Search

We investigated the effect of the assimilation of surface velocities from two HF-Radars (HFRs) and satellite sea surface temperature (SST) into an ocean circulation model, implemented on the North-Western Mediterranean Sea, using the ROMS 4D-Var Data Assimilation (DA) System. A three months period from September to November 2020 is analysed. A model free run is used as reference simulation to quantify the impact of the data assimilation procedure on the surface circulation. The DA framework consists of a sequence of 3-day windows during which the observations are assimilated into the model, followed by forecast simulations, starting from the last time step of the analysis run. The comparison of simulated velocities and SST against the assimilated observations shows that both forecast and analysis runs improve the solution as compared to the freerun, with error reduction up to the 47% and doubling of correlation values. If SST is not assimilated, the simulated sea surface temperature slightly degrades as a result of the DA procedure. Furthermore, we used surface velocities from Lagrangian drifters to test the model against independent observations. The results show that both the forecast and the analysis simulate surface circulation better than the free run within the areas covered by HFR observations, with error reduction ranging from 10% to 50%, and increase in correlation from 5% to more than doubled. Nevertheless, farther afield, outside the area covered by HFRs, improvements and degradations of the solution due to the DA procedure balance each other. The impact of assimilated observation on the alongshore transport of the upper 50 m is analysed along three transects adjacent to the two HFR areas and along a section in between. The transport increment is significant where the surface velocities are assimilated whereas it is negligible elsewhere. The effect of SST is to modify the velocity distribution along the transects, while keeping the transport increment unchanged. Modifications to the initial conditions have the most significant effect on the alongshore transport, whereas the corrections to boundary conditions and atmospheric forcing have different importance according to the analysed transects. The assimilation of surface velocities and SST affects more the ageostrophic component of the surface velocity field, rather than the geostrophic one. • 4D-Var DA of HF-Radar surface velocities and SST improve analysis and forecast. • Validation with independent data shows that benefits are limited to the HF-Radar area. • HF-Radar data change transport magnitude and SST modulates the velocity distribution. [ABSTRACT FROM AUTHOR]

Despite the increasing volume of oceanographic data, the available ocean salinity data remains patently insufficient which limits studying ocean dynamics, climate change and the calculation of salinity-related ocean elements. Considering that traditional salinity reconstruction methods often suffer from various factors such as additional constraints, priori physical assumptions and large of specific regression coefficients, a generative adversarial networks (GANs) based deep learning (DL) framework is proposed to directly construct a near real-time, high-resolution daily three-dimensional ocean subsurface salinity (3D OSS) dataset from a data-driven perspective in this study. Four models with different structural combinations are designed in the China's marginal seas. Experimental results demonstrate that the models can successfully reconstruct the high-precision and high-resolution 3D OSS on a daily scale in 12 depth levels (from 2 m to 200 m). The asymmetric inception-3DGAN model with all enhanced structures has the highest accuracy, the average root-mean-squared error (RMSE) is 0.135psu, the average coefficient of determination (R2) is 0.5641 and the percent bias is 0.436%. Comparing with model without enhanced structures, the average RMSE is decreased by 22.41% when adding all the enhancement structures. Besides, temporal and spatial error analysis are also conducted to evaluate the models' performance from different aspects. Finally, the model results are used to analyze common ocean elements, such as water mass properties, dynamic fields and geostrophic velocity fields, demonstrating that the 3D OSS dataset construction approach proposed in this study not only provides some new insights into ocean observations, but also has high application values. • High precision daily 3D OSS is reconstructed via T–S mapping from a data-driven view. • Four 3D GAN based models are progressively built by powerful enhancement techniques. • Results are analyzed with in situ observations and reanalysis data. • Best reconstruction performance is found for the asymmetric inception-3DGAN case. • Reliable near real-time data can be provided for operational oceanography research. [ABSTRACT FROM AUTHOR]

Numerical ocean circulation forecasts diverge from reality due to non-deterministic processes. Through data assimilation, ocean forecasting systems routinely make corrections to scalar ocean variables of temperature and salinity derived from ocean observations. Velocity observations, however, are vector quantities that require different processing. To assimilate existing ocean velocity observations, which generally are depth limited, vertical error covariances are needed to relate the velocity errors throughout the water column. In this paper we show a velocity assimilation approach that relies on historical observations and geostrophy to relate velocity, temperature, and salinity. Geopotential calculations transform temperature and salinity covariances into corresponding geopotential covariances. This paper describes the methods and examines details of the solution process for ocean velocity data assimilation. The approach employs two separate three-dimensional variational (3DVAR) analyses combined to create one set of initial conditions during each daily forecast cycle. The first 3DVAR analysis utilizes temperature, salinity, and sea surface height anomaly data and reduces the background error variance based on these observations. The second 3DVAR analysis uses the reduced error variance and performs velocity data assimilation using surface drifter data, and these observations further reduce the error variance of temperature and salinity for the next cycle. The second 3DVAR analysis for velocity data assimilation utilizes the new full depth error covariances. We show diagnostic results from a Gulf of Mexico simulation during extensive drifter deployments in the summer of 2020. Results reveal that the two-step data assimilation cycle applies vector velocity innovations to create increments that correct the background field towards a geostrophic balance in temperature and salinity. • Surface drifter positions create Eulerian innovations for assimilation. • Empirically derived vertical covariances distribute velocity increments. • Velocity forecast error is derived from empirical vertical covariances. • Velocity increments constrain forecasted geostrophic velocities. [ABSTRACT FROM AUTHOR]

We study the production and dissipation of the eddy kinetic energy (EKE) in a submesoscale eddy field forced with downfront winds using the Process Study Ocean Model (PSOM) with a horizontal grid resolution of 0.5 km. We simulate an idealized 100 m deep mixed-layer front initially in geostrophic balance with a jet in a domain that permits eddies within a range of O(1 km–100 km). The vertical eddy viscosities and the dissipation are parameterized using four different subgrid vertical mixing parameterizations: the k − ϵ , the KPP, and two different constant eddy viscosity and diffusivity profiles with a magnitude of O ( 10 − 2 m 2 s − 1 ) in the mixed layer. Our study shows that strong vertical eddy viscosities near the surface reduce the parameterized dissipation, whereas strong vertical eddy diffusivities reduce the lateral buoyancy gradients and consequently the rate of restratification by mixed-layer instabilities (MLI). Our simulations show that near the surface, the spatial variability of the dissipation along the periphery of the eddies depends on the relative alignment of the ageostrophic and geostrophic shear. Analysis of the resolved EKE budgets in the frontal region from the simulations show important similarities between the vertical structure of the EKE budget produced by the k − ϵ and KPP parameterizations, and earlier LES studies. Such an agreement is absent in the simulations using constant eddy-viscosity parameterizations. [ABSTRACT FROM AUTHOR]

Stochastic superparameterization, a stochastic parameterization framework based on a multiscale formalism, is developed for mesoscale eddy parameterization in coarse-resolution ocean modeling. The framework of stochastic superparameterization is reviewed and several configurations are implemented and tested in a quasigeostrophic channel model – an idealized representation of the Antarctic Circumpolar Current. Five versions of the Gent–McWilliams (GM) parameterization are also implemented and tested for comparison. Skill is measured using the time-mean and temporal variability separately, and in combination using the relative entropy in the single-point statistics. Among all the models, those with the more accurate mean state have the less accurate variability, and vice versa. Stochastic superparameterization results in improved climate fidelity in comparison with GM parameterizations as measured by the relative entropy. In particular, configurations of stochastic superparameterization that include stochastic Reynolds stress terms in the coarse model equations, corresponding to kinetic energy backscatter, perform better than models that only include isopycnal height smoothing. [ABSTRACT FROM AUTHOR]

In the near future we expect the resolution of many IPCC-class ocean models to enter the "eddy-permitting" regime. At this resolution models can produce reasonable eddy-like disturbances, but can still not properly resolve geostrophic eddies at all relevant scales. Adequate parameterizations representing sub-grid eddy effects are thus necessary. Most eddy-permitting models presently employ some kind of hyper-viscosity, which is shown to cause a significant amount of energy dissipation. However, comparison to higher resolution simulations shows that only enstrophy, but almost no energy, should be dissipated below the grid-scale. As a result of the artificial energy sink associated with viscous parameterizations, the eddy fields in eddy permitting models are generally not energetic enough. To overcome this problem, we propose a class of sub-grid parameterizations which dissipate enstrophy but little or no energy. The idea is to combine a standard hyperviscous closure with some mechanism to return dissipated energy to the resolved flow. Enstrophy dissipation remains ensured because the energy is returned at larger scales. Two simple ways to return the energy are proposed: one using a stochastic excitation and one using a negative Laplacian viscosity. Both approaches are tested in an idealized two-layer quasi-geostrophic model. Either approach is shown to greatly improve the solutions in simulations with typical eddy-permitting resolutions. The adaptation of the proposed parameterization for use in realistic ocean models is discussed. [ABSTRACT FROM AUTHOR]

Highlights: [•] Predictability of small-scale ocean variability from large-scales is demonstrated. [•] Spectral nudging of large-scales can recover small-scales of a quasi-geostrophic model. [•] A practical scheme for assimilating large scales and point observations is proposed. [•] Limited point observations are more valuable if large-scales are spectrally nudged. [Copyright &y& Elsevier]

Abstract: The impact of assimilation of sea surface velocity fields observed by a set of high-frequency (HF) radars is studied using a three-dimensional ocean circulation model configured along the Oregon coast. The study period is June–July 2008 featuring upwelling and separation of the coastal currents into the adjacent interior ocean. The nonlinear model is based on the Regional Ocean Modeling System (ROMS) and the data assimilation (DA) component on the AVRORA system utilizing the representer-based variational algorithm. Assimilation proceeds in a series of 3-day windows, providing an analysis solution in each window and a 3-day forecast into the next window. Experiments with two different initial condition error covariances are compared (one is dynamically balanced, based on the linearized equation of state, temperature-salinity relation, and geostrophic and thermal wind balance relations and the other is multivariate unbalanced). While the assimilation impact is statistically better in the case of the balanced covariance, the case with the unbalanced covariance also provides sensible improvement in terms of surface velocity and sea surface temperature (SST) model-data forecast statistics. The analysis of representer functions shows that even if the initial condition error covariance is unbalanced, the correction fields at the model initial time are partially balanced after each dynamical field is smoothed independently, due to inherent dynamical properties of the adjoint model. Assimilation of the HF radar surface currents improves not only surface velocity forecasts, but also geometry of the upwelling SST front and the sea surface height (SSH) slope near the coast, as verified against unassimilated satellite SSH and SST data. The assimilation also alters the latitudinal distribution of the time-averaged offshore transport. Combined HF radar velocity and other observations, e.g., altimetry, is needed to better constrain surface geostrophic currents in the entire model domain, including the area not covered by the HF radar data. [Copyright &y& Elsevier]

Abstract: A perceived limitation of z-coordinate models associated with spurious diapycnal mixing in eddying, frontal flow, can be readily addressed through appropriate attention to the tracer advection schemes employed. It is demonstrated that tracer advection schemes developed by Prather and collaborators for application in the stratosphere, greatly improve the fidelity of eddying flows, reducing levels of spurious diapycnal mixing to below those directly measured in field experiments, ∼1×10−5 m2 s−1. This approach yields a model in which geostrophic eddies are quasi-adiabatic in the ocean interior, so that the residual-mean overturning circulation aligns almost perfectly with density contours. A reentrant channel configuration of the MIT General Circulation Model, that approximates the Antarctic Circumpolar Current, is used to examine these issues. Virtual analogs of ocean deliberate tracer release field experiments reinforce our conclusion, producing passive tracer solutions that parallel field experiments remarkably well. [Copyright &y& Elsevier]

Abstract: Accurate representation of geostrophic and hydrostatic balance is an essential requirement for numerical modelling of geophysical flows. Potentially, unstructured mesh numerical methods offer significant benefits over conventional structured meshes, including the ability to conform to arbitrary bounding topography in a natural manner and the ability to apply dynamic mesh adaptivity. However, there is a need to develop robust schemes with accurate representation of physical balance on arbitrary unstructured meshes. We discuss the origin of physical balance errors in a finite element discretisation of the Navier–Stokes equations using the fractional timestep pressure projection method. By considering the Helmholtz decomposition of forcing terms in the momentum equation, it is shown that the components of the buoyancy and Coriolis accelerations that project onto the non-divergent velocity tendency are the small residuals between two terms of comparable magnitude. Hence there is a potential for significant injection of imbalance by a numerical method that does not compute these residuals accurately. This observation is used to motivate a balanced pressure decomposition method whereby an additional “balanced pressure” field, associated with buoyancy and Coriolis accelerations, is solved for at increased accuracy and used to precondition the solution for the dynamical pressure. The utility of this approach is quantified in a fully non-linear system in exact geostrophic balance. The approach is further tested via quantitative comparison of unstructured mesh simulations of the thermally driven rotating annulus against laboratory data. Using a piecewise linear discretisation for velocity and pressure (a stabilised P 1 P 1 discretisation), it is demonstrated that the balanced pressure decomposition method is required for a physically realistic representation of the system. [Copyright &y& Elsevier]

Abstract: Surface frontogenesis and thermohaline intrusion are examined in a submesoscale eddy-resolving primitive-equation model simulation of an idealized summer shelfbreak front of the Middle Atlantic Bight (USA). The initial front evolves into fully developed geostrophic turbulence characterized by large Rossby number, strong vortical asymmetry, and k −3 eddy kinetic energy and density spectra (k is the horizontal wavenumber). The baroclinic eddies are trapped in the upper 20m, and their growth is limited by conversion from the eddy to mean kinetic energy. The surface frontogenesis drives vigorous thermohaline intrusions of density-compensated temperature and salinity anomalies. Warmer saltier slope waters are subducted in the wave troughs, while colder fresher shelf waters are upwelled in the wave crests. The model results compare well with the observed distribution of subsurface salinity maxima. The total eddy induced onshore salt flux is comparable to the estimated salt budget in the Mid-Atlantic shelf. [Copyright &y& Elsevier]

Abstract: The dynamics of large-scale river plumes are investigated in idealized numerical experiments using the HYbrid Coordinate Ocean Model (HYCOM). The focus of this study is to address how the development and structure of a buoyant plume are affected by the outflow properties, as impacted by processes within the estuary and at the point of discharge to the coastal basin. Changes in the outflow properties involved vertical and horizontal redistribution of the river inflow and enhanced vertical mixing inside an idealized estuary. The development of the buoyant plume was evaluated in a rectangular, f-plane basin with flat and sloping bottom conditions and in the absence of other external forcing. The general behavior of a mid-latitude river plume was reproduced, with the development of a surface anticyclonic bulge off the estuary mouth and a surface along-shore coastal current which flows in the direction of Kelvin wave propagation (“downstream”); the momentum balance was predominantly geostrophic. Conditions within the estuary and the outflow properties at the river mouth (where observed profiles may be available) greatly impacted the fate of riverine waters. In flat bottom conditions, larger mixing at the freshwater source enhanced the estuarine gravitational circulation, promoting larger upward entrainment and stronger outflow velocities. Although the overall geostrophic balance was maintained, estuarine mixing led to an asymmetry of the currents reaching the river mouth and to a sharp anticyclonic veering within the estuary, resulting in reduced upstream flow and enhanced downstream coastal current. These patterns were altered when the plumes evolved in the presence of a bottom slope. The anticyclonic veering of the buoyant outflow was suppressed, the offshore intrusion decreased and the recirculating bulge was displaced upstream. The sloping bottom impacts were accompanied by enhanced transport and increased downstream extent of the coastal current in most cases. No major changes in the general properties and especially the vertical structure of the plumes were observed when the vertical coordinates were changed from cartesian–isopycnal, to sigma or to sigma–isopycnal. The findings offer a benchmark for coastal studies with HYCOM, where plume dynamics should be examined in tandem with additional circulation forcing mechanisms, resulting in transitions of the vertical coordinate system that are dictated by the prevailing dynamics. [Copyright &y& Elsevier]

Abstract: The vortex in cell (VIC) method has been found to be an attractive computational choice for solving a number of fluid dynamical problems. The Lagrangian advection of particles leads to stability of the method even for long time steps. Moreover, conservation of particle properties, such as potential vorticity, can be enshrined at the heart of the numerical procedure. In this paper we describe a numerical implementation of the VIC method for a reduced gravity, quasi-geostrophic model and make use of its novel aspects to explore the interaction of waves and turbulence on the sphere. The Lagrangian advection of particles is performed using a fourth order Runge–Kutta method and the stream-function is obtained by the inversion of potential vorticity using an underlying Eulerian grid. The scheme is tested in the simulation of Rossby and Rossby–Haurwitz waves. Encouraging results are obtained for various radii of deformations corresponding to both the atmosphere and ocean. [Copyright &y& Elsevier]

Abstract: An approximate expression is derived for the geostrophic streamfunction in approximately neutral surfaces, φ n, namelyThis expression involves the specific volume anomaly defined with respect to a reference point on the surface, Δp and ΔΘ are the differences in pressure and Conservative Temperature with respect to and , respectively, and is the thermobaric coefficient. This geostrophic streamfunction is shown to be more accurate than previously available choices of geostrophic streamfunction such as the Montgomery streamfunction. Also, by writing expressions for the horizontal differences on a regular horizontal grid of a localized form of the above geostrophic streamfunction, an over-determined set of equations is developed and solved to numerically obtain a very accurate geostrophic streamfunction on an approximately neutral surface; the remaining error in this streamfunction is caused only by neutral helicity. [Copyright &y& Elsevier]

Abstract: The probability density functions (PDFs) of sea level and geostrophic relative vorticity are examined using satellite altimeter data. It is shown that departures from a Gaussian distribution can generally be represented by two functions, and that the spatial distribution of these two functions is closely linked to the skewness and kurtosis of the PDF. The patterns indicate that strong jets tend to be identified by a zero contour in skewness coinciding with a low value of kurtosis. A simple model of the statistics of a meandering frontal region is presented which reproduces these features. Comparisons with mean currents and sea surface temperature gradients confirm the identification of these features as jets, and confirm the existence of several Southern Ocean jets unresolved by drifter data. Diagnostics from a range of idealized eddying model simulations show that there is a strong, simple relationship between kurtosis of potential vorticity and effective diffusivity. This suggests that kurtosis may provide a simple method of mapping mixing barriers in the ocean. [Copyright &y& Elsevier]

Abstract: A new high-resolution Eulerian numerical method is proposed for modelling quasigeostrophic ocean dynamics in eddying regimes. The method is based on a novel, second-order non-dissipative and low-dispersive conservative advection scheme called CABARET. The properties of the new method are compared with those of several high-resolution Eulerian methods for linear advection and gas dynamics. Then, the CABARET method is applied to the classical model of the double-gyre ocean circulation and its performance is contrasted against that of the common vorticity-preserving Arakawa method. In turbulent regimes, the new method permits credible numerical simulations on much coarser computational grids. [Copyright &y& Elsevier]

Abstract: We introduce a mixed discontinuous/continuous finite element pair for ocean modelling, with continuous quadratic layer thickness and discontinuous velocity. We investigate the finite element pair applied to the linear shallow-water equations on an f-plane. The element pair has the property that all geostrophically balanced states which strongly satisfy the boundary conditions have discrete divergence equal to exactly zero and hence are exactly steady states of the discretised equations. This means that the finite element pair has excellent geostrophic balance properties. We also show that the element pair applied to the non-rotating linear shallow-water equations does not have any spurious small eigenvalues. We illustrate these properties using numerical tests and provide convergence calculations which show that the numerical solutions have errors which decay quadratically with element edge length for both velocity and layer thickness. [Copyright &y& Elsevier]

Abstract: The spin-up timescale in large-scale ocean models, i.e., the time it takes to reach an equilibrium state, is determined by the slow processes in the deep ocean and is usually in the order of a few thousand years. As these equilibrium states are taken as initial states for many calculations, much computer time is spent in the spin-up phase of ocean model computations. In this note, we propose a new approach which can lead to a very large reduction in spin-up time for quite a broad class of existing ocean models. Our approach is based on so-called Jacobian–Free Newton–Krylov methods which combine Newton’s method for solving non-linear systems with Krylov subspace methods for solving large systems of linear equations. As there is no need to construct the Jacobian matrices explicitly the method can in principle be applied to existing explicit time-stepping codes. To illustrate the method we apply it to a 3D planetary geostrophic ocean model with prognostic equations only for temperature and salinity. We compare the new method to the ‘ordinary’ spin-up run for several model resolutions and find a considerable reduction of spin-up time. [Copyright &y& Elsevier]

Abstract: The development of an adaptive (in space and time) ocean model from an existing adaptive finite-volume Navier–Stokes model is described. A flexible and efficient quadtree spatial discretisation is used which requires collocation of all variables (i.e. an A-grid discretisation). We demonstrate that the use of an approximate projection method allows for implicit damping of instabilities generally associated with the A-grid, at the expense of a relatively small amount of numerical energy dissipation, while accurately preserving dispersive properties and geostrophic balance. The finite-volume formulation also maintains second-order spatial accuracy at all solid boundaries. Test cases demonstrate the efficacy of the adaptive ocean model, and the advantages it has in terms of efficient representation of multi-scale behaviour within a single model. The model is freely available as open-source code. [Copyright &y& Elsevier]

In contrast with the severe thinning of ice shelves along the coast of West Antarctica, large ice shelves (specifically, the Filchner–Ronne and Amery Ice Shelves) with deep grounding lines gained mass during the period 1994–2012. This positive mass budget is potentially associated with the marine ice production, which originates from the supercooled Ice Shelf Water plume carrying suspended frazil ice along the ice shelf base. In addition, the outflow of this supercooled plume from beneath the ice shelf arguably exerts a significant impact on the properties of Antarctic Bottom Water, as well as its production. However, knowledge of this buoyant and supercooled shear flow is still limited, let alone its structure that is generally assumed to be vertically uniform. In this study we extended the vertical one-dimensional model of ice shelf–ocean boundary current from Jenkins (2016) by incorporating a frazil ice module and a fairly sophisticated turbulence closure (i.e., k - ε model) with the effects of density stratification. On the basis of this extended model, the study reproduced the measured thermohaline properties of a perennially-prominent supercooled ice shelf–ocean​ boundary current underneath the Amery Ice Shelf in East Antarctica, and conducted extensive sensitivity runs to a wide range of factors, including advection of scalar quantities, far-field geostrophic currents, basal slope, and the distribution of frazil ice crystal size. Based on the simulation results, the following conclusions can be drawn: Firstly, it can be difficult to reasonably reproduce the vertical structure of the ice shelf–ocean boundary current using a constant eddy viscosity/diffusivity near the ice shelf base. Secondly, although there are no direct observations of the size of frazil ice crystals beneath the ice shelves, the size of the finest ice crystals that play an important role in controlling the ice shelf–ocean boundary current is strongly suggested. Lastly, but most importantly, the ice shelf–ocean boundary layer response to the vertical gradient of frazil ice concentration will significantly reduce the level of turbulence. Therefore, this study highlights the importance of the strong interaction between frazil ice formation and the hydrodynamics and thermodynamics of ice shelf–ocean boundary layer. This interaction must not only be included, but also be resolved at high resolutions in three-dimensional coupled ice shelf–ocean models applied to cold ice cavities, which will have a potential impact on the overall ice shelf mass balance and the Antarctic Bottom Water production. • A vertical 1D model of ice shelf-ocean boundary current with frazil ice is developed. • Sensitivity of frazil-laden ice shelf–ocean boundary current is extensively examined. • Interactions between frazil ice and ice shelf–ocean boundary current are highlighted. [ABSTRACT FROM AUTHOR]

Motivated by recent observations of coherent dipolar cyclone–anticyclone structures in the ocean, the modons, and their signature in the surface temperature field, we demonstrate that the classical modon solutions of the barotropic quasi-geostrophic equations can be generalized to include buoyancy or temperature as an active tracer. The properties of such "thermal" modons, and especially their ability to carry heat anomaly over long distances, depend on the relative sign of the associated vorticity and buoyancy anomalies. We show using numerical simulations with the thermal shallow water equations, and their quasi-geostrophic version, that the evolution of the modons is consistent with the observations. • Existence of coherent dipolar vortices is demonstrated in thermal shallow water model. • The evolution of these vortices is consistent with recent observations. • Their transport and mixing properties depend on relative signs of vorticity and buoyancy anomalies. [ABSTRACT FROM AUTHOR]

A pair of 12.5-year (July 2002–December 2014) HYbrid Coordinate Ocean Model (HYCOM) simulations that only differ in the wind stress formulation are used to investigate the effect of ocean surface currents on global monthly eddy kinetic energy (EKE) variation. The model results (2004–2014) show that the global monthly mean EKE is reduced by 37%, from 1.76 EJ (1018 J) to 1.10 EJ after ocean surface currents are included in the wind stress formulation. The monthly EKE budget indicates that the shear production and buoyancy work are positive (energy source) and the eddy wind work on the geostrophic currents is negative (energy sink) in the steady state (2004–2014) for both simulations. All of these three terms are reduced in the steady state when the ocean currents are included in the wind stress formulation. The global integral of the EKE difference budget suggests that the EKE reduction is primarily due to the reduction of the buoyancy work, followed by the reduction of the wind work on the geostrophic currents and the shear production. To our knowledge this is the first study to separate the eddy wind work into the geostrophic and ageostrophic components to investigate the impact of ocean surface currents on global and depth integrated EKE via the wind stress formulation using HYCOM simulations. • Including ocean surface currents in the wind stress formulation reduces global eddy kinetic energy. • Eddy component of wind work has a much larger magnitude than shear production and buoyancy work. • The reduction of global eddy kinetic energy is primarily due to the reduction of the buoyancy work. • Reduction of the wind work on geostrophic currents also contributes to the reduction of the global eddy kinetic energy. • Reduction of the shear production also contributes to the reduction of the global eddy kinetic energy. [ABSTRACT FROM AUTHOR]