. Certain assumptions and predictions of the analytical model are checked numerically. The importance of the applied numerical mixed layer scheme is assessed.
A high-resolution ocean model and satellite-based observations are used to understand the close
interaction of waves and turbulence in the Antarctic Circumpolar Current (ACC). This analysis
reveals the importance of radiation stresses in ACC dynamics which illustrate the organisation of
turbulence by waves, giving rise to systematic, long-range momentum transports. It is then shown
that in the ACC these long-range momentum transports, in the form of baroclinic jets and standing
barotropic Rossby waves, are limited in range by major topographic features. Downstream of these
topographic features standing Rossby waves become barotropically unstable, leading to regions of
enhanced eddy kinetic energy and mixing of tracers, which are often referred to as oceanic storm
The dynamics of these radiation stresses is investigated by analysing both lateral eddy momentum
fluxes, which are responsible for the conversion of energy between mean and eddy kinetic energy
reservoirs, and the kinematic suppression of eddy diffusivities by the mean flow, which are closely
related to the eddy momentum fluxes via the well-known Taylor identity. Additionally, the redistri-
bution of eddy energy through the eddy energy flux is shown to be of importance in understanding
the evolution of radiation stresses along the ACC.
Finally, implications for the observation of hydrographic fronts and the parameterization of eddy
fluxes are discussed.
In this talk, I will show how theory of shear instability can be used to examine stability characteristics of oceanic flows. In particular, simple methods to estimate critical Richardson number and marginality of observed flows have been developed and applied. The mechanistic link between dynamic instability and generated turbulence provides a natural way of parameterizing stratified turbulence in the ocean interior. The development and applications of a new shear instability theory that includes impacts of ambient turbulence on instability development will also be introduced.
An aquatic eddy covariance (EC) system was developed to measure the exchange of oxygen (O2) and hydrogen ions (H+) across the sediment-water interface. The system employs O2 optodes and a newly developed micro-flow cell H+ ion selective field effect transistor; these sensors displayed sufficient precision and response times required for measuring turbulent fluctuations. Discrete samples of total alkalinity and dissolved inorganic carbon (DIC) were used to determine the carbonate equilibria of the water column and relate the O2 and H+ fluxes to benthic processes. The ECOHES system was deployed in a eutrophic estuary (Waquoit Bay, Massachusetts, USA), and revealed that the benthic processes were a sink for acidity during the day and a source of acidity during the night, with H+ and O2 fluxes of ±0.0001 and ± 10 mmol m-2 h-1, respectively. H+ and O2 fluxes were also determined using benthic flux chambers, for comparison with the EC rates. Benthic chamber measurements co-varied with EC measurements but were of ~4 times lower magnitude, likely due to chamber artifacts on hydrodynamics, and the depletion or elevation of DIC and O2 within the enclosed chamber. The individual H+ and O2 fluxes were highly correlated in each data set (EC and chambers), and both methods yielded H+ fluxes that were not explained by O2 metabolism alone. This ECHOES system provides a new tool for the determining the influence of benthic biogeochemical cycling on coastal ocean acidification and carbon cycling.
Terms for the mechanical energy budget of the NE Pacific shelf and the adjoining Salish Sea are calculated from a realistic hindcast numerical simulation. Terms from the ROMS history, average, and diagnostic files are used to form closed budgets for all flux divergence terms in KE and APE equations. The fields give a compact, but spatially and temporally explicit way to compare the influence of tides, wind, rivers, eddies, and coastal trapped waves on the energy fields. Energy storage terms are also calculated, split up into barotropic and baroclinic terms, as well as tidal and subtidal. Available Potential Energy of the subtidal density field is calculated as the energy required going from a state of flattened density surfaces to an observed state. It is calculated for a number of different volumes, allowing exploration of the relative size of different physical energy reservoirs. There is a very large amount of APE stored in the density structure of the inland waters, consistent with the relative constancy of the estuarine exchange flow throughout the year. The conversion between APE and KE reserviors is expressed using a modified Bernoulli function.
In this talk I will discuss the results of several studies carried out in the relatively shallow waters of inner shelves that show the presence and often dominance of internal waves that drive complex flows, and can break, resulting in substantial mixing. In particular, I will highlight our ongoing work on turbulence associated with wave breaking and internal wave shoaling, arguing that this “internal wave surf zone” is a central physical feature of nearshore ecosystems like coral reefs or kelp forests. Finally, I will how these flows can be used to study fundamental features of turbulent stratified flows. To do, I will show results from several deployments of turbulence flux towers, like those commonly used in atmospheric studies, in nearshore waters. Measurements taken from several different sites are remarkably consistent, notably showing the dependence of the turbulent buoyancy flux on dissipation and buoyancy frequency in a way that implies that the most energetic turbulent events may produce relatively less mixing than might be assumed if the efficiency of mixing is assumed to be constant as is commonly done.
Particles suspended in the sea are central to many physical and biogeochemical processes. They affect light penetration through the water column, causing a significant impact on radiative transfer and primary production. Understanding the distribution of organic carbon, in the form of Particulate Organic Carbon (POC) and how it sinks from the photic zone to the ocean floor is a key stage in the carbon cycle and therefore is directly relevant to long-term climate prediction. The behaviour of suspended particles is a function of both the particles themselves (their nature, size, shape, density, settling velocity, swimming speed etc.) as well as their physical environment (density stratification, turbulent mixing regime). Biological particles (phytoplankton and zooplankton) may exist in thin layers: high-abundance congregations extending over large horizontal distances (O kilometre) but with small vertical extent (O decimetre) controlled by physical processes within the water column. Turbulent mixing may promote the aggregation of particles, so changing their settling velocity and enhancing carbon export, whilst SPM flocs may be re-suspended during periods of enhanced turbulence. Recent model and laboratory studies show that motile phytoplankton may accumulate in the centre of turbulent vortices in a process that has a variety of ecological implications.
Therefore, the generation of effective predictive models of particle transport requires in situ information on the particle characteristics as well as on their location within the detailed physical structure of the marine environment. To this end, we have developed a free-fall particle microstructure profiler, MSS-HOLO. The system comprises a streamlined holographic imaging system (holocam) attached to an MSS-90 free-fall turbulence profiler configured with sensors that measure (at 1024Hz): turbulent shear, microstructure temperature, optical backscatter, chlorophyll fluorescence and dissolved oxygen. The holocam sample-volume (total volume of 6.7ml) is positioned directly across the front of one of the shear sensors and is imaged at a resolution of 4.4microns, allowing the measurement of all particles from 20microns to about 6000microns diameter from within the sample volume. The combined system free-falls to the seabed at 0.45m/s, while the holocam samples at 15Hz allowing the recording of a hologram every 0.03m down the profile.
The MSS-HOLO system has been deployed at a variety of locations around the West Coast of the United Kingdom, in a fresh water lake, turbid coastal waters and clear oceanic waters. The data allow examination of the fine-scale vertical structure of particle types and abundance, revealing thin regions of high abundance nano- and microplankton (15 to 30microns diameter), broader distributions of compact flocs (100 to 200microns diameter), and even very large and delicate macroflocs (500microns to 2mm diameter) in the bottom boundary layer, most probably resuspended from the seabed. These measurements provide opportunity to gain new insight into in situ particle dynamics and allow existing models of particle transport and plankton behaviour to be tested.
The interaction of small-scale internal gravity waves with a larger-scale shear flow is analyzed with the aim to reveal the origin of large vertical viscosities, of order $1\rm m^2 \ssec$, which were predicted in a WKB approach but never substantiated in observed relations between wave-induced stress $\langle \v u' w'\rangle$ and the mean vertical shear $\p\v U/\p z$. The WKB theory is presented for a wave group propagating in a horizontal current as well as for a wave continuum governed by a radiation balance equation for the spectrum of the wave field. Essential for a non-vanishing stress is vertical asymmetry and horizontal anisotropy of the wave spectrum and, most important, a relaxation of the spectral perturbation, arising from the interaction with the mean flow, back to symmetry. We give a critical review of the theory.
The shelf seas cover a relativity small fraction of our oceans but account for 15-30% of the total oceanic primary production. This production is largely dependent on small amounts of vertical mixing that provide a diapycnal exchange of nutrients into the upper water column. The seasonal structure of the water column is primarily controlled by the balance between stratification and the input of mechanical energy in the form of mixing from wind stress at the surface and, in the case of shelf seas, tidal shear at the seabed. However in most model simulations the vertical exchange schemes are based on boundary layer physics which fail to reproduce the diapycnal fluxes across this critical interface, leading to the inclusion of physically unjustified mixing schemes, compromising the predictive capacity of any simulation. This current state is unacceptable if we are to provide realistic estimates of the flux of heat, nutrients and momentum across the pycnocline. The aim of PycnMix is to produce a ‘step change’ in the representation of pycnocline mixing processes in regional scale shelf sea models. As part of this project we will present here the preliminarily results from the reanalysis of over 20 years of microstructure measurements from across the seasonally stratified North West European Shelf Seas, forming the world’s largest, consistently processed, observational database of shelf sea pycnocline turbulence measurements. The data has been interrogated in various parameter space scenarios (e.g gradient Richardson number, Froude number) and coherent behaviour identified within similarly forced flows. Here we will present an overview of the vast dataset alongside preliminary results of turbulence data analysed in dimensionless parameter space that provides a tantalising early indication that pycnocline turbulence is indeed better behaved than it often appears.
Turbulent mixing is considered to be an essential contributor to driving the global overturning circulation, but remains unresolved in ocean general circulation models and is typically parameterized by a vertical diffusion of the respective fluid properties. Instead of assuming a prescribed value for this vertical diffusivity, recently developed parameterizations involve internal wave energetics, taking into account that breaking internal waves are thought to be a major source of small-scale turbulence.
The model IDEMIX (``Internal Wave Dissipation, Energy and Mixing'') predicts the propagation and dissipation of oceanic internal gravity waves as well as the corresponding diapycnal diffusivities based on a simplification of the spectral radiation balance of the wave field and can be used as a mixing module for global numerical simulations. The aim of this study is to validate the model through a comparison with observations. Since direct observations of turbulent mixing are sparse, we follow the approach by Whalen et al. (2012) and compute finescale strain variance from Argo-float CTD-profiles to estimate the turbulent kinetic energy dissipation and the related diapycnal diffusivity.
These estimates are sensitive to the shear-to-strain ratio, which has to be set to a constant value when using CTD-data only, or the version of the Garrett-Munk model, that features as a reference in the parameterization, but these uncertainties are smaller than the general uncertainty of finescale parameterizations (factor 2-4). The spatial variation is influenced both by bottom topography and surface winds and features a noticeable seasonal cycle. Idealized simulations (without currents or temperature variations) show, that the magnitude and pattern (especially in the Gulf Stream) of the energy dissipation rate cannot be explained when meso-scale eddies and the dissipation of their energy are not simulated. The observed seasonal cycle, too, can in the model only be explained by the seasonal variations in eddy kinetic energy. A detailed fine-tuning of the IDEMIX-module will be attempted based on parameters like the bandwidth of the Garret-Munk spectrum or the symmetrization time scale of the internal wave field, using a global ocean general circulation model with a special focus on seasonal variations.
The functioning of estuarine circulation has been investigated by means of numerical models in numerous studies under idealized conditions. This led to a deep understanding of the theory of estuarine residual flows. The question remains, how estuarine circulation is established in real estuaries in response to their topographical and forcing characteristics. The present study uses a highly accurate three-dimensional numerical model simulation to calculate estuarine circulation in a curved tidally energetic channel of the Wadden Sea in the south-eastern North Sea. The results show how established forcing mechanisms of residual circulation such as horizontal buoyancy gradients, wind stress and stratification act in a combined way. In general, estuarine circulation is always positively correlated to wind stress, with frequent reversals of residual flow for wind stress directed towards higher buoyancy. Only during calm weather conditions, longitudinal and lateral estuarine circulation are highly correlated with the respective buoyancy gradients. The correlation between estuarine circulation intensity and and stratification is similar: only under calm weather conditions, estuarine circulation increases with stratification. The lateral circulation is strongly determined by a barotropic residual eddy which results from the curvature of the tidal channel: at the inner shoal, up-estuarine advection of denser North Sea water dominates such that a significant lateral density gradient is established, leading to near-bottom residual circulation towards the outer shoal. Only strong wind towards the inner shoal can revert the sense of this residual circulation.
Large eddy simulation (LES) allows to explicitly resolve all scales of turbulent motions larger than the numerical grid (the applied filter width), while only the effects of turbulent eddies smaller than the grid size on the resolved-scale flow have to be parameterized by a so called subgrid-scale (SGS) model. Using a sufficiently fine grid spacing, LES results are almost independent of the (SGS) turbulence parameterization, which makes LES an ideal method e.g. for improving turbulence parameterizations in larger scale models. For more than 40 years, LES has been a tool for fundamental research of turbulent atmospheric flows, while LES applications for oceanic flows started around 1995, still beeing on the fringes in ocean science.
We as meteorologists have used LES to intensively study the convective atmospheric boundary layer (ABL) above the ocean with a focus on coherent structures (roll-like and cellular patterns) during cold air outbreaks and how they affect the total vertical transport of heat and moisture. A further research topic has been the convective transport above leads in the marginal ice zone and the influence of specific parameters like lead width and wind velocity on this transport. Recent results show that convective rolls are not created by flow internal instability mechanisms (inflection point instability) created but by surface heterogeneities of the sea ice. Furthermore, simulations suggest that the evolving coherent structures do not increase the total vertical transport of heat.
The PArallelized LES Model (PALM), developed at IMUK during the last 15 years, has been used for these studies. Recently, PALM has been extended for an ocean option including salinity and the equation of state for seawater. The model also allows coupled simulations of the atmospheric and oceanic mixed layer (OML). The air-water interface is assumed to be flat and the coupling scheme conserves turbulent fluxes of mass (humidity, salinity), momentum and heat within each grid cell of the interface. The coupled model is applied to study turbulent interactions between the humid atmospheric ABL and the salt water OML. First results indicate that the convective structures in the ABL and the OML are coorganized.
The presentation will give a short introduction to LES and the PArallelized LES-Model PALM, including the atmosphere-ocean coupling and will then focus on past and present studies with PALM of transport processes in the CBL above the ocean, as well as on effects of air-sea interactions on the atmospheric and oceanic fluxes.
Turbulent kinetic energy cascades in fluid dynamical systems are usually characterized by scale invariance. However, subgrid-scale (SGS) parametrizations of in large eddy simulations do not necessarily fulfill this constraint. Up to now, scale invariance has been considered only in the context of isotropic, incompressible, and three-dimensional turbulence. Here we extend the theory to anisotropic turbulence in compressible flows that obey the hydrostatic approximation. We present a criterion to check if the symmetries of the governing equations are correctly translated into the equations used in a numerical model including the corresponding SGS parametrizations (model equations). We validate the criterion by recovering the breakdown of scale invariance in the classical Smagorinsky model and by confirming scale invariance for the Dynamic Smagorinsky Model. We further apply the criterion to the primitive equations completed by horizontal and vertical diffusion as used in a GCM. Our assumption is that the numerical resolution extends into the macroturbulent inertial range of the mesoscales, which is governed by a forward energy cascade. The aforementioned criterion then allows us to formulate both the horizontal and vertical mixing lengths for the free atmosphere in accordance with scale invariance. High-resolution runs with the Kühlungsborn Mechanistic General Circulation Model (KMCM) using triangular spectral truncation at wavenumber 330 are presented, being the first simulations of a -5/3 slope of the kinetic energy spectrum in the upper troposphere and lower stratosphere without numerical dissipation or hyperdiffusion. In particular, a dynamic vertical mixing length leads to a steepening of the spectrum in the synoptic scales and a shallowing in the mesoscales.
Tidal straining is known to be a crucial factor for the generation of residual currents and sediment transport in estuaries and Regions of Freshwater Influence (ROFIs) of the coastal ocean. Essential for this process is the presence of a lateral density gradient, resulting either from freshwater runoff or from differential heating. Here, we show that near sloping topography, tidal straining triggered by the projection of a purely vertical density gradient onto the slope may be observed even in the absence of freshwater runoff and differential heating. A one-dimensional (slope-normal) numerical model is applied to capture the basic mechanisms and implications of this new process, focusing on residual currents and sediment transport. We identify the key non-dimensional parameters governing the problem, and use them to examine a wide range of physically relevant parameters. Our main result is that for numerous parameter constellations, tidal straining results in an effective upslope transport of sediment with increased transport rates for coarse (quickly sinking) sediments. This process may be relevant in a broad context as it does not require lateral density anomalies caused by a ROFI or by differential heating, in contrast to classical tidal straining.
Thirty years after its discovery, the deep cycle of equatorial turbulence remains poorly understood. Large-scale models are unable to represent it accurately, leading to critical errors in ocean heat uptake. Since 2005, moored turbulence measurements in the equatorial Pacific (and more recently in the Atlantic and Indian basins) have expanded the deep cycle database dramatically beyond the sporadic shipboard observations that were available previously. Preliminary analysis of this data has “finally clinched the connection between mixing and climate, and holds the promise of improving global climate models” (Xie, S.P., Nature, Aug 2013). Observational work has been supplemented by linear stability analyses and large-eddy simulations which have greatly clarified the physics of deep cycle turbulence.
Most observations of the deep cycle have been made at 140W, in the equatorial Pacific cold tongue, and mostly in boreal fall. Does the deep cycle occur at other times of the year? In other locations around the equator? At different points in the ENSO cycle? How does the resulting diapycnal heat flux vary, and what large-scale phenomena control it? We are now at the stage of assessing the temporal and longitudinal extents of the deep cycle and the mechanisms that drive it. In this talk I will review the history of deep cycle research and will present the most recent results.
The Peruvian upwelling regime shows pronounced submesoscale variability including filaments and sharp density fronts. Submesoscale frontal processes can drive large vertical velocities and enhance vertical tracer fluxes in the upper ocean. The associated high temporal and spatial variability poses a large challenge to observational approaches targeting these processes.
In this study the role of submesoscale processes for both the ventilation of the near-coastal oxygen minimum zone off Peru and the physical-biogeochemical coupling at these scales is investigated. For our study we use satellite based sea surface temperature measurements and multiple high-resolution glider observations of temperature, salinity, oxygen and chlorophyll fluorescence carried out in January and February 2013 off Peru near 14°S during active upwelling. Additionally, high-resolution regional ocean circulation model outputs (ROMS) are analysed.
At the beginning of our observations a previously upwelled, productive and highly oxygenated water body is found in the mixed layer. Subsequently, a cold filament forms and the waters are moved offshore. After the decay of the filament and the relaxation of the upwelling front, the oxygen enriched surface water is found in the previously less oxygenated thermocline suggesting the occurrence of frontal subduction. A numerical model simulation is used to analyse the evolution of Lagrangian floats in several upwelling filaments, whose vertical structure and hydrographic properties agree well with the observations. The temporal evolution of the floats supports our interpretation that the subduction of previously upwelled water indeed occurs in filaments off Peru.
Filaments are common features in eastern boundary upwelling systems, which all encompass large oxygen minimum zones. However, most state of-the-art large and regional scale physical-biogeochemical ocean models do not resolve submesoscale filaments and the associated downward transport of oxygen and other solutes.
Diapyncal mixing processes in a deep ocean channel in the Lucky Strike region are investigated using microstructure data collected by an autonoumous underwater vehicle (AUV), moored current time series, lowered ADCP profiles and lowered CTD profiles. The distribution of the flow, the density, and the dissipation rate of turbulent kinetic energy in the deep ocean channel in the central valley of the Mid-Atlantic Ridge near 37°N are presented. Within the channel, mostly unidirectional, northward flow across a sill was present. The spatial distribution of the dissipation rate inside the channel was inferred using a horizontally profiling microstructure probe attached to an AUV. To the authors’ knowledge, this is the first successful realization of a horizontal, deep-ocean microstructure survey. The magnitude of the dissipation rate was distributed asymmetrically relative to the position of the sill. Elevated dissipation rates were present in a segment 1 to 4 km downstream of the sill with peak values of 1 ∙ 10−7 W/kg. Flow speeds with a maximum of 20~cm/s and elevated density finestructure were observed within the same segment.
Lowered and moored velocity observations showed large variability on semi-diurnal time-scales. The measurements indicated a hydraulic jump to be established at least temporarily downstream of the sill. Consistently, upward displacement of the isopycnals was observed in the area where the hydraulic jump is expected from the velocity distribution.
The spatial distributions of the flow, density and dissipation rate provide a consistent picture indicating deep ocean mixing to depend heavily on the local bottom topography and flow conditions. Furthermore, the results nicely illustrate that horizontally-profiling AUV-based observations may be an efficient tool to study deep-ocean turbulence over complex terrain.
We study the impact of vertical eddy heat and salinity fluxes using a suite of ocean-only simulations with the Max-Planck-Institute Ocean Model (MPIOM) at resolutions ranging from 1.5o to 0.1o, augmented by additional 1.5o-simulations in which the magnitude of eddy-induced tracer transports parameterized following Gent et al. (1995) is enhanced. Meso-scale eddies resolved by the 0.1o-MPIOM
produce upward vertical heat and salinity fluxes with considerable strength below a surface layer. The global integral of eddy heat flux exceeds 1 PW down to 2000 m. The global integral of eddy salinity flux, when converted into a fresh water flux, reaches about 0.2 Sv at 2000 m. The global-mean divergence of these eddy fluxes acts to cool and freshen water masses at intermediate depths, thereby reducing the major long-standing biases in the low-resolution MPIOM. This bias reduction cannot be fully achieved by increasing parameterized eddy-induced transports. Furthermore, the resolved eddies affect the global mean heat budget in a way fundamentally different from the parameterized eddies. The cooling and freshing induced by eddy forcing is balanced to a considerable degree by the diffusive processes in the 0.1o-simulation, but mostly by mean-flow advection in the low-resolution models. The unsatisfactory performance of the parameterized eddy-induced transports is partly related to the fact that these transports are derived from biased density fields and hence do not always occur where they should.
Currently, ocean processes smaller than the model resolution or faster than the model time step are parametrized using deterministic closure schemes. The schemes, which represent the effect of unresolved processes, are mostly based on empirical formulae and are traditionally down-gradient. However many of those ocean processes are turbulent and anti-diffusive, therefore potentially best represented by stochastic backscatter closure schemes, especially at eddy-permitting resolutions. A representation of the mesoscale eddy forcing as a function of the non-viscous straining and shearing of the resolved flow, wind forcing, stratification and model resolution is developed using probability distribution functions (PDFs) diagnosed from high-resolution quasigeostrophic eddy-resolving runs. The PDFs represent the eddy-eddy and eddy-mean flow interactions and associated Reynolds stresses necessary to mimic the effects of the eddies. The PDFs are used as a parametrization of sub-grid eddy effects, leading to improvements in the mean and variability of the coarse-resolution parametrized simulation over the unparametrized model. The effects of the parametrizations are shown to allow for upgradient fluxes and energy backscatter, therefore respecting the inverse energy cascade of quasi-geostrophic turbulence.
A new set of numerical simulations of turbulent plane Couette flow in a large box of dimension (20πh, 2h, 6πh) at Reynolds number (Reτ) = 125, 180, 250 and 550 is described and compared with simulations at lower Reynolds numbers, Poiseuille flows and experiments. The simulations present a logarithmic near-wall layer and are used to verify and revise previously known results. It is confirmed that the fluctuation intensities in the streamwise and spanwise directions do not scale well in wall units. The scaling failure occurs both near to and away from the wall. On the contrary, the wallnormal intensity scales in inner units in the near-wall region and in outer units in the core region. The spectral ridge found by Hoyas & Jiménez (Phys. Fluids, vol. 18, 2003, 011702) for the turbulent Poiseuille flow can also be seen in the present flow. Away from the wall, very large-scale motions are found spanning through all the length of the channel.
Dynamics of the upper layers in the central Baltic Sea were investigated, using towed instruments (T/S) and acoustic profiler along with a high-resolution numerical model, to better understand the physical conditions. Data were collected in July 2012 by cruising in large meandering patterns about once per day during 14 days. The surveys covered an area of 28 by 28 km in hydrography mode using Scanfish, towed and ship ADCP, and in fine-structure mode using towed CTD-chain, towed and ship ADCP. In addition, a freefalling microstructure profiler was used over an area of 15 by 15 km. All the CTD casts showed low-salinity intrusions around 20 m below the surface with varying thickness. Full spatio-temporal variability of water mass distribution was obtained with DIVAND (N-dimensional variation analysis). The observations were then compared to numerical model results. For that, the General Estuarine Transport Model (GETM) with a horizontal resolution of 600 m was used. In the vertical, 80 adaptive layers were employed. The numerical model helped us to explore the source and dynamics of the intrusions and there major driving mechanisms.
Ocean models have progressed substantially in the past couple of decades in terms of predictive capabilities and in terms of computational efficiency. However, there are still facets of these models that need refining. One of the most important processes which ocean models tend to underestimate is the level of mixing at the pycnocline. This is due to lack of depth resolution in regional models and deficiency in knowledge about the driving mechanism of turbulence at the pycnocline in terms of bulk parameters.
Pycnocline mixing, however, regulates fluxes of nutrients and other substances, between the surface mixed layer and deeper waters. This process therefore must be appropriately represented in order to obtain reliable predictions about, for instance, phytoplankton blooms or CO2 air-sea fluxes.
As a first step towards parameterising this process in regional models, a large-eddy simulation (LES) has been utilised. Large-eddy simulation is similar to an ocean model, but has far higher resolution and includes non-hydrostatic physics. The LES is configured to resolve the fine scale processes as well as incorporating tidal forcing (a process which has a larger wavelength than the domain size). I will present the relatively new technique of incorporating a tide into a large-eddy simulation and show differences in turbulent mixing between a tidally forced ocean model (namely NEMO) and the tidally forced LES.
Eastern boundary upwelling regimes are characterized by submesoscale frontal processes which are thought to be of major importance for air-sea gas exchange. The recent knowledge on submesoscale processes is mainly based on high-resolution numerical model studies. Direct observations of these highly variable processes remain a major challenge but are essential for model validation.
Here the role of time-variable submesoscale frontal processes for evolution of air-sea gas exchange is studied in the Peruvian upwelling regime near 14°S. The study is based on repeated transects of ship-based temperature, velocity, wind and CO2 underway measurements across an upwelling front in February 2013. Additionally, linear stability analysis is used to investigate the stability of the upwelling front.
During the three-day experiment, the temperature front collapses within 12 hours after a reduction of down-frontal wind-speed. As the front decays, surface temperature and velocity anomalies appear on wavelengths close to the mixed layer deformation radius, coinciding with Rossby numbers of O(1). Based on these results it is suggested that submesoscale dynamics such as mixed layer instabilities are involved in the frontal decay. This is found to be in agreement with the results from the linear stability analysis, which predicts a mixed layer mode with similar wavelengths.
During the phase when the front is active about two times higher surface pCO2 values are observed than after the frontal decay. The air-sea exchange of CO2 seems to be capped off after the frontal decay by the overlying layer of warm water. These results point to the importance of submesoscale processes on spatial scales of O(1-10 km) for the air-sea gas-exchange in upwelling regimes.
Recent results from a tracer experiment in the deepest layers of the central Baltic Sea suggest that boundary mixing plays an essential role for the net vertical transport of dissolved substances. Here, we discuss results from an extensive field campaign, aimed at clarifying the relevance of boundary mixing in the Bornholm Basin by direct observations of turbulence and mixing in the near-bottom region. The dataset includes high-resolution shear-microstructure and velocity data from seven transects across one of the lateral slopes of the Bornholm Basin (Southern Baltic Sea). Turbulence is found to be strongly enhanced in a near-bottom region varying between 0 and 7 m thickness. Most of the bottom boundary layer is stably stratified and characterized by small Ozmidov scales, suggesting that near-bottom turbulence is generally not controlled by law-of-the-wall dynamics. Dissipation rates inside the bottom boundary layer show strong cross-slope variability with the smallest values observed inside the stably stratified halocline region, and the largest above and below it. Turbulence in these regions is mainly driven by bottom friction, and, in contrast to turbulent boundary layers over a flat bottom, is likely to be efficient due to a permanent tendency for restratification. Our data suggest that most of the basin-wide energy dissipation and a substantial fraction of net diapycnal mixing occur inside the turbulent near-bottom region.
The presence of a seasonal thermocline likely plays a key role in restraining methane released from a seabed source in the deeper water column, thereby inhibiting exchange to the atmosphere. The bubble plume itself, however, generates an upward motion of fluid, e.g. upwelling and may thereby be partially responsible for an early breakdown of the seasonal thermocline. Measurements at site 22/4b, located at (57o55’N, 1o38’E) in the UK Central North Sea, 200 km east of the Scottish mainland, where gas is still being released since a blow out in 1990, have been used to identify the generation of the seasonal thermocline, and thus, the depth of the upper mixed layer and its breakdown in autumn. Data derived from two landers, containing an Acoustic Doppler Current Profiler and a Conductivity Temperature Depth recorder, were used to determine the mixed layer depth and the breakdown of the thermocline. Mixing of upper layer fluid into the lower layer has been inferred from large amplitude variations in the near-bottom temperature.
The ADCPs estimate velocity profiles in four beam directions using Doppler shifted frequency from acoustic pings sent out and received by four different transducers in a specific configuration. Besides that, the intensity of the backscattered sound per transducer is also recorded. Bubbles from the nearby plume contaminate the signal during part of the tidal cycle, but in bubble free periods, the mixed layer depth can be estimated using the acoustic backscatter signal as local maxima. Results show that the thermocline broke down between mid-October and early November, several weeks earlier than the breakdown of the thermocline in nearby/comparable areas, likely caused by bubble-induced downwelling at the site. The early breakdown of the thermocline was accompanied by multiple occurrence of a strong jet-like structure, associated with the seasonal tidal mixing front.
The influence of turbulence on the distribution of phytoplankton is a classical topic in biological oceanography. Recent studies have shown how it is necessary to include accurately the physical structure of the water column to assess properly this influence. In this work we use a simple 1D model of phytoplankton and a set of physical framework to show how introducing a more realistic turbulent layer could alter phytoplankton vertical distribution. Sinking rate and the effect of a vertical variable sinking rate could be particularly relevant to determine vertical phytoplankton distribution. In this case, vertical variation of sinking rate is driven by turbulence, based on relationships from empirical data. Furthermore, the model is used to assess the likely consequences of increasing stratification and wind-induced surface turbulence associated with global warming. Our results show that, even under a scenario of reduced biological biomass, the redistribution of pelagic material due to increasing turbulence levels could effectively increase the vertical flux of organic matter into the deep ocean.
A pilot field experiment was conducted to assess the importance of zooplankton generated mixing in stratified lakes. The objective of this work is to improve understanding of thermocline mixing in the lake interior, which is crucially important for better predictions of transport of dissolved substances, and consequent impacts upon lake ecosystem functioning. In this experiments, turbulence was directly measured in the thermocline of a lake during a vertical migration of crustacean zooplankton (Daphnia). Profiles of turbulence were measured with a temperature microstructure instrument in Monkswood Reservoir (UK), a small manmade lake with small wind fetch, steep sides, and a flat bottom with a depth of approximately 11 m. Twenty-seven profiles were measured over the course of three hours during sunset on 1 September 2014, during which there was no wind. Zooplankton tows were conducted before and after the turbulence profiling to verify the distribution of Daphnia before and after sunset. The zooplankton tows showed an increase in daphnia above 6 m depth and a decrease below 6 m after sunset. There was also an increase in the dissipation rate of turbulent kinetic energy to 10^-7 W/kg (two orders of magnitude above the background levels) at 6 m depth over the course of the time series. Given the uncertainty in measuring the length scales of turbulence associated with small zooplankton, it is not certain if the observed turbulence during the time of migration was due the migration or due to other causes, such as the onset of penetrative convection associated with nighttime cooling. In Summer 2015, these measurements will be repeated in Vobster Quay (UK), a gravel pit lake with a maximum depth of 40 m that has a significantly higher Daphnia population and stronger stratification than Monkswood Reservoir.
Remote sensing techniques provide unique opportunities to observe oceanic turbulence. In case of presence of traces on the water surface or in the surface water layer satellite images represent the snapshots of the surface currents. Spatial scale of the features being captured during such surveys ranges from hundreds meters (high resolution optical and radar imagery) to hundreds or even thousands kilometers (passive microwave, optical, and infrared imagery). The present study is focused on the coherent vortical structures being observed in satellite imagery of the inner (mainly European) seas. Such vortical structures, or eddies, play essential role in marine ecosystems. They widely participate in horizontal and vertical water transport and thus affect the biogeochemical processes and in particular bioproductivity of the marine ecosystems.
The study is based on analysis of a wide range of radar (Envisat Advanced Synthetic Aperture Radar (ASAR)), infrared (Advanced Very High Resolution Radiometer (AVHRR), Moderate Resolution Imaging Spectroradiometer (MODIS)), and visible-range (Medium Resolution Imaging Spectrometer (MERIS), MODIS) satellite imagery. Among the regions of interest there were the North, Baltic, Black, Caspian, Mediterranean, and Red Seas. The images were acquired in the period between 2006 and 2013. Total amount of images exceeds 13.000.
On analysis of the images mentioned above, location and length scale of almost 33.000 submeso- and mesoscale eddies were defined. The collected dataset of eddy manifestations was further processed providing a comprehensive analysis of spatial and temporal distribution of the eddies. For gaining the details of the factors governing the generation of the eddies, observed distribution of eddies was compared with distribution of other hydrophysical and meteorological parameters retrieved from satellite observations and numerical modelling. As a result, it was noticed that strong mesoscale activity is being observed in the areas with strong jet currents, while submesoscale eddies tend to be located (i) in the vicinity of the sharp thermal fronts and/or (ii) with shallow thermocline depths.
Transport by surface currents in the Gulf of Finland has been studied by use of surface drifters deployed within the period 2010-2014. Drifters were usually deployed in pairs or triplets in order to study the relative separation rate. The correlation between drifter motions within a cluster was reduced as the separation distance increased, but some coordination of motion was often observed even for separation distances of several kilometers. Analysis of relative dispersion revealed that the drifter separation followed a Richardson's law dispersion, characterized by a time dependent relative distance increase of x(t) ~ t3/2. This behavior was prominent for separation distances ranging from ~50 m up to ~5 km.
The Gulf of Finland is characterized as an elongated estuarine sea basin, and it is generally accepted that the sea surface current field is dominated by the wind induced forcing, either directly through shear stress acting on the sea surface or indirectly via surface waves (Stokes drift). We examined the wind effect on surface drifter motion by using data from the Kalbadagrund weather station. The aim of the study was to investigate the correlation between drifter motion and wind direction and speed, and to examine if the wind forcing could explain the transition from highly correlated to uncorrelated drifter motion within a cluster.