Total kinetic energy in four global eddying ocean circulation models and over 5000 current meter records

We compare the total kinetic energy (TKE) in four global eddying ocean circulation simulations with a global dataset of over 5000, quality controlled, moored current meter records. At individual mooring sites, there was considerable scatter between models and observations that was greater than estimated statistical uncertainty. Averaging over all current meter records in various depth ranges, all four models had mean TKE within a factor of two of observations above 3500 m, and within a factor of three below 3500 m. With the exception of observations between 20 and 100 m, the models tended to straddle the observations. However, individual models had clear biases. The free running (no data assimilation) model biases were largest below 2000 m. Idealized simulations revealed that the parameterized bottom boundary layer tidal currents were not likely the source of the problem, but that reducing quadratic bottom drag coefficient may improve the fit with deep observations. Data assimilation clearly improved the model-observation comparison, especially below 2000 m, despite assimilated data existing mostly above this depth and only south of 47 °N. Different diagnostics revealed different aspects of the comparison, though in general the models appeared to be in an eddying-regime with TKE that compared reasonably well with observations.     

Evaluations of threshold and curvature mixed layer depths by various mixing schemes in the Mediterranean Sea

Predictive ability of five different embedded turbulent mixing models that range from second-order turbulent closure to bulk mixing parameterization is examined in the Mediterranean Sea. Each is embedded in the HYbrid Coordinate Ocean Model (HYCOM). Mixed layer depth (MLD), which is one of the most important upper ocean variables, is used to evaluate the treatment of turbulent processes in each model run. In addition to overall spatial and temporal variability, analyses of MLD are presented using an extensive set (3976) of temperature and salinity profiles from various data sources during 2003–2006. Results obtained from simulations (with no data assimilation and relaxation only to salinity) for the five mixing models are compared with observed MLDs obtained from in situ temperature and salinity profile observations. To ensure the robustness of the validation statistics MLD is computed using both curvature and threshold based methodologies. Results indicate that while all mixing schemes represent the MLD well, the bulk mixing models have substantial accuracy deficiencies relative to the higher order mixing models. The modeled MLDs are slightly deeper than observed MLDs with the mean bias error ∼10 m for the higher order mixing models while the bulk mixing model bias error is 15 m or more. The RMS error for the higher order mixing models is ∼40 m while it is ∼50 m for the bulk mixing models. The bulk mixing models had substantially larger errors particularly for the curvature MLD definition.     

Sea level anomalies in straits of Malacca and Singapore

This paper studies sea level anomaly (SLA) behaviour in Malacca and Singapore straits which serve part of a major maritime trade route between Indian and Pacific Ocean using both observed data and numerical model. Spatio-temporal behaviour of SLA in the region is analyzed based on 15 years of in-situ and remote sensing data. Results show that SLA signatures can be distinctly different in the two straits, with vastly opposite behaviours during certain months. By further analyzing spatial dependency of observed SLA in the region, SLA in Malacca and Singapore straits are found to be under the influence of Indian Ocean and South China Sea, respectively. Based on this insight, a numerical model is built with the appropriate non-tidal forcing derived from meteorological model and satellite dataset to properly represent SLA in Malacca and Singapore straits with Root Mean Square Error of less than 10 cm. With this well calibrated model, the effect of different types of forcing on volume flux through the straits is investigated. Combined tidal and non-tidal forcing in the model gives 4 to 7 × 10¹¹ m³ of annual net westward volume flux through the straits which is four to seven times higher than that of tidal forcing alone. Furthermore with this combined forcing, a distinct seasonal trend with westward net flow during northeast monsoon (November to March) and eastward net flow during southwest monsoon (May to September) can be observed through the straits in the model which is not observed with tidal forcing. The findings of this paper highlight the importance of these non-tidal forcing in the model to obtain accurate SLA and flow representation in the straits that is vital to environmental fate and transport modelling during operational forecast.     

Offshore wind farm impacts on surface waves and circulation in Eastern Lake Ontario

A coupled wave and hydrodynamic model was applied to the Kingston Basin of eastern Lake Ontario, a region with bathymetric variability due to channels and shoals, to assess the potential impacts on surface waves and wind-driven circulation of an offshore wind farm. The model was used to simulate a series of storm events with time-varying wind forcing and validated against wave, current and water level observations. The wind farm was simulated by adding semi-permeable structures in the surface wave model to represent the turbine monopiles, and by adding an energy loss term to the fluid momentum equations in the hydrodynamic model to represent the added drag of the monopiles on the flow. The results suggest that the wind farm would have a small influence on waves and circulation throughout the wind farm area, with spatial variability due to focussing of wave energy and re-direction of the flow. Overall, the results indicate that the wave height in coastal areas will be minimally affected with changes in significant wave height predicted to be < 3%. Larger changes to the strength of circulation occur inside the wind farm region with localized changes in current magnitude of up to 8 cm s^− 1. The results of this study may help to understand the impacts of future offshore wind farms and other offshore structures in the Great Lakes.     

Eddies and a mesoscale deflection of the slope current in the Faroe–Shetland Channel

The mesoscale dynamics of the Scottish side of the Faroe–Shetland Channel have been investigated using synoptic in situ and remote sensing observations. A cold core cyclonic eddy, identified from an AVHRR image, had a diameter of about 50 km and surface current speeds of up to 50 cm s^-1; it appeared to be attached to the 800 m isobath as it moved north-eastward along the edge of the channel at about 8 cm s^-1. Speeds in the slope current were about 50 cm s^-1 but increased to 70 cm s^-1 where the current was compressed by the eddy. Offshore, over the 1000 m isobath in the cooler water, speeds in the current were slower (ca. 20 cm s^-1). North-west of the Shetlands the offshore edge of the slope current was deflected across the channel for a distance of about 70 km from the shelf edge. The speed of drifters in the slope current increased to over 60 cm s^-1 as they moved anti-cyclonically around this deflection. CTD profiles suggest that the movement of the surface waters was mirrored in the deep water of the channel. The deflection carried a very large quantity of North Atlantic Water into the central part of the channel; its cause and ultimate fate are not known, although it is likely to have had a significant impact on the dynamics of the channel.     

Atlantic inflow to the Nordic Seas: current structure and volume fluxes from moored current meters,VM-ADCP and SeaSoar-CTD observations, 1995–1999

This study deals with the inflow of warm and saline Atlantic water to the Nordic Seas, an important factor for climate, ecology and biological production in Northern Europe. The investigations are carried out along the Svinøy standard hydrographic section, which cuts through the Atlantic inflow to the Norwegian Sea just to the north of the Faroe–Shetland Channel. In the Svinøy section, we consider the Atlantic inflow as water with salinity above 35.0, corresponding to temperatures above 5°C. Current measurements for the period April 1995 to February 1999, positioned on the continental slope in water depths between 490 and 990 m, are combined with VM-ADCP, SeaSoar-CTD and CTD transects to estimate long-term transports and spatial features of the Atlantic inflow. A well-defined two-branched Norwegian Atlantic Current was revealed with an eastern and a western branch. The eastern branch appears as a narrow, topographically trapped, near barotropic, 30–50 km wide current, with a maximum speed of 117 cm/s. The western branch is also about 30–50 km wide, and appears as an unstable frontal jet about 400 m deep with a maximum speed of 87 cm/s. Between these two prominent branches, the observations show an average eddy field with a recirculation to the southwest. Transport estimates from the current records in the eastern branch show an annual mean inflow of 4.2 Sv (1 Sv=10⁶ m³/s) with variation on a 25 h time scale ranging from −2.2 to 11.8 Sv, and between 2.0 and 8.0 Sv on a monthly time scale. The current record in the core of the eastern branch mirrors the estimated transport on a monthly time scale with a correlation coefficient of 0.86. Except for the year 1995–1996, this nearly four-year current record shows evidence of a systematic annual cycle with summer to winter variations in the proportion of 1 to 2. Comparison between the North Atlantic Oscillation (NAO) index and the current record on a three-month time scale shows a strong connection for most of the period. This reflects the strong coupling between the westerly winds and the inflow. The baroclinic transport west of the eastern branch, including the frontal jet, is inferred from hydrography in combination with VM-ADCP transects, and has a total mean of 3.4 Sv. Thus, investigations to date indicate a yearly mean Atlantic inflow of 7.6 Sv in the Svinøy section.     

Corrections to ocean surface forcing in the California Current System using 4D variational data assimilation

The option for surface forcing correction, recently developed in the 4D-variational (4DVAR) data assimilation systems of the Regional Ocean Model System (ROMS), is presented. Assimilation of remotely-sensed (satellite sea surface height anomaly and sea surface temperature) and in situ (from mechanical and expendable bathythermographs, Argo floats and CTD profiles) oceanic observations has been applied in a realistic, high resolution configuration of the California Current System (CCS) to sequentially correct model initial conditions and surface forcing, using the Incremental Strong constraint version of ROMS-4DVAR (ROMS-IS4DVAR). Results from both twin and real data experiments are presented where it is demonstrated that ROMS-IS4DVAR always reduces the difference between the model and the observations that are assimilated. However, without corrections to the surface forcing, the assimilation of surface data can degrade the temperature structure at depth. When using surface forcing adjustment in ROMS-IS4DVAR the system does not degrade the temperature structure at depth, because differences between the model and surface observations can be reduced through corrections to surface forcing rather than to temperature at depth. However, corrections to surface forcing can generate abnormal spatial and temporal variability in the structure of the wind stress or surface heat flux fields if not properly constrained. This behavior can be partially controlled via the choice of decorrelation length scales that are assumed for the forcing errors. Abnormal forcing corrections may also arise due to the effects of model error which are not accounted for in IS4DVAR. In particular, data assimilation tends to weaken the alongshore wind stress in an attempt to reduce the rate of coastal upwelling, which seems to be too strong due to other sources of error. However, corrections to wind stress and surface heat flux improve systematically the ocean state analyses. Trends in the correction of surface heat fluxes indicate that, given the ocean model used and its potential limitations, the heat flux data from the Coupled Ocean–Atmosphere Mesoscale Prediction System (COAMPS) used to impose surface conditions in the model are generally too low except in spring-summer, in the upwelling region, where they are too high. Comparisons with independent data provide confidence in the resulting forecast ocean circulation on timescales ～14 days, with less than 1.5 °C, 0.3 psu, and 9 cm RMS error in temperature, salinity and sea surface height anomaly, respectively, compared to observations.     

Absolutely referenced geostrophic velocity and transport on a section across the North Atlantic Current

A transect of CTD profiles crossing the North Atlantic Current (NAC) along WOCE line ACM6 near 42.5°N during August 1–7, 1993, provides geostrophic shear velocity profiles, which were absolutely referenced using simultaneous POGO transport float measurements and velocity measurements from a ship-mounted acoustic doppler current profiler (ADCP). The NAC absolute transport was 112±23×10⁶ m³ s⁻¹, which includes a portion of the transport of the Mann Eddy, a large permanent anticyclonic eddy commonly adjacent to the NAC. The NAC transport estimated relative to a level of no motion at the bottom would have underestimated the true total absolute transport by 20%. A surprisingly large 58×10⁶ m³ s⁻¹ flowed southward just inshore of the NAC. This flow, centered near 1500 dbars about 200 km offshore of the shelf-break, was fairly barotropic with a peak velocity of greater than 20 cm s⁻¹, and the water mass characteristics were of Labrador Sea Water. These absolute transport observations suggest southward recirculation inshore of the NAC at 42.5°N and a stronger NAC than has previously been observed.     

Using stochastic population process models to predict the impact of climate change

More than ten years ago a paper was published in which stochastic population process models were fitted to time series of two marine polychaete species in the western Wadden Sea, The Netherlands (Van der Meer et al., 2000). For the predator species, model fits pointed to a strong effect of average sea surface winter temperature on the population dynamics, and one-year ahead model forecasts correlated well with true observations (r = 0.90). During the last decade a pronounced warming of the area occurred. Average winter temperature increased with 0.9 °C. Here we show that despite the high goodness-of-fit whilst using the original dataset, predictive capability of the models for the recent warm period was poor.     

Transformations of biogenic particulates from the pelagic to the deep ocean realm

This overview compares and contrasts trends in the magnitude of the downward Particulate Organic Carbon (POC) flux with observations on the vertical profiles of biogeochemical parameters in the NE subarctic Pacific. Samples were collected at Ocean Station Papa (OSP, 50°N, 145°W), between 18–22 May 1996, on pelagic stocks/rate processes, biogenic particle fluxes (drifting sediment traps, 100–1000 m), and vertical profiles of biogeochemical parameters from MULVFS (Multiple Unit Large Volume Filtration System) pumps (0–1000 m). Evidence from thorium disequilibria, along with observations on the relative partitioning of particles between the 1–53 μm and >53 μm classes in the 50 m mixed layer, indicate that there was little particle aggregation within the mixed layer, in contrast to the 50–100 m depth stratum where particle aggregation predominated. Vertical profiles of thorium/uranium also provided evidence of particle decomposition occuring at depths ca. 150 m; heterotrophic bacteria and mesozooplankton were likely responsible for most of this POC utilisation. A water column carbon balance indicated that the POC lost from sinking particles was the predominant source of carbon for bacteria, but was insufficient to meet their demands over the upper 1000 m. While, the vertical gradients of most parameters were greatest just below the mixed layer, there was evidence of sub-surface increases in microbial viability/growth rates at depths of 200–600 m. The C:N ratios of particles intercepted by free-drifting and deep-moored traps increased only slightly with depth, suggesting rapid sedimentation even though this region is dominated by small cells/grazers, and the upper water column is characterised by long particle residence times (>15 d), a fast turnover of POC (2 d) and a low but constant downward POC flux.     

The western boundary current in the Bransfield Strait,Antarctica

The Bransfield Strait west of the Antarctic Peninsula has been considered as a highly productive region for all trophic levels from primary production, to zooplankton aggregations, especially krill, to birds and mammals. The western boundary current, referred to as the Bransfield Current, plays an important role in determining the transport and retention of biota in the Bransfield Strait. Following the study of surface current characteristics in the strait using 39 tracks of mixed-layer drifters deployed between 1988 and 1990, a high-resolution transect of temperature, salinity and current measurements crossing the Bransfield Current was conducted between 13 and 14 March 2004, for understanding its horizontal and vertical structure and dynamics. The results from current, temperature and salinity measurements using a vessel mounted narrow band acoustic doppler current profiler and conductivity–temperature–depth (CTD) sensors revealed the magnitude of this current of approximately 50 cm/s within a horizontal distance of 15 km associated with a narrow and deep density front 4–6 km wide and 500 m deep. The comparison between the direct current measurements and the geostrophic current estimates from the density field implies that the Bransfield Current is geostrophically balanced. The mechanism forming this current is explored with Sverdrup dynamics. Results indicate that the negative wind stress curl and β-effect lead to a southwestward transport in the Bransfield Strait. When this transport is restricted by land and shelves, a narrow western boundary current is formed.     

Diagnosis of vertical velocities with the QG omega equation: an examination of the errors due to sampling strategy

Vertical motion at the mesoscale plays a key role in ocean circulation, ocean-atmosphere interaction, and hence climate. It is not yet possible to make direct Eulerian measurements of vertical velocities less than 1000 m day⁻¹. However, by assuming quasi-geostrophic (QG) balance, vertical velocities O (10 m day⁻¹) can be diagnosed from the geostrophic velocity field and suitable boundary conditions. Significant errors in the accuracy of this diagnosis arise from the necessary compromise between spatial resolution and synopticity of a hydrographic survey. This problem has been addressed by sampling the output of a numerical ocean model to simulate typical oceanographic surveys of mesoscale fronts. The balance between the number of observations and the synopticity of observations affects the apparent flow and in particular the diagnosed vertical motion. A combination of effects can typically lead to errors of 85% in the estimation of net vertical heat flux. An analytical two-layer model is used to understand components of this error and indicate the key parameters for the design of mesoscale sampling.     

Investigation of a longfin inshore squid's swimming characteristics and an underwater locomotion during acceleration

In the present study, locomotion of a real longfin inshore squid (Doryteuthis pealeii) was numerically investigated. Geometry of a real squid was obtained using computed tomography (CT) images. In addition to a two-dimensional axisymmetric squid model, a modified squid model with no cavities around her head and an ellipse shaped model were generated with a fineness ratio (the ratio of body length to maximum body diameter) of 7.56. These numerical models were exposed to an acceleration with two different velocity programs. Added mass and basset forces on bodies were calculated during acceleration of the squid models starting from rest. Pressure and viscous drag forces were also calculated due to pressure variation along the squid models and friction on the surface of the models. The effect of a nozzle diameter on jet velocities and propulsive efficiency at all bodies were evaluated when time dependent velocity profiles (from 0 to 10 m/s in 0.5 and 1 s time durations) were set for the inlet of computational domain. The modified squid model required least thrust force during acceleration phase of time dependent velocity profile compared to the other models while the 0.02 m nozzle diameter provided largest propulsive efficiency for all models.     

Solitary wave generation dynamics at Luzon Strait

A high resolution modeling study is undertaken, with a 2.5-dimensional nonhydrostatic model, of the generation of internal waves induced by tidal motion over the ridges in Luzon Strait. The model is forced by the barotropic tidal components K1, M2, and O1. These tidal components, along with the initial density field, were extracted from data and models. As the barotropic tide moves over the Luzon Strait sills, there is a conversion of barotropic tidal energy into baroclinic tidal energy. Depressions are generated that propagate towards the Asian Seas International Acoustics Experiment (ASIAEX) test site on the Chinese continental shelf. Nonlinear effects steepen the depressions, frequency and amplitude dispersion set in, and disintegration into large amplitude solitary waves occurs. The effects of varying the initial density field, tidal component magnitudes, as well as adding a steady background current to represent the occasional excursions of the Kuroshio Current into the strait, are considered.Depressions are generated at each of the two sills in Luzon Strait which radiate away, steepening and evolving into internal solitary wave trains. Baroclinic fluxes of available potential energy, kinetic energy and linear are calculated for various parameter combinations. The solitary wave trains produced in the simulations generally consist of large amplitude wave trains alternating with small amplitude wave trains. During strong tidal flow, Kelvin–Helmholtz type instabilities can develop over the taller double-humped sill. The solitary waves propagating towards the ASIAEX test site have been observed to reach amplitudes of 120–250 m, depending on the tidal strength. ASIAEX observations indicate amplitudes up to 150 m and the Windy Island Experiment (WISE) measurements contain magnitudes over 200 m. The model results yield solitary wave amplitudes of 70–300 m and half widths of 0.60–3.25 km, depending on parameter values. These are in the range of observations. Measurements by Klymak et al. (2006), in the South China Sea, exhibit amplitudes of 170 m, half widths of 3 km and phase speeds of 2.9 m s^?1. Model predictions indicate that the solitary waves making up the wave packet each experience different background currents with strong near surface shear.The energy in the leading soliton of the large amplitude wave trains ranges between 1.8 and 9.0 GJ m^?1. The smaller value, produced using barotropic tidal currents based on the Oregon State University data base, is the same as the energy estimated to be in a solitary wave observed by Klymak et al. (2006). Estimates of the conversion of barotropic tidal energy into radiating internal wave energy yield conversion rates ranging between 3.6% and 8.3%.     

How does multiscale modelling and inclusion of realistic palaeobathymetry affect numerical simulation of the Storegga Slide tsunami?

The ∼8.15 ka Storegga submarine slide was a large (∼3000 km³), tsunamigenic slide off the coast of Norway. The resulting tsunami had run-up heights of around 10–20 m on the Norwegian coast, over 12 m in Shetland, 3–6 m on the Scottish mainland coast and reached as far as Greenland. Accurate numerical simulations of Storegga require high spatial resolution near the coasts, particularly near tsunami run-up observations, and also in the slide region. However, as the computational domain must span the whole of the Norwegian-Greenland sea, employing uniformly high spatial resolution is computationally prohibitive. To overcome this problem, we present a multiscale numerical model of the Storegga slide-generated tsunami where spatial resolution varies from 500 m to 50 km across the entire Norwegian-Greenland sea domain to optimally resolve the slide region, important coastlines and bathymetric changes. We compare results from our multiscale model to previous results using constant-resolution models and show that accounting for changes in bathymetry since 8.15 ka, neglected in previous numerical studies of the Storegga slide-tsunami, improves the agreement between the model and inferred run-up heights in specific locations, especially in the Shetlands, where maximum run-up height increased from 8 m (modern bathymetry) to 13 m (palaeobathymetry). By tracking the Storegga tsunami as far south as the southern North sea, we also found that wave heights were high enough to inundate Doggerland, an island in the southern North Sea prior to sea level rise over the last 8 ka.     

The supply of nutrients due to vertical turbulent mixing: A study at the Porcupine Abyssal Plain study site in the northeast Atlantic

As part of a multidisciplinary cruise to the Porcupine Abyssal Plain (PAP) study site (49°00′N 16°30′W), in June and July 2006, observations were made of the vertical nitrate flux due to turbulent mixing. Daily profiles of nitrate and turbulent mixing, at the central PAP site, give a mean nitrate flux into the euphotic zone of 0.09 (95% confidence intervals: 0.05–0.16) mmol N m⁻² d⁻¹. This is a factor of 50 lower than the mean observed rate of nitrate uptake within the euphotic zone (5.1±1.3 mmol N m⁻² d⁻¹). By using our direct observations to ‘validate’ a previously published parameterisation for turbulent mixing, we further quantify the variability in the vertical turbulent flux across a roughly 100×100 km region centred on the PAP site, using hydrographic data. The flux is uniformly low (0.08±0.26 mmol N m⁻² d⁻¹, the large standard deviation being due to a strongly non-Gaussian distribution) and is consistent with direct measurements at the central site. It is demonstrated that on an annual basis convective mixing supplies at least 40-fold more nitrate to the euphotic zone than turbulent mixing at this location. Other processes, such as those related with mesoscale phenomena, may also contribute significantly.     

Dynamics at an elongated,intermediate depth seamount in the North Atlantic (Sedlo Seamount, 40°20′N, 26°40′W)

Observations from a five-mooring array deployed in the vicinity of Sedlo Seamount over a 4-month period, together with supporting hydrographic and underway ADCP measurements, are described. Sedlo Seamount is an elongated, intermediate depth seamount with three separate peaks, rising from 2200 m water depth to summit peaks between 950 and 780 m depth, located at 40°20′N, 26°40W. Currents measured in depth range 750 and 820 m – the layer close to the summit depth of the shallowest southeast peak – showed a mean anti-cyclonic flow around the seamount, with residual current velocities of 2–5 cm s⁻¹. Significant mesoscale variability was present at this level, and this is attributed to the weak and variable background impinging flow. Stronger, more persistent currents were found at the summit mooring as a result of tidal rectification and some weak amplification. Below 1300 m, currents were extremely weak, even close to the seabed. Time series of relative vorticity for the depth layer 750–820 m showed persistent anti-cyclonic vorticity except for two periods of cyclonic vorticity. A mean relative vorticity of −0.06f (f=the local Coriolis frequency) was calculated from a triangle of current meters located at the flanks of the seamount. Modelling results confirmed that anti-cyclonic flow above the seamount was likely due to Taylor Cone generation driven by a combination of steady impinging and tidally rectified flow. The closed circulation pattern over the seamount was found to extend to ∼150 m above the summit level, consistent with simple idealised theory and the supporting hydrographic observations. At shallower depths (<500 m) model simulations predicted a predominantly cyclonic recirculation most likely controlled by topographic steering along the zonal axis of the seamount. There was some indication of flow reversal at these depths from Acoustic Doppler Current Profiler (ADCP) measurements carried out at one hydrographic survey. The model results were in good agreement with observations at the seamount summit, but were unable to reproduce the mesoscale variability patterns recorded in shallower layers. Kinetic energy patterns derived from the model revealed high variability in the oceanic far field downstream of the seamount summit probably as a result of complex flow interaction along the chain of seamount peaks. Possible impacts of the flow dynamics on the biological functioning at Sedlo Seamount and its surroundings are discussed.     

Comparison between M5′ model tree and neural networks for prediction of significant wave height in Lake Superior

Prediction of wave height is of great importance in marine and coastal engineering. Soft computing tools such as artificial neural networks (ANNs) are recently used for prediction of significant wave height. However, ANNs are not as transparent as semi-empirical regression-based models. In addition, neural networks approach needs to find network parameters such as number of hidden layers and neurons by trial and error, which is time consuming. Therefore, in this work, model trees as a new soft computing method was invoked for prediction of significant wave height. The main advantage of model trees is that, compared to neural networks, they represent understandable rules. These rules can be readily expressed so that humans can understand them. The data set used for developing model trees comprises of wind and wave data gathered in Lake Superior from 6 April to 10 November 2000 and 19 April to 6 November 2001. M5′ algorithm was employed for building and evaluating model trees. Training and testing data include wind speed (U₁₀) as the input variable and the significant wave height (H_s) as the output variable. Results indicate that error statistics of model trees and feed-forward back propagation (FFBP) ANNs were similar, while model trees was marginally more accurate. In addition, model tree shows that for wind speed above 4.7 m/s, the wave height increases nonlinearly by the wind speed.     

Observations of the Storfjorden overflow

The mixing and spreading of the Storfjorden overflow were investigated with density and horizontal velocity profiles collected at closely spaced stations. The dense bottom water generated by strong winter cooling, enhanced ice formation and the consequent brine rejection drains into and fills the depression of the fjord and upon reaching a 120-m deep sill, descends like a gravity current following the bathymetry towards the shelf edge. The observations covered an approximate 37-km path of the plume starting from about 68 km downstream of the sill. The plume is identified as two layers: a dense layer 1 with relatively uniform vertical structure underlying a thicker layer 2 with larger vertical density gradients. Layer 1, probably remnants from earlier overflows, almost maintains its temperature–salinity characteristics and spreads to a width of about 6 km over its path, comparable to spread resulting from Ekman veering. Layer 2, on the other hand, is a mixing layer and widens to about 16 km. The overflow, in its core, is observed to have salinities greater than 34.9, temperatures close to the freezing point, and light transmissivity typically 5% less than that of the ambient waters. The overall properties of the observed part of the plume suggest dynamical stability with weak entrainment. However local mixing is observed through profiles of the gradient Richardson number, the non-dimensional ratio of density gradient over velocity gradient, which show portions with supercritical values in the vicinity of the plume–ambient water interface. The net volume transport associated with the overflow is estimated to be 0.06 Sv (Sv≡10⁶ m³ s⁻¹) out of a section closest to the sill and almost double that as it leaves the section furthest downstream. The weak entrainment is estimated to account for the doubling of the volume transport between the two sections. A simple model proposed by Killworth (J. Geophys. Res. 106 (2001) 22267), giving the path of the overflow from a constant rate of vertical descent along the slope, compares well with our observations.     

Observations on the Rim Current structure,CIW formation and transport in the western Black Sea

CTD and ADCP measurements together with a sequence of satellite images indicate pronounced current meandering and eddy activity in the western Black Sea during April 1993. The Rim Current is identified as a well-defined meandering jet stream confined over the steepest topographic slope and associated cyclonic–anticyclonic eddy pairs located on both its sides. It has a form of highly energetic and unstable flow system, which, as it propagates cyclonically along the periphery of the basin, is modified in character. It possesses a two-layer vertical structure with uniform upper layer speed in excess of 50 cm/s (maximum value ∼100 cm/s), followed by a relatively sharp change across the pycnocline (between 100 and 200 m) and the uniform sub-pycnocline currents of 20 cm/s (maximum value ∼40 cm/s) observed up to the depth of ∼350 dbar, being the approximate limit of ADCP measurements. The cross-stream velocity structure exhibits a narrow core region (∼30 km), flanked by a narrow zone of anticyclonic shear on its coastal side and a broader region of cyclonic shear on its offshore side. The northwestern shelf circulation is generally decoupled from the influence of the basinwide circulation and is characterized by much weaker currents, less than 10 cm/s. The southward coastal flow associated with the Danube and Dinepr Rivers is weak during the measurement period and is restricted to a very narrow coastal zone.The data suggest the presence of temperature-induced overturning prior to the measurements, and subsequent formation of the Cold Intermediate Water mass (CIW) within the Northwestern Shelf (NWS) and interior of the western basin. The newly formed shelf CIW is transported in part along the shelf by the coastal current system, and in part it flows downslope across the shelf and intrudes into the Rim Current convergence zone. A major part of the cold water mass, however, seems to be trapped within the northwestern shelf. The CIW mass, injected into the Rim Current zone from the shelf and the interior region, is then circulated around the basin.