The stability of the North Atlantic thermohaline circulation in a coupled ocean-atmosphere general circulation model

 The stability of the Atlantic thermohaline circulation against meltwater input is investigated in a coupled ocean-atmosphere general circulation model. The meltwater input to the Labrador Sea is increased linearly for 250 years to a maximum input of 0.625 Sv and then reduced again to 0 (both instantaneously and linearly decreasing over 250 years). The resulting freshening forces a shutdown of the formation of North Atlantic deepwater and a subsequent reversal of the thermohaline circulation of the Atlantic, filling the deep Atlantic with Antarctic bottom water. The change in the overturning pattern causes a drastic reduction of the Atlantic northward heat transport, resulting in a strong cooling with maximum amplitude over the northern North Atlantic and a southward shift of the sea-ice margin in the Atlantic. Due to the increased meridional temperature gradient, the intertropical convergence zone over the Atlantic is displaced southward and the westerlies in the Northern Hemisphere gain strength. We identify four main feedbacks affecting the stability of the thermohaline circulation: the change in the overturning circulation of the Atlantic leads to longer residence times of the surface water in high-northern latitudes, which allows them to accumulate more precipitation and runoff from the continents. As a consequence the stratification in the North Atlantic becomes more stable. This effect is further amplified by an enhanced northward atmospheric water vapour transport, which increases the freshwater input into the North Atlantic. The reduced northward oceanic heat transport leads to colder sea-surface temperatures and an intensification of the atmospheric cyclonic circulation over the Norwegian Sea. The associated Ekman transports cause increased upwelling and increased freshwater export with the East Greenland Current. Both the cooling and the wind-driven circulation changes largely compensate for the effects of the first two feedbacks. The wind-stress feedback destabilizes modes without deep water formation in the North Atlantic, but has been neglected in almost all studies so far. After the meltwater input stops, the North Atlantic deepwater formation resumed in all experiments and the meridional overturning returned within 200 years to a conveyor belt pattern. This happened although the formation of North Atlantic deep water was suppressed in one experiment for more than 300 years and the Atlantic overturning had settled into a circulation pattern with Antarctic bottom water as the only source of deep water. It is a clear indication that cooling and wind-stress feedback are more effective, at least in our model, than advection feedback and increased atmospheric water vapour transport. We conclude that the conveyor belt-type thermohaline circulation seems to be much more stable than hitherto assumed from experiments with simpler models. Received 31 January 1996/Accepted 22 August 1996     

一个压力坐标下的海洋环流模式

A new oceanic general circulation model in pressure coordinates is formulated. Since the bottom pressure changes with time, the vertical coordinate is actually a pressure-σ coordinate. The numerical solution of the model is based on an energy-conservation scheme of finite difference. The most important new feature of the model is that it is a truly compressible ocean model and it is free of the Boussinesq approxima tions. Thus, the new model is quite different from many existing models in the following ways: 1) the exact form of mass conservation, 2) the in-situ instantaneous pressure and the UNESCO equation of state to calculate density, 3) the in-situ density in the momentum equations, 4) finite difference schemes that conserve the total energy. Initial tests showed that the model code runs smoothly, and it is quite stable. The quasi-steady circulation patterns generated by the new model compare well with existing models, but the time evolution of the new model seems different from some existing models. Thus, the non-Boussinesq models may provide more accurate information for climate study and satellite observations.     

Multi-decadal thermohaline variability in an ocean–atmosphere general circulation model

A century scale integration of a near-global atmosphere–ocean model is used to study the multi-decadal variability of the thermohaline circulation (THC) in the Atlantic. The differences between the coupled and two supplementary ocean-only experiments suggest that a significant component of this variability is controlled by either a collective behavior of the ocean and the atmosphere, particularly in the form of air-sea heat exchange, or sub-monthly random noise present in the coupled system. Possible physical mechanisms giving rise to the mode of this THC variability are discussed. The SST anomaly associated with the THC variability resembles an interdecadal SST pattern extracted from observational data, as well as a pattern associated with the 50–60 year THC variability in the GFDL coupled model. In each case, a warming throughout the subpolar North Atlantic but concentrated along the Gulf Stream and its extension is indicated when the THC is strong. Concomitantly, surface air temperature has positive anomalies over the warmer ocean, with the strongest signal located downwind of the warmest SST anomalies and intruding into the western Eurasian Continent. In addition to the thermal response, there are also changes in the atmospheric flow pattern. More specifically, an anomalous northerly wind develops over the Labrador Sea when the THC is stronger than normal, suggesting a local primacy of the atmospheric forcing in the thermohaline perturbation structure.     

An eddy-permitting oceanic general circulation model and its preliminary evaluation

1. Introduction The current oceanic general circulation models(GCMs) used in climate studies, especially those serv-ing as component models in coupled GCMs, havean average resolution around 2? (IPCC, 2001). Theoceanic GCMs with relatively coarse resolutions canreproduce the major observed features of the large-scale circulations, but there are some insurmount-able di?culties. It was documented by Gates (1992)that the main defects of the coarse resolution oceanicGCMs include the repres…     

Sensitivity of the thermohaline circulation in coupled oceanic GCM — atmospheric EBM experiments

We analyze the sensitivity of the oceanic thermohaline circulation (THC) regarding perturbations in fresh water flux for a range of coupled oceanic general circulation — atmospheric energy balance models. The energy balance model (EBM) predicts surface air temperature and fresh water flux and contains the feedbacks due to meridional transports of sensible and latent heat. In the coupled system we examine a negative perturbation in run-off into the southern ocean and analyze the role of changed atmospheric heat transports and fresh water flux. With mixed boundary conditions (fixed air temperature and fixed surface fresh water fluxes) the response is characterized by a completely different oceanic heat transport than in the reference case. On the other hand, the surface heat flux remains roughly constant when the air temperature can adjust in a model where no anomalous atmospheric transports are allowed. This gives an artificially stable system with nearly unchanged oceanic heat transport. However, if meridional heat transports in the atmosphere are included, the sensitivity of the system lies between the two extreme cases. We find that changes in fresh water flux are unimportant for the THC in the coupled system.     

Stability of the interhemispheric thermohaline circulation in a coupled box model

The coupled atmosphere–ocean box model of the interhemispheric thermohaline circulation (THC) formulated by Scott et al. [Scott, J.R., Marotzke, J., Stone, P.H., 1999. Interhemispheric THC in a coupled box model. J. Phys. Oceanogr., 29, 351–365.] is solved analytically, by introducing the approximation that the time variations of salinity in the ocean are much slower than the time variations in the temperature. The analytic solution shows that there is an unstable limit cycle near the bifurcation where the flow becomes unstable, as suggested by Scott et al.'s numerical solutions. The solution also leads to an analytic expression for the conditions under which the instability discovered by Scott et al. sets in, which is more general than that found by Scott et al. In particular, it includes the effect of coupling the THC to the atmospheric meridional transports of heat and moisture. It shows that the stability of THC is much more sensitive to the representation of the atmospheric heat transport, i.e., to how it depends on the meridional temperature gradient, than it is in hemispheric models. In particular, it shows that interhemispheric ocean models that use mixed boundary conditions, or couple the ocean to a diffusive representation of the atmospheric heat transport, are less susceptible to this kind of instability than when the ocean is coupled to a representation of the atmospheric meridional heat transport which is more sensitive to the meridional temperature gradient, as is implied by observations and theory.     

Influence of the oceanic biology on the tropical Pacific climate in a coupled general circulation model

The influence of chlorophyll spatial patterns and variability on the tropical Pacific climate is investigated by using a fully coupled general circulation model (HadOPA) coupled to a state-of-the-art biogeochemical model (PISCES). The simulated chlorophyll concentrations can feedback onto the ocean by modifying the vertical distribution of radiant heating. This fully interactive biological-ocean-atmosphere experiment is compared to a reference experiment that uses a constant chlorophyll concentration (0.06 mg m⁻³). It is shown that introducing an interactive biology acts to warm the surface eastern equatorial Pacific by about 0.5°C. Two competing processes are involved in generating this warming: (a) a direct 1-D biological warming process in the top layers (0–30 m) resulting from strong chlorophyll concentrations in the upwelling region and enhanced by positive dynamical feedbacks (weaker trade winds, surface currents and upwelling) and (b) a 2-D meridional cooling process which brings cold off-equatorial anomalies from the subsurface into the equatorial mixed layer through the meridional cells. Sensitivity experiments show that the climatological horizontal structure of the chlorophyll field in the upper layers is crucial to maintain the eastern Pacific warming. Concerning the variability, introducing an interactive biology slightly reduces the strength of the seasonal cycle, with stronger SST warming and chlorophyll concentrations during the upwelling season. In addition, ENSO amplitude is slightly increased. Similar experiments performed with another coupled general circulation model (IPSL-CM4) exhibit the same behaviour as in HadOPA, hence showing the robustness of the results.     

Climate predictability experiments with a general circulation model

The atmospheric response to the evolution of the global sea surface temperatures from 1979 to 1992 is studied using the Max-Planck-Institut 19 level atmospheric general circulation model, ECHAM3 at T 42 resolution. Five separate 14-year integrations are performed and results are presented for each individual realization and for the ensemble-averaged response. The results are compared to a 30-year control integration using a climate monthly mean state of the sea surface temperatures and to analysis data. It is found that the ECHAM3 model, by and large, does reproduce the observed response pattern to El Nino and La Niña. During the El Nino events, the subtropical jet streams in both hemispheres are intensified and displaced equatorward, and there is a tendency towards weak upper easterlies over the equator. The Southern Oscillation is a very stable feature of the integrations and is accurately reproduced in all experiments. The inter-annual variability at middle- and high-latitudes, on the other hand, is strongly dominated by chaotic dynamics, and the tropical SST forcing only modulates the atmospheric circulation. The potential predictability of the model is investigated for six different regions. Signal to noise ratio is large in most parts of the tropical belt, of medium strength in the western hemisphere and generally small over the European area. The ENSO signal is most pronounced during the boreal spring. A particularly strong signal in the precipitation field in the extratropics during spring can be found over the southern United States. Western Canada is normally warmer during the warm ENSO phase, while northern Europe is warmer than normal during the ENSO cold phase. The reason is advection of warm air due to a more intense Pacific low than normal during the warm ENSO phase and a more intense Icelandic low than normal during the cold ENSO phase, respectively.     

Century-scale variability in a randomly forced,two-dimensional thermohaline ocean circulation model

The response of a two-dimensional thermohaline ocean circulation model to a random freshwater flux superimposed on the usual mixed boundary conditions for temperature and salinity is considered. It is shown that for a wide range of vertical and horizontal diffusivities and a box geometry that approximates the Atlantic Ocean, 200–300 yr period oscillations exist in the basic-state, interhemispheric meridional overturning circulation with deep convection in the north. These fluctuations can also be described in terms of propagating salinity anomalies which travel in the direction of the thermohaline flow. For large horizontal (K _h = 15 × 103 m²/s) and small vertical (K _v = 0.5 × 10^–4 m²/s) diffusivities, the random forcing also excites deca-millennial oscillations in the basic structure of the thermohaline circulation. In this case, the meridional circulation pattern slowly oscillates between three different stages: a large positive cell, with deep convection in the North Atlantic and upwelling in the south; a symmetric two-cell circulation, with deep convection in both polar regions and upwelling near the equator; and a large negative cell, with deep convection in the South Atlantic and upwelling in the north. Each state can persist for 0 (10 kyr).     

基于海气耦合环流模式的ENSO预测

Predictions of ENSO are described by using a coupled atmosphere-ocean general circulation model. The initial conditions are created by forcing the coupled system using SST anomalies in the tropical Pacific at the background of the coupled model climatology. A series of 24-month hindcasts for the period from November 1981 to December 1997 are carried out to validate the performance of the coupled system. Correlations of SST anomalies in the Nino3 region exceed 0.54 up to 15 months in advance and the rms errors are less than 0.9℃. The system is more skillful in predicting SST anomalies in the 1980s and less in the 1990s. The model skills are also seasonal-dependent, which are lower for the predictions starting from late autumn to winter and higher for those from spring to autumn in a year-time forecast length. The prediction, beginning from March, persists 8 months long with the correlation skill exceeding 0.6, which is important in predictions of summer rainfall in China. The predictions are succesful in many aspects for the 1997-2000 ENSO events.     

Simulation of climate change over Europe using a global variable resolution general circulation model

 This study presents results from a downscaling simulation of the impact of a doubling of CO₂ concentration. A multidecadal coupled simulation of a 1% per year increase of CO₂ concentration with the Hadley Centre ocean-atmosphere model provides its sea-surface temperatures and deep soil climatological temperatures as a boundary condition to two 10-year integrations with a version of the ARPEGE-IFS atmosphere model. This global spectral model has a horizontal resolution varying between 60 km in the Mediterranean Sea and 700 km in the southern Pacific. The global impact as well as the regional impact over Europe in this time slice are examined and compared with results from other studies. Over Europe, our main focus, the model impact consists of a warming of about 2 ^°C, relatively uniform and with little seasonal dependence. There are precipitation increases of about 10% over the northern part in winter and spring, and 30% over the southern part in winter only. Precipitation decreases by 20% in the southern part in autumn. The day-to-day variability of the precipitation increases, except over the southern area in summer. No strong impact is found on the soil moisture. Budgets of physical fluxes are examined at the top of the atmosphere and at the land-atmosphere interface. Received: 26 February 1997/Accepted: 21 October 1997     

Fluctuation dissipation in a general circulation model

This paper considers the climate response to step function changes in the solar constant in two versions of a general circulation model with simplified geography. The NCAR CCM0 model is applied to an all-land planet with no topography (Terra Blanda). In one version there is moisture in the air (as well as self-generated clouds) as evaporated from an ideal surface at a fixed 80% of saturation. In the other version there is no moisture in the atmosphere. We examine the decay of natural anomalies in the large-scale temperature field in each model and compare the time dependence of the ensemble average with the average temporal behavior of the response to step function changes in the solar constant. The fluctuation-dissipation theorem of statistical mechanics makes specific predictions about the relationship between the two curves. We conduct the experiments for both versions of the model since the sensitivity is quite different for each. The theorem is found to hold reasonably well in each case.     

Sensitivity of the thermohaline circulation and climate to ocean exchanges in a simple coupled model

 A new simple, coupled climate model is presented and used to investigate the sensitivity of the thermohaline circulation and climate to ocean vertical and horizontal exchange. As formulated, the model highlights the role of thin, ocean surface layers in the communication between the atmosphere and the subsurface ocean. Model vertical exchange is considered to be an analogue to small-scale, diapycnal mixing and convection (when present) in the ocean. Model horizontal exchange is considered to be an analogue to the effects of the wind-driven circulation. For small vertical exchange in the ocean, the model exhibits only one steady-state solution: a relatively cold, mid-high-latitude climate associated with a weak, salinity-driven circulation (“off ” mode). For large vertical and horizontal exchange in the ocean, the model also exhibits only one steady-state solution: a relatively warm, mid-high-latitude climate associated with a strong, thermally-driven circulation (“on” mode). For sufficiently weak horizontal exchange but large enough vertical exchange, both modes are possible stable, steady-state solutions. When model parameters are calibrated to fit tracer distributions of the modern ocean-atmosphere system, only the “on” mode is possible in this standard case. This suggests that the wind-driven circulation in consort with diapycnal mixing suppresses the “off ” mode in the modern ocean-atmosphere system. Since both diapycnal mixing and the wind-driven circulation would be expected to increase in a cold climate with greater meridional temperature gradients and enhanced winds, vertical and horizontal exchange in the ocean are probably associated with strong negative feedbacks which tend to stabilize climate. These results point to the need to resolve ocean wind-driven circulation and to greatly improve the treatment of ocean diapycnal mixing in more complete models of the climate system. Received: 16 November 1999 / Accepted: 19 June 2000     

Multi-year predictability in a coupled general circulation model

Multi-year to decadal variability in a 100-year integration of a BMRC coupled atmosphere-ocean general circulation model (CGCM) is examined. The fractional contribution made by the decadal component generally increases with depth and latitude away from surface waters in the equatorial Indo-Pacific Ocean. The relative importance of decadal variability is enhanced in off-equatorial “wings” in the subtropical eastern Pacific. The model and observations exhibit “ENSO-like” decadal patterns. Analytic results are derived, which show that the patterns can, in theory, occur in the absence of any predictability beyond ENSO time-scales. In practice, however, modification to this stochastic view is needed to account for robust differences between ENSO-like decadal patterns and their interannual counterparts. An analysis of variability in the CGCM, a wind-forced shallow water model, and a simple mixed layer model together with existing and new theoretical results are used to improve upon this stochastic paradigm and to provide a new theory for the origin of decadal ENSO-like patterns like the Interdecadal Pacific Oscillation and Pacific Decadal Oscillation. In this theory, ENSO-driven wind-stress variability forces internal equatorially-trapped Kelvin waves that propagate towards the eastern boundary. Kelvin waves can excite reflected internal westward propagating equatorially-trapped Rossby waves (RWs) and coastally-trapped waves (CTWs). CTWs have no impact on the off-equatorial sub-surface ocean outside the coastal wave guide, whereas the RWs do. If the frequency of the incident wave is too high, then only CTWs are excited. At lower frequencies, both CTWs and RWs can be excited. The lower the frequency, the greater the fraction of energy transmitted to RWs. This lowers the characteristic frequency (reddens the spectrum) of variability off the equator relative to its equatorial counterpart. At low frequencies, dissipation acts as an additional low pass filter that becomes more effective, as latitude increases. At the same time, ENSO-driven off-equatorial surface heating anomalies drive mixed layer temperature responses in both hemispheres. Both the eastern boundary interactions and the accumulation of surface heat fluxes by the surface mixed layer act to low pass filter the ENSO-forcing. The resulting off-equatorial variability is therefore more coherent with low pass filtered (decadal) ENSO indices [e.g. NINO3 sea-surface temperature (SST)] than with unfiltered ENSO indices. Consequently large correlations between variability and NINO3 extend further poleward on decadal time-scales than they do on interannual time-scales. This explains why decadal ENSO-like patterns have a broader meridional structure than their interannual counterparts. This difference in appearance can occur even if ENSO indices do not have any predictability beyond interannual time-scales. The wings around 15–20°S, and sub-surface variability at many other locations are predictable on interannual and multi-year time-scales. This includes westward propagating internal RWs within about 25° of the equator. The slowest of these take up to 4 years to reach the western boundary. This sub-surface predictability has significant oceanographic interest. However, it is linked to only low levels of SST variability. Consequently, extrapolation of delayed action oscillator theory to decadal time-scales might not be justified.     

Feedbacks affecting the response of the thermohaline circulation to increasing CO2: a study with a model of intermediate complexity

A three-dimensional ocean model with an idealized geometry and coarse resolution coupled to a two-dimensional (zonally averaged) statistical-dynamical atmospheric model is used to simulate the response of the thermohaline circulation to increasing CO ₂ concentration in the atmosphere. The relative roles of different factors in the slowing down and recovery of the thermohaline circulation were studied by performing simulations with ocean only and partially coupled models. The computational efficiency of the model allows an extensive and thorough study of the causes of changes in the strength of the thermohaline circulation, through a large number of extended runs. The evolution of the atmosphere-to-ocean surface heat fluxes is shown to be the dominant factor in causing the weakening of the circulation in response to an increasing external forcing as well as in controlling the subsequent recovery. The feedback between heat flux and the sea surface temperature is necessary for the ocean circulation to recover. The rate of the recovery, however, is not sensitive to the magnitude of the feedback, and changes in the atmosphere, while contributing to the recovery, play a secondary role. In the case of very strong feedback, substantial changes in the SST structure are shown not to be a necessary condition for the recovery of the circulation. Subsurface changes in the density structure accompany recovery despite nearly fixed SST in one of the uncoupled experiments. Changes in the zonal distribution of heat fluxes serve as a positive feedback for both decrease and recovery of the meridional overturning, and are as important as changes in the zonal-mean values of heat fluxes. The secondary role of the moisture fluxes is explained by a smaller magnitude of their contribution to the surface buoyancy flux. Imposing amplified changes in the moisture fluxes leads to a significant slow down of the circulation, accompanied, however, by changes in the heat flux. The changed heat flux, in its turn, makes a significant contribution to the future slow down. This feedback complicates the evaluation of the relative roles of the different fluxes.     

A low latitude general circulation model

Summary The primitive equations of motion are solved to find peak latitudes of families of trajectories of air moving from low latitudes towards the pole from a state of rest under the action of a constant meridional pressure gradient. The general case is discussed first and then calculations are applied to the 20° West meridian in January and July for the Northern Hemisphere.The resulting mass transposition indicates that a belt of convergence should occur at about 32° N in January; this agrees well with the latitude of the axis of the mean sub-tropical high pressure belt at that time of year.The nature of the climatological model in low latitudes suggested by the calculation is then discussed.

---

Zusammenfassung Die vereinfachten Bewegungsgleichungen werden gelöst, um Scheitelbreiten für Trajektorienscharen der Luftbewegungen zu finden, die ausgehend von dem Ruhezustand unter Einwirkung eines konstanten meridionalen Druckgradienten polwärts gerichtet sind. Der allgemeine Fall wird diskutiert und es werden Berechnungen für den 20. westlichen Längengrad der Nordhemisphäre für Januar und Juli durchgeführt.Der resultierende Massentransport zeigt, daß im Januar in 32° N ein Konvergenzgürtel auftreten sollte. Dies stimmt gut überein mit der geographischen Breite der Achse des mittleren subtropischen Hochdruckgürtels zu dieser Jahreszeit.Schließlich wird die Form des klimatologischen Modells für niedere Breiten, das durch diese Berechnungen nahegelegt wird, diskutiert.

---

Résumé Résolution des équations simpflifiées du mouvement pour déterminer l'origine des faisceaux de trajectoires des courants aériens qui, partant du repos, sont dirigés vers le pôle sous l'effet d'un gradient constant de pression méridien. Discussion du cas général; calcul pour le 20^e méridien Ouest de l'hémisphère Nord en janvier et en juillet.Le transport de masse résultant montre qu'il doit y avoir en janvier, a 32° de latitude Nord, une ceinture de convergence ce qui concorde bien avec la latitude de l'axe de la zone moyenne des hautes pressions subtropicales en cette saison.Forme du modèle climatologique aux basses latitudes suggéré par ces calculs.

---

---

With 1 Figure     

Effect of seasonal forcing on global circulation in a world ocean general circulation model

 Effects of the seasonal variation in thermohaline and wind forcing on the abyssal circulation are investigated by using an ocean general circulation model. To isolate effects of the seasonality in the thermohaline forcing from those in the wind forcing, we carry out three experiments with (1) annual-mean wind forcing and perpetual-winter thermohaline forcing, (2) annual-mean wind forcing and seasonal thermohaline forcing, and (3) seasonal wind forcing and seasonal thermohaline forcing. The deep water under the seasonal thermohaline forcing becomes warmer than under the perpetual-winter thermohaline forcing. Although the perpetual-winter thermohaline forcing is widely used and believed to reproduce the deep water better than the annual-mean forcing, the difference between the results of the perpetual-winter and the seasonal thermohaline forcing is significant. The seasonal variation of the Ekman convergence and divergence produces meridional overturning cells extending to the bottom because the period of seasonal cycle is shorter than the adjustment timescale by baroclinic Rossby waves. The heat transport owing to those Ekman flows and temperature anomalies makes the upper water (0–200 m) colder at low to mid-latitudes (40S–40N) and warmer at high latitudes. Also the deep water becomes warmer owing to the warming of the northern North Atlantic, the main source region of North Atlantic Deep Water. The model is also synchronously (i.e., without acceleration) integrated with seasonal forcing for 5400 y. A past study suggested that under seasonal forcing, a sufficient equilibrium state can be achieved after only decades of synchronous integration following more than 10 000 y of accelerated integration. Here, the result so obtained is compared with that of the 5400-y synchronous integration. The difference in the global average temperature is as small as 0.12 °C, and most of the difference is confined to the Southern Ocean. Received: 1 May 1998 / Accepted: 5 January 1999     

A coupled ocean-atmosphere climate model: temperature versus salinity effects on the thermohaline circulation

A simple nonlinear three-box ocean model of the North Atlantic Ocean including the rudiments of eddy mixing, vertical stratification and thermohaline circulation is first presented. It is subject to uniform latitudinal differential heating, q, and net evaporation m _e , and includes a linear equation of state. Two quite different limiting steady-state solutions exist. The first has a warm saline surface water and a cold, low-salinity deep ocean; deep water is primarily formed in higher latitudes by the prevalence of differential heating. A second limiting solution consists of a warm saline deep ocean underlying a cool, low-salinity surface ocean; deep water is formed primarily in lower latitudes as a consequence of large differential evaporation. A coupled ocean-atmosphere model, in which the oceanic surface heat fluxes are determined internally but with differential evaporation at the ocean surface m _e remaining an external parameter, is next presented. The atmosphere component is a simple energy balance model that emphasizes the vertical fluxes of radiative, sensible and latent heat fluxes but does not include temperature-albedo feedback. Model response depends on the external parameters m _e and , controlling the magnitude of the thermohaline-driven circulation, and on the magnitudes of the eddy mixing coefficients and the solar constant. For small m _e , a steady-state solution corresponding to a cold fresh deep ocean is found, qualitatively similar to the modern ocean. For large m _e , a steady-state solution with a warm saline deep ocean occurs; this solution resembles conceptual models that have been proposed for the warm saline Cretaceous ocean. There exists an intermediate region of values of m _e for which the solutions are more complex. On the lower end of this region, both the cold fresh deep-ocean and warm saline deep-ocean circulations coexist as stable equilibria. On the upper end, the cold-deep ocean becomes unstable, manifesting oscillations with growing amplitude, and ultimately reaches the warm saline deep-ocean solution. In the neighborhood of a cusp on the , m _e plane, that is, for relatively small , more complex behaviour occurs, which has not yet been fully analyzed. The model response in the region of complexity is not sensitive to changes in the solar constant but is sensitive to the eddy mixing coefficients.