Steering of upper ocean currents and fronts by the topographically constrained abyssal circulation

A two-layer theory is used to investigate (1) the steering of upper ocean current pathways by topographically constrained abyssal currents that do not impinge on the bottom topography and (2) its application to upper ocean – topographic coupling via flow instabilities where topographically constrained eddy-driven deep mean flows in turn steer the mean pathways of upper ocean currents and associated fronts. In earlier studies the two-layer theory was applied to ocean models with low vertical resolution (2–6 layers). Here we investigate its relevance to complex ocean general circulation models (OGCMs) with high vertical resolution that are designed to simulate a wide range of ocean processes. The theory can be easily applied to models ranging from idealized to complex OGCMs, provided it is valid for the application. It can also be used in understanding some persistent features seen in observed ocean frontal pathways (over deep water) derived from satellite imagery and other data. To facilitate its application, a more thorough explanation of the theory is presented that emphasizes its range of validity. Three regions of the world ocean are used to investigate its application to eddy-resolving ocean models with high vertical resolution, including one where an assumption of the two-layer theory is violated. Results from the OGCMs with high vertical resolution are compared to those from models with low vertical resolution and to observations. In the Kuroshio region upper ocean – topographic coupling via flow instabilities and a modest seamount complex are used to explain the observed northward mean meander east of Japan where the Kuroshio separates from the coast. The Japan/East Sea (JES) is used to demonstrate the impact of upper ocean – topographic coupling in a relatively weak flow regime. East of South Island, New Zealand, the Southland Current is an observed western boundary current that flows in a direction counter to the demands of Sverdrup flow and counter to the direction simulated in nonlinear global flat bottom and reduced gravity models. A model with high vertical resolution (and topography extending through any number of layers) and a model with low vertical resolution (and vertically compressed but otherwise realistic topography confined to the lowest layer) both simulate a Southland Current in the observed direction with dynamics depending on the configuration of the regional seafloor. However, the dynamics of these simulations are very different because the Campbell Plateau and Chatham Rise east and southeast of New Zealand are rare features of the world ocean where the topography intrudes into the stratified water column over a relatively broad area but lies deeper than the nominal 200 m depth of the continental shelf break, violating a limitation of the two-layer theory. Observations confirm the results from the high vertical resolution model. Overall, the model simulations show increasingly widespread upper ocean – topographic coupling via flow instabilities as the horizontal resolution of the ocean models is increased, but fine resolution of mesoscale variability and the associated flow instabilities are required to obtain sufficient coupling. As a result, this type of coupling is critical in distinguishing between eddy-resolving and eddy-permitting ocean models in regions where it occurs.     

The Gulf Stream pathway and the impacts of the eddy-driven abyssal circulation and the Deep Western Boundary Current

A hydrodynamic model of the subtropical Atlantic basin and the Intra-Americas Sea (9–47°N) is used to investigate the dynamics of Gulf Stream separation from the western boundary at Cape Hatteras and its mean pathway to the Grand Banks. The model has five isopycnal Lagrangian layers in the vertical and allows realistic boundary geometry, bathymetry, wind forcing, and a meridional overturning circulation (MOC), the latter specified via ports in the northern and southern boundaries. The northward upper ocean branch of the MOC (14 Sv) was always included but the southward Deep Western Boundary Current (DWBC) was excluded in some simulations, allowing investigation of the impacts of the DWBC and the eddy-driven mean abyssal circulation on Gulf Stream separation from the western boundary. The result is resolution dependent with the DWBC playing a crucial role in Gulf Stream separation at 1/16° resolution but with the eddy-driven abyssal circulation alone sufficient to obtain accurate separation at 1/32° resolution and a realistic pathway from Cape Hatteras to the Grand Banks with minimal DWBC impact except southeast of the Grand Banks. The separation from the western boundary is particularly sensitive to the strength of the eddy-driven abyssal circulation. Farther to the east, between 68°W and the Grand Banks, all of the 1/16° and 1/32° simulations with realistic topography (with or without a DWBC) gave similar generally realistic mean pathways with clear impacts of the topographically constrained eddy-driven abyssal circulation versus very unrealistic Gulf Stream pathways between Cape Hatteras and the Grand Banks from otherwise identical simulations run with a flat bottom, in reduced-gravity mode, or with 1/8° resolution and realistic topography. The model is realistic enough to allow detailed model-data comparisons and a detailed investigation of Gulf Stream dynamics. The corresponding linear solution with a Sverdrup interior and Munk viscous western boundary layers, including one from the northward branch of the MOC, yielded two unrealistic Gulf Stream pathways, a broad eastward pathway centered at the latitude of Cape Hatteras and a second wind plus MOC-driven pathway hugging the western boundary to the north. Thus, a high resolution model capable of simulating an inertial jet is required to obtain a single nonlinear Gulf Stream pathway as it separates from the coast. None of the simulations were sufficiently inertial to overcome the linear solution need for a boundary current north of Cape Hatteras without assistance from pathway advection by the abyssal circulation, even though the core speeds of the simulated currents were consistent with observations near separation. In the 1/16° simulation with no DWBC and a 1/32° simulation with high bottom friction and no DWBC the model Gulf Stream overshot the observed separation latitude. With abyssal current assistance the simulated (and the observed) mean Gulf Stream pathway between separation from the western boundary and 70°W agreed closely with a constant absolute vorticity (CAV) trajectory influenced by the angle of the coastline prior to separation. The key abyssal current crosses under the Gulf Stream at 68.5–69°W and advects the Gulf Stream pathway southward to the terminus of an escarpment in the continental slope. There the abyssal current crosses to deeper depths to conserve potential vorticity while passing under the downward-sloping thermocline of the stream and then immediately retroflects eastward onto the abyssal plain, preventing further southward pathway advection. Thus specific topographic features and feedback from the impact of the Gulf Stream on the abyssal current pathway determined the latitude of the stream at 68.5–69°W, a latitude verified by observations. The associated abyssal current was also verified by observations.     

Timescales and dynamics of the formation of a thermocline

In a state of equilibrium, the constraint of a balanced heat budget for the ocean strongly influences the depth of the tropical thermocline because that depth controls the rate at which the ocean absorbs heat from the atmosphere. Thus, an increase in the oceanic heat loss in high latitudes results in a shoaling of the equatorial thermocline so that the heat gain also increases. How does the ocean adjust to such a new equilibrium state after an abrupt change in the heat flux in high latitudes? The adjustment of the wind-driven circulation of the upper ocean is shown to involve two timescales. The first is the familiar adiabatic wave-adjustment time associated with the horizontal redistribution of warm water above the thermocline in shallow water models. (This is essentially the time it takes Rossby and Kelvin waves to propagate from the disturbed extra-equatorial region to the equator.) The second adjustment-time is associated with the diabatic processes that come into play once the waves from higher latitudes modify the thermal structure in low latitudes and hence the flux of heat into the ocean; it is the timescale on which the ocean recovers a balanced heat budget. The identification of this timescale is the main result of this paper.Through a series of simulations of an idealized ocean basin, we identify the diabatic timescale and argue that it is determined by the strength of the upwelling and the intensity of the air–sea heatfluxes. By simulating the formation of a thermocline from isothermal conditions, we are able to relate this timescale to other relevant timescales such as that associated with diffusive processes and the adiabatic timescale invoked by Gu and Philander [Gu, D., Philander, S.G.H., 1997. Interdecadal climate fluctuations that depend on exchanges between the tropics and extra-topics. Science 275, 805–807].     

The effect of ageostrophy on the stability of thin oceanic vortices

This paper examines the stability of vortices in a two-layer ocean on the f-plane. The mean depth of the upper layer is assumed to be much smaller than the depth of the lower layer. Using the primitive equations, we derive an asymptotic criterion for baroclinic instability of compensated (i.e. confined to the upper layer) vortices. Surprisingly, it coincides exactly with a similar criterion derived from the quasigeostrophic equations [Benilov, E.S., 2003. Instability of quasigeostrophic vortices in a two-layer ocean with thin upper layer. J. Fluid Mech. 475, 303–331]. Thus, to leading order in , ageostrophy does not affect the stability properties of thin compensated vortices. As a result, whether a vortex is stable or not, depends on its shape, not amplitude (although the growth rate of an unstable vortex does depend on its amplitude).     

Impact of 1/8° to 1/64° resolution on Gulf Stream model–data comparisons in basin-scale subtropical Atlantic Ocean models

We investigate the impact of 1/8°, 1/16°, 1/32°, and 1/64° ocean model resolution on model–data comparisons for the Gulf Stream system mainly between the Florida Straits and the Grand Banks. This includes mean flow and variability, the Gulf Stream pathway, the associated nonlinear recirculation gyres, the large-scale C-shape of the subtropical gyre and the abyssal circulation. A nonlinear isopycnal, free surface model covering the Atlantic from 9°N to 47°N or 51°N, including the Caribbean and Gulf of Mexico, and a similar 1/16° global model are used. The models are forced by winds and by a global thermohaline component via ports in the model boundaries. When calculated using realistic wind forcing and Atlantic model boundaries, linear simulations with Munk western boundary layers and a Sverdrup interior show two unrealistic mean Gulf Stream pathways between Cape Hatteras and the Grand Banks, one proceeding due east from Cape Hatteras and a second one continuing northward along the western boundary until forced eastward by the regional northern boundary. The northern pathway is augmented when a linear version of the upper ocean global thermohaline contribution to the Gulf Stream is added as a Munk western boundary layer. A major change is required to obtain a realistic pathway in nonlinear models. Resolution of 1/8° is eddy-resolving but mainly gives a wiggly version of the linear model Gulf Stream pathway and weak abyssal flows except for the deep western boundary current (DWBC) forced by ports in the model boundaries. All of the higher resolution simulations show major improvement over the linear and 1/8° nonlinear simulations. Additional major improvement is seen with the increase from 1/16° to 1/32° resolution and modest improvement with a further increase to 1/64°. The improvements include (1) realistic separation of the Gulf Stream from the coast at Cape Hatteras and a realistic Gulf Stream pathway between Cape Hatteras and the Grand Banks based on comparisons with Gulf Stream pathways from satellite IR and from GEOSAT and TOPEX/Poseidon altimetry (but 1/32° resolution was required for robust results), (2) realistic eastern and western nonlinear recirculation gyres (which contribute to the large-scale C-shape of the subtropical gyre) based on comparisons with mean surface dynamic height from the generalized digital environmental model (GDEM) oceanic climatology and from the pattern and amplitude of sea surface height (SSH) variability surrounding the eastern gyre as seen in TOPEX/Poseidon altimetry, (3) realistic upper ocean and DWBC transports based on several types of measurements, (4) patterns and amplitude of SSH variability which are generally realistic compared to TOPEX/Poseidon altimetry, but which vary from simulation to simulation for specific features and which are most realistic overall in the 1/64° simulation, (5) a basin wide explosion in the number and strength of mesoscale eddies (with warm core rings (WCRs) north of the Gulf Stream, the regional eddy features best observed by satellite IR), (6) realistic statistics for WCRs north of the Gulf Stream based on comparison to IR analyses (low at 1/16° resolution and most realistic at 1/64° resolution for mean population and rings generated/year; realistic ring diameters at all resolutions), and (7) realistic patterns and amplitude of abyssal eddy kinetic energy (EKE) in comparison to historical measurements from current meters.     

Influence of bottom topography roughness on the jet and inertial recirculation of a mid-latitude gyre

Several numerical experiments are conducted to examine the influence of mesoscale, bottom topography roughness on the inertial circulation of a wind-driven, mid-latitude ocean gyre. The ocean model is based on the quasi-geostrophic formulation, and is eddy-resolving as it features high vertical and horizontal resolutions (six layers and a 10 km grid). An antisymmetrical double-gyre wind stress curl forces the baroclinic modes and generates a strong surface jet. In the case of a flat bottom, inertia and inverse energy cascade force the barotropic mode, and the resulting circulation features strong, barotropic, inertial gyres. The sea-floor roughness inhibits the inertial circulation in the deep layers; the barotropic component of the flow is then forced by eddy-topography interactions, and its energy concentrates at the scales of the topography. As a result, the baroclinicity of the flow is intesified: the barotropic mode is reduced with regard to the baroclinic modes, and the bottom flow (constrained by the mesoscale sea-floor roughness) is decoupled from the surface flow (forced by the gyre-scale wind). Rectified, mesoscale bottom circulation induces an interfacial form stress at the thermocline, which enhances horizontal shear instability and opposes the eastward penetration of the jet. The mean jet is consequently shortened, but the instantaneous jet remains very turbulent, with meanders of large meridional extent. The sea-floor roughness modifies the energy pathways, and the eddies have an even more important role in the establishment of the mean circulation: below the thermocline, rectification processes are dominant, and eddies transfer energy toward permanent mesoscale circulations strongly correlated with topography, whereas above the thermocline mean flow and eddy generation are influenced by the mean bottom circulation through interfacial stress. The topography modifies the vorticity of the barotropic and highest baroclinic modes. Vorticity accumulates at the small topographic scales, and the vorticity content of the highest modes, which is very weak in the flat-bottom case, increases significantly. Few changes occur in surface-intensified modes. In the deep layers of the model, the inverse correlation between relative vorticity and topography at small scales ensures the homogenization of the potential vorticity, which mainly retains the largest scales of the bottom flow and the scale of β.     

An upper ocean general circulation model for climate studies: global simulation with seasonal cycle

The global ocean circulation with a seasonal cycle has been simulated with a two-and-a-half layer upper-ocean model. This model was developed for the purpose of coupling to an atmospheric general circulation model for climate studies on decadal time scales. The horizontal resolution is 4° latitude by 5° longitude and is thus not eddy-resolving. Effects of bottom topography are neglected. In the vertical, the model resolves the oceanic mixed layer and the thermocline. A thermodynamic sea-ice model is coupled to the mixed layer. The model is forced at the surface with seasonally varying (a) observed wind stress, (b) heat fluxes, as defined by an atmospheric equilibrium temperature, and (c) Newtonian-type surface salt fluxes. The second layer is coupled to the underlying deep ocean through Newtonian-type diffusive heat and salt fluxes, convective overturning, and mass entrainment in the upwelling regions of the subpolar gyres. The overall global distributions of mixed layer temperature, salinity and thickness are favorably reproduced. Inherent limitations due to coarse horizontal resolution result in large mixed-layer temperature errors near continental boundaries and in weak current systems. Sea ice distributions agree well with observations except in the interiors of the Ross and Weddell Seas. A realistic time rate of change of heat storage is simulated. There is also realistic heat transport from low to high latitudes.     

The Kuroshio Current System as a jet and twin “relative” recirculation gyres embedded in the Sverdrup circulation

By analyzing the results of a realistic ocean general circulation model (OGCM) and conducting a series of idealized OGCM experiments, the dynamics of the Kuroshio Current System is examined. In the realistic configuration, the Kuroshio Current System is successfully simulated when the horizontal resolution of OGCMs is increased from 1/2° to 1/10°. The difference between the two experiments shows a jet, the model’s Kuroshio Extension, and a pair of cyclonic and anticyclonic, “relative,” recirculation gyres (RRGs) on the northern and southern flanks of the jet. We call them recirculation gyres because they share some features with ordinary recirculation gyres in previous studies, and we add the adjective “relative” to emphasize that they may not be apparent in the total field. Similar zonal jet and RRGs are obtained also in the idealized model with a rectangular basin and a flat bottom with a horizontal resolution of 1/6°. The northern RRG is generated by the injection of high potential vorticity (PV) created in the viscous sublayer of the western boundary current, indicating the importance of a no-slip boundary condition. Since there is no streamline with such high PV in the Sverdrup interior, the eastward current in the northern RRG region has to lose its PV anomaly by viscosity before connecting to the interior. In the setup stage this injection of high PV is carried out by many eddies generated from the instability of the western boundary current. This high PV generates the northern RRG, which induces the separation of the western boundary current and the formation of the zonal jet. In the equilibrium state, the anomalous high PV values created in the viscous sublayer are carried eastward in the northern flank of the zonal jet. The southern RRG is due to the classical Rhines–Young mechanism, where low PV values are advected northward within the western boundary inertial sublayer, and closed, PV-conserving streamlines form to the south of the Kuroshio Extension, allowing slow homogenization of the low PV anomalies. The westward-flowing southern branch of this southern RRG stabilizes the inertial western boundary current and prevents its separation in the northern half of the Sverdrup subtropical gyre, where the western boundary current is unstable without the stabilizing effect of the southern RRG. Therefore, in the equilibrium state, the southern RRG should be located just to the north of the center of the Sverdrup subtropical gyre, which is defined as the latitude of the Sverdrup streamfunction maximum. The zonal jet (the Kuroshio Extension) and the northern RRG gyre are formed to the north of the southern RRG. This is our central result. This hypothesis is confirmed by a series of sensitivity experiments where the location of the center of the Sverdrup subtropical gyre is changed without changing the boundaries of the subtropical gyre. The locations of the zonal jets in the observed Kuroshio Current System and Gulf Stream are consistent as well. Sensitivities of the model Kuroshio Current System are also discussed with regard to the horizontal viscosity, strength of the wind stress, and coastline.     

Design and Numerical Simulation of an Arctic Ocean Circulation and Thermodynamic Sea－Ice Model

In this paper, the first version of a new Arctic Ocean circulation and thermodynamic sea-ice model is presented by the authors based on the framework of a twenty-layer World Oceanic general circulation model developed by Zhang et al. in 1994, The model’s domain covers the Arctic Ocean and Greenland-Norwegian Seas with the horizon-tal resolution of 200 km × 200 km on a stereographic projection plane. In vertical, the model uses the Eta-coordinate (Sigma modified to have quasi-horizontal coordinate surfaces) and has ten unevenly-spaced layers to cover the deep-est water column of 3000 m. Two 150-year integrations of coupling the ocean circulation model with the sea-ice model have been performed with seasonally cyclic surface boundary conditions. The only difference between the two experiments is in the model’s geography. Some preliminary analyses of the experimental results have been done fo-cused on the following aspects: (1) surface layer temperature, salinity and current; (2) the "Atlantic Layer"; (3) sea-ice cover and its seasonal variation. In comparison with the available observational data, these results are accept-able with reasonable accuracy.     

Overturning and wind-driven circulation in a low-order ocean–atmosphere model

A low-order ocean–atmosphere model is presented which combines coupling through heat exchange at the interface and wind stress forcing. The coupling terms are derived from the boundary conditions and the forcing terms of the constituents. Both the ocean and the atmosphere model are based on Galerkin truncations of the basic fluid dynamical equations. Hence, the coupled model can readily be extended to include more physics and more detail. The model presented here is the simplest of a hierarchy of low-order ocean–atmosphere models. The behaviour of the coupled model is investigated by means of geometric singular perturbation theory and bifurcation analysis. Two ways are found in which the slow time scales can play a role in the coupled dynamics. In the first scenario, a limit cycle on the overturning time scale is created. The associated oscillatory behaviour is governed by internal ocean dynamics. In the second scenario, intermittent behaviour occurs between periodic and chaotic regimes in parameter space.     

Stability dependent parameterisations of vertical mixing in ocean general circulation models

Summary Parameterisations of mixing induced through shear instability, internal wave breaking, and double diffusion are investigated in simulations of ocean climate using a global ocean general circulation model (OGCM). Focus is placed on the sensitivity of the large scale circulation, water mass formation and transport of heat as measures of the model's ability to represent current climate. The model resolution is typical of OGCMs being coupled to atmospheric. GCMs in climate models and the parameterisations investigated are all computationally inexpensive enough to allow for integrations on long time scales. Under the assumption of constant vertical eddy coefficients (the control case), the model climatology displays acceptable values of North Atlantic Deep Water formation, Antarctic Circumpolar Current (ACC) transport, and Indonesian through-flow but an excessively deep and diffuse pycnocline structure with weak stratification in the deep ocean. It is found that various circulation and water mass properties are sensitive to the choice of parameterisation of vertical mixing and that determining a scheme which works satisfactorily over all regions (tropical, mid-latitude, and polar) of the domain is not straightforward. Parameterisations of internal wave breaking or upper ocean shear instability lead to some improvements in the model water mass formation. ACC and poleward heat transport when compared to the control case whereas parameterisations of double diffusive processes did not. Based on these and other results, various recommendations are made for mixing parameterisations in ocean climate models.With 8 Figures     

Impact of ocean model resolution on CCSM climate simulations

The current literature provides compelling evidence suggesting that an eddy-resolving (as opposed to eddy-permitting or eddy-parameterized) ocean component model will significantly impact the simulation of the large-scale climate, although this has not been fully tested to date in multi-decadal global coupled climate simulations. The purpose of this paper is to examine how resolved ocean fronts and eddies impact the simulation of large-scale climate. The model used for this study is the NCAR Community Climate System Model version 3.5 (CCSM3.5)—the forerunner to CCSM4. Two experiments are reported here. The control experiment is a 155-year present-day climate simulation using a 0.5° atmosphere component (zonal resolution 0.625 meridional resolution 0.5°; land surface component at the same resolution) coupled to ocean and sea-ice components with zonal resolution of 1.2° and meridional resolution varying from 0.27° at the equator to 0.54° in the mid-latitudes. The second simulation uses the same atmospheric and land-surface models coupled to eddy-resolving 0.1° ocean and sea-ice component models. The simulations are compared in terms of how the representation of smaller scale features in the time mean ocean circulation and ocean eddies impact the mean and variable climate. In terms of the global mean surface temperature, the enhanced ocean resolution leads to a ubiquitous surface warming with a global mean surface temperature increase of about 0.2?°C relative to the control. The warming is largest in the Arctic and regions of strong ocean fronts and ocean eddy activity (i.e., Southern Ocean, western boundary currents). The Arctic warming is associated with significant losses of sea-ice in the high-resolution simulation. The sea surface temperature gradients in the North Atlantic, in particular, are better resolved in the high-resolution model leading to significantly sharper temperature gradients and associated large-scale shifts in the rainfall. In the extra-tropics, the interannual temperature variability is increased with the resolved eddies, and a notable increases in the amplitude of the El Ni?o and the Southern Oscillation is also detected. Changes in global temperature anomaly teleconnections and local air-sea feedbacks are also documented and show large changes in ocean–atmosphere coupling. In particular, local air-sea feedbacks are significantly modified by the increased ocean resolution. In the high-resolution simulation in the extra-tropics there is compelling evidence of stronger forcing of the atmosphere by SST variability arising from ocean dynamics. This coupling is very weak or absent in the low-resolution model.     

Sensitivity studies with the North Atlantic sigma coordinate Princeton Ocean Model

The sigma coordinate, Princeton Ocean Model (POM) has been configured for the North Atlantic Ocean between 5°N and 50°N as part of data assimilation, model predictability and intercomparison studies. The model uses a curvilinear orthogonal grid with higher resolution in the western North Atlantic and lower resolution in the eastern North Atlantic. A series of experiments, each one of a 10-year duration, are performed to evaluate the sensitivity of the ocean mean state and variability to model parameters and model configuration; these experiments include open vs. closed boundary conditions, low vs. high resolution grids, and different choices of diffusion and viscosity. The results show that the use of closed boundaries together with near-boundary buffer zones where temperature and salinity are relaxed towards the observed values give less realistic flows, weaker recirculation gyres and less realistic Gulf Stream separation than do open boundary conditions. The experiments show that the sensitivity of the ocean variability in the model to the choice of the Smagorinsky diffusion and viscosity coefficients significantly differs from one region to another and largely depends on other attributes such as the mean position of the Gulf Stream in each simulation. A 50% change in model resolution in the Gulf Stream region has a larger effect on ocean variability than a change of diffusivity by a factor of 10. In areas where either the high or the low resolution models have sufficient resolution, as in the Gulf of Mexico, they are able to produce variability comparable to that observed from altimeter data; elsewhere, model variability is underestimated.     

FGOALS耦合模式两个版本的海洋热吸收与气候敏感度的关系研究

海洋在气候变暖过程中的重要性通常用海洋热吸收来衡量,热吸收的大小影响全球变暖的幅度。本文利用FGOALS-g2、FGOALS-s2（以下分别缩写为g2、s2）两个耦合模式的CO₂浓度以每年1%速率增长（1pctCO₂）试验,评估和分析海洋热吸收与气候敏感度的关系。结果表明:进入海洋净热通量（s2模式大于g2模式）会使得s2模式的海洋热吸收总体比g2模式大;更为重要的是,由于s2模式中的海洋热吸收主要集中在上层,使得耦合模式s2中的瞬态气候响应（TCR,或称气候敏感度）比g2大。当CO₂浓度加倍时,在两个耦合模式中,海洋热吸收的空间分布呈现显著性的差异,s2模式中上层热吸收明显比深层大,上层热吸收主要位于太平洋和印度洋,而g2模式中上层和深层热吸收差别较小,深层主要位于大西洋和北冰洋。进一步研究表明,海洋热吸收分布特征与两个耦合模式海洋环流变化有关。在g2模式中北大西洋经圈翻转环流（AMOC）强度强且深度大,在CO₂浓度加倍时,AMOC减弱小,这样AMOC可将热量带到海洋的深层,增加海洋深层热吸收。而在s2模式中,平均AMOC弱且浅,在CO₂浓度加倍时,AMOC减弱明显,热量不易到达深层,主要集中在海洋上层,对气候敏感度影响更快且更强。海洋环流导致热吸收及其空间差异同时影响到气候敏感度的差异。因此,探讨海洋热吸收与气候敏感度之间的关系,利于明确气候敏感度不确定性的来源。     

Modeled Sverdrup flow in the North Atlantic from 11 different wind stress climatologies

In studies of large-scale ocean dynamics, often quoted values of Sverdrup transport are computed using the Hellerman–Rosenstein wind stress climatology. The Sverdrup solution varies, however, depending on the wind set used. We examine the differences in the large-scale upper ocean response to different surface momentum forcing fields for the North Atlantic Ocean by comparing the different Sverdrup interior/Munk western boundary layer solutions produced by a 1/16° linear numerical ocean model forced by 11 different wind stress climatologies. Significant differences in the results underscore the importance of careful selection of a wind set for Sverdrup transport calculation and for driving nonlinear models. This high-resolution modeling approach to solving the linear wind-driven ocean circulation problem is a convenient way to discern details of the Sverdrup flow and Munk western boundary layers in areas of complicated geometry such as the Caribbean and Bahamas. In addition, the linear solutions from a large number of wind sets provide a well-understood baseline oceanic response to wind stress forcing and thus, (1) insight into the dynamics of observed circulation features, by themselves and in conjunction with nonlinear models, and (2) insight into nonlinear model sensitivity to the choice of wind-forcing product.The wind stress products are evaluated and insight into the linear dynamics of specific ocean features is obtained by examining wind stress curl patterns in relation to the corresponding high-resolution linear solutions in conjunction with observational knowledge of the ocean circulation. In the Sverdrup/Munk solutions, the Gulf Stream pathway consists of two branches. One separates from the coast at the observed separation point, but penetrates due east in an unrealistic manner. The other, which overshoots the separation point at Cape Hatteras and continues to flow northward along the continental boundary, is required to balance the Sverdrup interior transport. A similar depiction of the Gulf Stream is commonly seen in the mean flow of nonlinear, eddy-resolving basin-scale models of the North Atlantic Ocean. An O(1) change from linear dynamics is required for realistic simulation of the Gulf Stream pathway. Nine of the eleven Sverdrup solutions have a C-shaped subtropical gyre, similar to what is seen in dynamic height contours derived from observations. Three mechanisms are identified that can contribute to this pattern in the Sverdrup transport contours. Along 27°N, several wind sets drive realistic total western boundary current transport (within 10% of observed) when a 14 Sv global thermohaline contribution is added (COADS, ECMWF 10 m re-analysis and operational, Hellerman–Rosenstein and National Centers for Environmental Prediction (NCEP) surface stress re-analysis), a few drive transport that is substantially too high (ECMWF 1000 mb re-analysis and operational and Isemer–Hasse) and Fleet Numerical Meteorology and Oceanography Center (FNMOC) surface stresses give linear transport that is slightly weaker than observed. However, higher order dynamics are required to explain the partitioning of this transport between the Florida Straits and just east of the Bahamas (minimal in the linear solutions vs. 5 Sv observed east of the Bahamas). Part of the Azores Current transport is explained by Sverdrup dynamics. So are the basic path of the North Atlantic Current (NAC) and the circulation features within the Intra-Americas Sea (IAS), when a linear rendition of the northward upper ocean return flow of the global thermohaline circulation is added in the form of a Munk western boundary layer.     

A world ocean model for greenhouse sensitivity studies: resolution intercomparison and the role of diagnostic forcing

We have developed an improved version of a world ocean model with the intention of coupling to an atmospheric model. This article documents the simulation capability of this 1° global ocean model, shows improvements over our earlier 5° version, and compares it to features simulated with a 0.5° model. These experiments use a model spin-up methodology whereby the ocean model can subsequently be coupled to an atmospheric model and used for order 100-year coupled model integrations. With present-day computers, 1° is a reasonable compromise in resolution that allows for century-long coupled experiments. The 1° ocean model is derived from a 0.5°-resolution model developed by A. Semtner (Naval Postgraduate School) and R. Chervin (National Center for Atmospheric Research) for studies of the global eddy-resolving world ocean circulation. The 0.5° bottom topography and continental outlines have been altered to be compatible with the 1° resolution, and the Arctic Ocean has been added. We describe the ocean simulation characteristics of the 1° version and compare the result of weakly constraining (three-year time scale) the three-dimensional temperature and salinity fields to the observations below the thermocline (710 m) with the model forced only at the top of the ocean by observed annual mean wind stress, temperature, and salinity. The 1° simulations indicate that major ocean circulation patterns are greatly improved compared to the 5° version and are qualitatively reproduced in comparison to the 0.5° version. Using the annual mean top forcing alone in a 100-year simulation with the 1° version preserves the general features of the major observed temperature and salinity structure with most climate drift occurring mainly beneath the thermocline in the first 50–75 years. Because the thermohaline circulation in the 1° version is relatively weak with annual mean forcing, we demonstrate the importance of the seasonal cycle by performing two sensitivity experiments. Results show a dramatic intensification of the meridional overturning circulation (order of magnitude) with perpetual winter surface temperature forcing in the North Atlantic and strong intensification (factor of three) with perpetual early winter temperatures in that region. These effects are felt throughout the Atlantic (particularly an intensified and northward-shifted Gulf Stream outflow). In the Pacific, the temperature gradient strengthens in the thermocline, thus helping counter the systematic error of a thermocline that is too diffuse.Partial support is provided by the Office of Health and Environmental Research of the US Department of Energy The National Center for Atmospheric Research is sponsored by the National Science Foundation     

On deep mean flow generation mechanisms and the abyssal circulation of numerical model gyres

Mesoscale resolution ocean general circulation model (EGCM) experiments have been carried out under a variety of different model physical assumptions, and the different model systems often produce very different deep mean flow fields. The flat bottom, rectangular basin experiments exhibit two distinct types of deep mean flow, which are here called “corotating” and “counterrotating”. Counterrotating deep flow, in which two adjacent deep gyres, with circulation of opposite senses, underlie the upper ocean eastward jet and its recirculation, has been found only in models with adiabetic two-layer model physics. None of the more complex model systems exhibit counterrotating deep flows; this type of flow is apparently restricted to a particular range of forcing/dissipation parameter space and/or particular model physical assumptions.Since the deep flow in these EGCM systems is generally weak, geostrophic dynamics provides the basic deep flow interior balance and the mean vertical velocity field, through the lower layer vorticity equation, largely determines the deep interior flow. The dynamical constraints on the mean vertical velocity field introduced by different model physical equations are reviewed and the adiabatic quasi-geostrophic (QG) two-layer model system is shown to be strongly constrained in several respects. In particular, the idea that eddy and mean heat flux divergence (or “layer thickness flux divergence”) drive the mean vertical velocity does not generalize to more complicated dynamical systems in which there is the possibility of altering the mean vertical density profile and/or in which the horizontal flow can be divergent. As a consequence of the constraints, there can be no basin net vorticity input to the lower layer via vortex stretching in the QG system.Because of the adiabatic QG constraints and the particular parametric regime in which the published adiabatic QG EGCM experiments exist, a very plausible explanation can be found for the existence of the deep cyclonic circulation of the model subtropical gyre. It is this cyclonic circulation that causes these deep flows to differ so dramatically from those of the more physically complex model systems. Because all the published adiabatic QG experiments that have non-trivial deep flows exhibit the counterrotating behavior, and because available ocean data do not support the existence of such a gyre in the North Atlantic, it seems important to thoroughly understand the reasons for the existence or absence of the deep cyclonic circulations. If they are an invitable feature of adiabatic QG systems, these models may need to be treated with caution as tools for understanding the mean ocean circulation.     

A Coupled General Circulation Model for the Tropical Pacific Ocean and Global Atmosphere

On the basis of Zeng’s theoretical design, a coupled general circulation model (CGCM) is developed with its characteristics different from other CGCMs such as the unified vertical coordinates and subtraction of the standard stratification for both atmosphere and ocean, available energy consideration, and so on. The oceanic component is a free surface tropical Pacific Ocean GCM between 30oN and 30oS with horizontal grid spacing of 1o in latitude and 2o in longitude, and with 14 vertical layers. The atmospheric component it a global GCM with low-resolution of 4o in latitude and 5o in longitude, and two layers or equal man in the vertical between the surface and 200 hPa. The atmospheric GCM includes comprehensive physical processes. The coupled model is subjected to seasonally-varying cycle. Several coupling experiments, ranging from straight forward coupling without flux correction to one with flux correction, and to so-called predictor-corrector monthly coupling (PCMC), are conducted to show the existence and final controlling of the climate drift in the coupled system. After removing the climate drift with the PCMC scheme, the coupled model is integrated for more than twenty years. The results show reasonable simulations of the annual mean and its seasonal cycle of the atmospheric and oceanic circulation. The model also produces the coherent interannual variations of the climate system, manifesting the observed El Ni?o / Southern Oscillation (ENSO).     

A numerical world ocean general circulation model

This paper describes a numerical model of the world ocean based on the fully primitive equations. A “Standard” ocean state is introduced into the equations of the model and the perturbed thermodynamic variables are used in the modle’s calculations. Both a free upper surface and a bottom topography are included in the model and a sigma coordinate is used to normalize the model’s vertical component. The model has four unevenly-spaced layers and 4 × 5 horizontal resolution based on C-grid system. The finite-difference scheme of the model is designed to conserve the gross available energy in order to avoid fictitious energy generation or decay.The model has been tested in response to the annual mean surface wind stress, sea level air pressure and sea level air temperature as a preliminary step to its further improvement and its coupling with a global atmospheric general circulation model. Some of results, including currents, temperature and sea surface elevation simulated by the model are presented.     

The deep water stratification of ocean general circulation models

Abstract

A central problem in climate and ocean modelling is the accurate simulation of the climatological state of the oceanic density field. A constant vertical diffusivity for heat and salt is frequently employed in ocean general circulation models (OGCMs) and it is usually assigned a value designed to optimize the depth of the pycnocline. One undesired consequence of this choice is a poor representation of the deep water, which is usually insufficiently stratified. In contrast to the uniform diffusivity of many models, some observational studies suggest that the vertical diffusivity is not constant but increases with depth, possibly in inverse proportion to the local buoyancy frequency. Numerical experiments with an OGCM are presented that demonstrate that allowing the vertical diffusivity to increase below the pycnocline substantially increases the stratification of the abyssal water mass of these models without significantly affecting the pycnocline depth, and hence may lead to a better representation of the vertical density structure.