Andrew J. Weaver - Abstracts

On the circulation of the North Pacific Ocean: Climatology, seasonal cycle and interpentadal variability

Paul G. Myers and Andrew J. Weaver

A finite element diagnostic model is used to study the circulation of the North Pacific ocean. With the inclusion of the JEBAR term, the model produces a very realistic picture of the circulation. All major currents are reproduced with the calculated transports agreeing very well with the observations. The three dimensional velocity structure is diagnosed from the thermal wind equation, assuming a reference velocity at the bottom. This bottom reference velocity is calculated from the Ekman, thermohaline and total transport (from the finite element model) velocities. The diagnosed velocity fields are compared with a number of observational sections.

The effect of using different wind stress climatologies is also examined. Due to the dominance of the JEBAR term in the solution, the resulting circulations are all similar. Analysis of the seasonal cycle in the model supports the suggestion of Sakamoto and Yamagata that JEBAR rectification can explain the decreased amplitude of the seasonal cycle and the out of phase relationship between observations and the predictions of flat-bottomed Sverdrup theory. Finally, density fields from 1955-1959 and 1970-1974 are used to examine aspects of interpentadal variability in the North Pacific Ocean.

Validation of sub-grid scale mixing schemes using CFCs in a global ocean model

Daniel Y. Robitaille and Andrew J. Weaver

Three sub-grid scale mixing parameterizations (lateral/vertical; isopycnal/diapycnal; Gent and McWilliams, 1990) are used in a global ocean model in an attempt to determine which yields the best ocean climate. Observations and model Freon 11 distributions, in both the North and South Atlantic, are used in the model validation. While the isopycnal/diapycnal mixing scheme does improve the deep ocean potential temperature and salinity distributions, when compared to results from the traditional lateral/vertical mixing scheme, the Freon 11 distribution is significantly worse due to too much mixing in the southern ocean. The Gent and McWilliams (1990) parameterization, on the other hand, significantly improves the deep ocean potential temperature, salinity and Freon 11 distributions when compared to both of the other schemes. The main improvement comes from a reduction of Freon uptake in the southern ocean where the "bolus" transport cancels the mean advection of tracers and hence causes the Deacon Cell to disappear. These results suggest that the asymmetric response found in CO₂ increase experiments, whereby the climate over the southern ocean does not warm as much as in the northern hemisphere, may be an artifact of the particular mixing schemes used.

On the variability of temperature, salinity and the circulation of the Labrador Sea

Thierry H. Reynaud, Andrew J. Weaver and Richard J. Greatbatch

The Labrador Sea is a key region for the formation of deep water in the North Atlantic and long time-scale variability of its water masses are thought to play an important role in climate. This study attempts to quantify the decadal-interannual variability of the water temperature and salinity in the Labrador Sea as well as their impact on the circulation. The original MEDS- NODC temperature and salinity data are first mapped onto a 1/2 degree x 1/2 degree grid for the periods 1950-1964 and 1965-1981. The results show that fresher and colder conditions prevailed during the latter period at the surface, possibly associated with the Great Salinity Anomaly, while saltier and warmer waters are found at deep and intermediate depths. The Mellor et al (1982) method is then applied to diagnose the volume transport variability between these two periods. The results show a weakening over the domain of 4 to 6 Sv in the 1965-1981 period. This weakening arises from changes in the density field that reduce the local JEBAR contribution to the total volume transport. Neither the variability of the local wind forcing nor the reference boundary values used in the model could account for the weakening found in the volume transport. The objective analysis was then repeated on a higher resolution grid for two selected sub-regions located near Hamilton Bank and Flemish Cap.

Can regulation of fresh water runoff in Hudson Bay affect the climate of the North Atlantic?

P.H. Leblond, J.R. Lazier and A.J. Weaver

In response to claims that freshwater regulation arising from hydroelectric developments in the basins of Hudson Bay and James Bay might quench deep-water formation in the North Atlantic, and thus northward heat transport by ocean currents, which in turn might lead to a cooling of western Europe, we have examined the oceanographic consequences of changes in the timing of the freshwater runoff regime in that area. We critically review the information available on the sequence of phenomena linking anthropogenic changes in runoff to a possible impact on the North Atlantic thermohaline circulation: the spreading of estuarine plumes under ice; the effect of lowered salinity on the rate of ice formation; regional effects on the scale of Hudson Bay; the export of fresh water to the Labrador Sea; its impact on deep convection in that area; the relative importance of such changes to the North Atlantic circulation. At each step of this chain of events, we have attempted to compare anthropogenic effects with other factors and to place them within the perspective of natural variability. Our conclusion is straightforward and unambiguous: a close examination of the oceanic response does not support the contention that freshwater runoff regulation in the basins of Hudson and James Bays could have a significant, perhaps even a detectable effect on the climate of the North Atlantic.

Summer mean circulation in the western North Atlantic

Thierry H. Reynaud, Andrew J. Weaver and Richard J. Greatbatch

The Mellor et al. (1982) method is applied to a new high-resolution analysis of the temperature and salinity fields in order to determine the summer transport and circulation of the northwestern Atlantic Ocean. This high-resolution analysis is carried out using an objective analysis scheme which is a modification of that used by Levitus (1982). In view of the strong topographic control exhibited by the circulation features in the area (i.e., the shelf break Labrador and Greenland currents), the scheme preferentially searches for data along, rather than across, isobaths, the horizontal resolution is 1/3 by 1/3 degree, with 37 vertical levels. The data were obtained from the Marine Environment Data Service archived data and was supplemented by a subset of the National Oceanographic Data Center data from J. Reid and by additional data for the 1980s from Fukumori and Wunsch (1991). Summer mean transports of 49 and 46 Sv are found in the Labrador Sea and the Irminger Sea, respectively. Most of the transport through the region is determined by the transport through the eastern boundary, emphasizing the importance of the eastern Atlantic for determining the circulation in the west. The local wind stress forcing plays a relatively unimportant role in driving the transport in the northwestern Atlantic Ocean. The current structure is obtained by combining the results from the Mellor et al. (1982) method together with a level of no motion (at the bottom) calculation. Using these two methods, the bottom currents are evaluated, and hence the current structure for the whole domain is determined. The results show that the bottom currents follow the planetary potential vorticity (f/H) lines closely. The strongest currents are found along the shelf breaks, offshore from the coasts of Greenland and Labrador. The results also indicate the presence of cross-shelf flow on the western side of the Labrador Sea. The cross-shelf transport (~ 5Sv) of fresh waters may well be important in modifying the salinity characteristics, and hence convective properties, of the central Labrador Sea.

Sea surface temperature - evaporation feedback and the ocean's thermohaline circulation

Tertia M.C. Hughes and Andrew J. Weaver

A simple surface boundary condition which includes the dependence of evaporation on sea surface temperature is developed for use in uncoupled ocean models. A positive feedback with the thermohaline circulation is found, with accelerated overturning warming the high latitudes by advection, enhancing evaporation and hence the sea surface salinity, which then feeds back on to the overturning. Although nonlinear interactions with convection do occur in some instances, the time-dependent component of the evaporation is in general small, which would tend to support the use of fixed freshwater fluxes as a good first approximation to air-sea interaction. In agreement with this, two examples with internal variability of the thermohaline circulation on decadal and millennial timescales respectively are not fundamentally altered under the new feedback compared to control runs under mixed boundary conditions, although both the period and the duration of the variability are shortened in some cases.

On the incompatibility of ocean and atmosphere models and the need for flux adjustments

Andrew J. Weaver and Tertia M.C. Hughes

The surface heat and freshwater fluxes from equilibrium ocean (OGCM) and atmospheric (AGCM) general circulation model climates are examined in order to determine the minimum flux correction required to prevent climate drift upon coupling. It is shown that a dramatic climate drift of the coupled system is inevitable unless climatological ocean heat and salt transports are used as constraints for tuning the AGCM present-day climatology. It is further shown that the magnitude of the mismatch between OGCM and AGCM fluxes is not as important for climate drift as the difference in OGCM and implied AGCM heat and freshwater transports. Hence a Minimum Flux Correction is proposed, which is zonally-uniform in each basin and of small magnitude compared to present flux-corrections. This minimum flux correction acts only to correct the AGCM implied oceanic transports of heat and freshwater. A slight extension is also proposed to overcome the minor drift in the surface waters when the minimum flux correction is used. Finally, it is shown that the current methods used to determine flux corrections are all essentially equivalent leading to correction fields which are significantly larger than both AGCM and climatological fields over large regions.

Climate stability as deduced from an idealized coupled atmosphere-ocean model

Benyang Tang and Andrew J. Weaver

The stability of an idealized climate system is investigated using a simple coupled atmosphere-ocean box model. Motivated by the results from general circulation models, the main physical constraint imposed on the system is that the net radiation at the top of the atmosphere is fixed. The specification of an invariant equatorial atmospheric temperature, consistent with paleoclimatic data, allows the hydrological cycle to be internally determined from the poleward heat transport budget, resulting in a model that has a plausible representation of the hydrological cycle-thermohaline circulation interaction. The model suggests that the stability and variability of the climate system depends fundamentally on the mean climatic state (total heat content of the system). When the total heat content of the climate system is low, a stable present-day equilibrium exists with high-latitude sinking. Conversely, when the total heat content is high, a stable equatorial sinking equilibrium exists. For a range of intermediate values of the total heat content, internal climatic oscillations can occur through a hydrological cycle-thermohaline circulation feedback process. Experiments conducted with the model reveal that under a 100-year 2 x CO₂ warming, the thermohaline circulation first collapses but then recovers. Under a 100-year 4 x CO₂ warming, the thermohaline circulation collapses and remains collapsed. Recent paleoclimatic data suggest that the climate system may behave very differently for a warmer climate. Our results suggest that this may be attributed to the enhanced poleward freshwater transport, which causes increased instability of the present-day thermohaline circulation.

Low-frequency Internal Oceanic Variability Under Seasonal Forcing

Paul G. Myers and Andrew J. Weaver

A series of numerical experiments are conducted using the Bryan-Cox ocean general circulation model to investigate the potential existence of low-frequency variability of the thermohaline circulation under seasonal forcing. Experiments are performed with different combinations of a seasonal cycle being present or not on the restoring temperature, the surface freshwater flux fields (mixed boundary conditions), and the surface wind forcing. Despite the presence of forcing on the dominant seasonal time scale, it is found that the system may oscillate at the decadal period or longer. The decadal variability is excited by changes in the net surface density flux which are due to the advection of temperature and salinity anomalies in the model domain. The magnitude of the seasonal cycle also plays an important role in determing the time scale of the variability. Violent overturning events may occur on the century time scale under seasonal forcing. The magnitudes of the flushes are reduced compared with those found in similar experiments without the presence of a seasonal cycle.

Interdecadal Variability in an Idealized Model of the North Atlantic

Andrew J. Weaver, Stella M. Aura and Paul G. Myers

A coarse resolution model is developed to study the thermohaline circulation of the North Atlantic. This model is driven by the annual mean Hellerman and Rosenstein wind stress field, Levitus sea surface restoring temperatures and Schmitt, Bogden and Dorman freshwater flux fields (mixed boundary conditions) together with various parameterizations of Arctic freshwater export into the North Atlantic. The model simulations indicate the existence of self-sustained, internal variability of the thermohaline circulation with a period of about 20 years. Associated with the variability is a large variation in the deep water formation rate in the Labrador Sea, and hence the poleward heat transport in the North Atlantic. It is shown that the variability is insensitive to the freshwater flux and wind forcing used and that the timescale for this thermally-driven convective/advective oscillation is set by the cooling time of the Labrador Sea. The variability is robust to various parameterizations of Arctic freshwater export but may be suppressed if there is a strong freshwater flux through the Canadian Archipelago (or equivalently, large precipitation) into the Labrador Sea. The importance of topography, although poorly resolved in this coarse resolution study, is addressed and the results are compared with a coupled atmosphere-ocean simulation and observations taken over the North Atlantic.

The ocean as a source for rapid interglacial climate fluctuations

Andrew J. Weaver and Tertia M.C. Hughes

Recent remarkable findings in Greenland ice core data have suggested that the climate of the last (Eemian) interglacial was not as stable as that of our present (Holocene) interglacial. Rapid transitions between warm and cold periods were found to occur over a period of several decades. The North Atlantic climate during the Eemian period was further shown to be characterized by three states: one significantly warmer than today; one similar to today and one significantly colder than today. While there is some concern as to the reliability of the ice core data corresponding to the latter half of the Eemian we suggest that the "conveyor" for North Atlantic deep water formation has three distinct modes of operation, consistent with the existence of the three Eemian climatic states. This suggestion is based on the results from an idealized global ocean model driven by quasi-realistic winds, freshwater fluxes, and sea-surface temperatures. We further show that rapid transitions between the modes can be excited through the addition of a simple random forcing to the mean freshwater flux forcing field. The model results suggest that a source for the observed, but controversial, Eemian climate variability may well lie in the dynamics of the ocean's thermohaline circulation which responds to an enhanced hydrological cycle associated with the warmer mean Eemian climate.

Global climate change: Lessons from the past

Andrew J. Weaver and Chris Green

Over the past few years increasing public, political and scientific concern has been directed towards potential climate change associated with increasing greenhouse gases. The most recent and sophisticated climate forecasts have suggested that global warming will occur at a rate of 3.5 deg C per century associated with a 1%/year increase in atmospheric CO₂ (close to the IPCC Business as Usual Scenario for atmospheric greenhouse gas emissions). In the high latitudes of the northern hemisphere, the climate forecasts for a doubling of atmospheric CO₂ further suggest an amplification of the warming (about 8-9 degrees C versus about 3 degrees C for low latitude regions) due to the reduction of sea ice cover and the accompanying decrease in surface albedo (amount of incoming solar radiation reflected back to space). In the region around Antarctica, very little change (or even slight cooling) is predicted over the next few centuries, due to the efficient absorption of heat by the ocean there. Predictions of global climate change pose a public policy problem: what, if anything, should be done to head off the threat created by emissions of greenhouse gases? In answering this question, the most widely accepted approach (at least among economists) is to weigh the benefits of a given action against its costs. The benefits of actions taken to head off (prevent or mitigate) global climate change are the damages thereby avoided. These are the damages that would occur in the absence of policy actions taken to limit the build-up of greenhouse gases in the atmosphere. Since some climate changes may already be in the pipeline, as a result of past emissions of greenhouse gases, not all damages may be avoided. Economists are able to provide a quantitative measure of benefits by focusing on the economic damages avoided. Typically this requires an in depth analysis of the impact that climate has on the various sectors of economic activity. Of necessity such measures ignore the existence "value" some human beings place on the earth's present climate, independent of any economic damages that might accompany a changed (presumed warmer) global climate. The costs of policy actions are also measured in economic terms. These costs are quantified as the reductions in output that are estimated to occur if greenhouse gas emissions are reduced below the level that would occur in the absence of policy action. Since a large fraction of greenhouse gas emissions, especially CO₂, are released as a result of energy production and use, cutbacks in emissions imply cutbacks in fossil-fuel energy use. With 85% of the world's energy provided by fossil fuels, and with non-fossil fuel energy substitutes (other than nuclear energy, which has its own environmental problems) currently incapable of meeting baseload energy requirements, reduced CO₂ emissions imply reduced energy use. In turn, reduced energy use is likely to result in a reduction in economic output (or gross domestic product [GDP]). A reduction in output or output growth is inevitable if the reduction in energy use exceeds the rate at which technological improvements make possible the long term decline in fossil fuel energy input per unit of economic output. Economists have estimated that simply stabilizing global emissions at 1990 levels (which would slow but certainly not prevent climate change) will reduce global GDP by approximately 2% per annum by 2040 and 4% per annum, plus or minus 1%, by 2100. The comparison of the benefits and costs of actions in preventing/mitigating climate change requires a scientifically based climate change scenario and numerous economic assumptions. The most important of these assumptions are those underlying projections of the long term rate of output growth and those relating to (a) the rate of induced technological changes that affect energy use per unit of output and (b) the development of non-fossil fuel energy alternatives. The most widely employed scenario is based upon IPCC modelling, which projects that "business as usual" (no policy action taken to control emissions) would produce a 3 to 6 degrees C global warming in 2100 (relative to 1900), with a best guesstimate being 4 degrees C. This scenario is based on atmospheric general circulation models (GCMs) that project a relatively smooth path of rising global average temperature. As a result, estimates of economic damages are implicitly based on the assumption of relatively smooth, gradual changes in climate. The possibility of climatic "regime changes", such as those which we shall discuss below, are all but ignored. Suffice it to say that the economists' benefit-cost calculus, already hard pressed to deal with the long term and uncertain nature of climate change has not yet been, and perhaps cannot be, fruitfully applied to the far more catastrophic possibility posed by climate regime changes.

Decadal-millennial internal oceanic variability in coarse resolution ocean general circulation models

Andrew J. Weaver

The ocean's thermohaline circulation, driven by fluxes of freshwater and heat through the ocean's surface, is an important mechanism for the transport of heat from low to high latitudes. Changes in the intensity of the thermohaline circulation, and hence its poleward heat transport, would have a significant effect on global climate. Here, the results of a number of experiments conducted using a coarse resolution ocean general circulation model (OGCM) in idealized basins are reviewed to illustrate the relative importance of freshwater flux, thermal, and wind forcing in exciting decadal-millennial variability of the thermohaline circulation. A brief discussion of the shortcomings of these models and some suggestions for future research are also presented. Recent experiments are also discussed for a coarse OGCM simulation of the North Atlantic. This model is driven by annual mean Hellerman and Rosenstein winds, Levitus sea surface restoring temperature and Schmitt, Bogden and Dorman freshwater flux fields (mixed boundary conditions). Various parameterizations of Arctic freshwater export into the North Atlantic are included to examine the internal variability properties of the North Atlantic thermohaline circulation. It is found that self-sustained, internal variability with about a 20 year period exists under steady forcing provided there is a sufficiently weak Arctic freshwater flux through the Canadian Archipelago into the Labrador Sea. The variability is robust to various parameterizations Arctic freshwater export. Large variations in the deep water formation rate, especially in the Labrador Sea, and hence the poleward transport of heat occur over an oscillation. The importance of topography is also addressed and a detailed physical discussion of the mechanism and timescale for the oscillation is presented.

Freshwater Flux Forcing of Decadal/Interdecadal Oceanic Variability

Andrew J. Weaver, Ed S. Sarachik and Jochem Marotzke

The ocean's thermohaline circulation (THC), driven by fluxes of freshwater and heat through the ocean's surface, is an important mechanism for the transport of heat from low to high latitudes. Changes in the intensity of the THC, and hence its poleward heat transport, would have a significant effect on global climate. In recent years numerous observations of decadal/interdecadal variability in and around the North Atlantic have been reported. The timescale associated with such variability suggests that its source lies within the ocean. Here we present the results of three numerical experiments which show the importance of freshwater flux forcing (precipitation - evaporation; denoted P-E) in exciting decadal/ interdecadal variability of the THC. If a sufficiently strong local minimum exists in this P-E forcing field (as observed over the Greenland Sea), self-sustained oscillations of the THC may be excited. We propose that such variability may be important in interpreting observations of decadal/interdecadal variability in the air-sea-ice climate system.

Stability and Variability of the Thermohaline Circulation

Andrew J. Weaver, Jochem Marotzke, Patrick F. Cummins and Ed S. Sarachik

The stability and internal variability of the ocean's thermohaline circulation is investigated using a coarse resolution general circulation model of an idealized ocean basin, in one hemisphere. The model circulation is driven, in addition to wind forcing, by restoring the surface temperature to prescribed values, and by specifying freshwater fluxes in the surface salinity budget (mixed boundary conditions). All forcing functions are constant in time. The surface freshwater forcing is the dominant factor in determining the model's stability and internal variability. Increasing the relative importance of freshwater flux versus thermal forcing, we find, in turn, one stable steady state of the model, two stable ones, one stable and one unstable equilibrium, or no stable steady states at all. If the freshwater forcing is sufficiently strong, there exist self-sustained oscillations in the deep water formation rate, which last thousands of years. One type of oscillation occurs on the timescale of decades and is associated with the advection of high-latitude salinity anomalies. The other type has a diffusive timescale of centuries or longer, and marks periods of complete absence of deep water formation followed by violent overturning events (flushes). When a stochastic component is added to the steady freshwater flux forcing, internal decadal variability persists if the background steady freshwater flux is sufficiently strong. Periodic flushes also exist under stochastic forcing; with increasing magnitude of the stochastic term the frequency of the flush events increases while their intensity decreases.

Stability and Variability of the Thermohaline Circulation and its Link to Climate

Andrew J. Weaver and Tertia M.C. Hughes

The thermohaline circulation, driven by fluxes of heat and freshwater through the ocean's surface, is an important mechanism for the transport of heat from low to high latitudes. Changes in the intensity of the thermohaline circulation, and hence its poleward heat transport, would have a significant effect on global climate. In this article we review some recent contributions to the understanding of the stability and variability properties of the thermohaline circulation. We begin by discussing recent observations of the global ocean thermohaline circulation and present evidence of decadal-millennial climate variability. A review of a hierarchy of models, starting from idealized box models and then continuing to more complicated zonally-averaged models, is then presented. We further summarize recent results obtained using uncoupled ocean general circulation models and finish with a discussion of fully coupled, three-dimensional, atmosphere-ocean general circulation models. The importance of freshwater flux versus thermal forcing in determining the stability and variability properties of the ocean's thermohaline circulation is underlined throughout this review.

On the Importance of Vertical Resolution in Certain Ocean General Circulation Models

Andrew J. Weaver and Ed S. Sarachik

In centred difference models of ocean circulation, two grid point computational modes can be excited if grid Reynolds and Peclet numbers are greater than two. The Bryan-Cox General Circulation Model (GCM) is used to show the dramatic effect that this instability has on the equatorial thermohaline circulation. In many recent numerical calculations researchers have used 12 vertical levels. It is shown that this resolution produces an artificial cell at the equator when typical values of the vertical diffusivity and viscosity parameters are used. This artificial cell rotates counter to the primary cell driven by deep water formation at high latitudes, is driven by downwelling at the eastern boundary near the equator and is 40% the strength of the primary cell for the parameters used in the present study. When the vertical resolution is increased the cell vanishes. It is suggested therefore that higher vertical resolution should be used in Bryan-Cox GCM deep ocean modelling studies when current values of the vertical diffusivity and viscosity parameters are used.

The Role of Mixed Boundary Conditions in Numerical Models of the Ocean's Climate

Andrew J. Weaver and Ed S. Sarachik

Several simple numerical experiments are conducted, using both single and double hemisphere ocean basins under symmetric steady forcing and using low vertical eddy viscosity, to study the ocean's thermohaline circulation. It is shown that a stable steady state obtained under a restoring surface boundary condition on salinity becomes unstable upon a switch to a flux boundary condition. The Polar Halocline Catastrophe of F. Bryan (1986a) occurs. It is shown that further integration of this collapsed state ultimately yields a steady, stable one-cell circulation with the approach being essentially chaotic but with significant energy at decadal period. The two hemisphere ocean passes through many stages in which violent overturning occurs (order 80+ Sv). These flushes occur in both hemispheres and are of one-cell structure. The time period between these flushes varies from several hundred to about one thousand years. A single 12 vertical level hemispheric basin, spun up from an initial state of rest under mixed boundary conditions (restoring boundary condition on temperature and flux boundary condition on salinity), never reaches a steady state. Three characteristic stages are observed in the integration: A stage where the system oscillates with decadal timescale, a stage when the system undergoes a violent overturning flush, and a quiescent stage in which either deep water is forming or the thermohaline circulation is in a collapsed state. These three characteristic stages are also present in 33 level single and double hemisphere runs. The decadal timescale is associated primarily with the advection of positive salinity anomalies into the region of deep water formation from the mid-ocean region between the subtropical and subpolar gyres. Upon increasing the resolution to 33 levels a steady state is reached. The resulting steady state is fundamentally different from the one obtained under the same resolution and restoring boundary conditions in that it is more energetic and has much warmer basin mean temperature. These differences are due to a change in the location of deep water formation. The dependence of the results on the type of convection scheme used, vertical resolution and timestepping procedure (synchronous or asynchronous integration) is also studied in order to separate physical processes from those which might be numerical artifacts. Sufficient vertical resolution is shown to be crucial in obtaining realistic models of the thermohaline circulation. It is shown that a steady state which is stable under asynchronous integration and mixed boundary conditions becomes unstable upon a switch to synchronous integration. It is further shown that the steady state obtained under restoring boundary conditions only changes slightly upon a switch to synchronous integration. Under mixed boundary conditions the steady state is shown to be very sensitive to the choice of surface tracer timestep even while integrating asynchronously. Upon a switch in this timestep a polar halocline catastrophe may be induced. The implications of the present study for future ocean climate models are discussed.

Evidence for decadal variability in an ocean general circulation model: An advective mechanism

Andrew J. Weaver and Ed S. Sarachik

A series of numerical experiments involving long-time integrations are conducted using the Bryan-Cox Ocean General Circulation Model under mixed surface boundary conditions (i.e., a Newtonian restoring surface boundary condition on temperature and a specified flux boundary condition on salinity). Under steady forcing the system oscillates with significant energy at decadal period. This oscillation is shown to be an advective phenomenon, associated with the propagation of salinity and temperature anomalies from the region between the subtropical and subpolar gyres, where they are generated, to the eastern boundary, where deep water is formed. Furthermore, the oscillation is characterized by the fluctuation of the thermohaline circulation between a state in which deep water is formed and a collapsed state with no deep water formation. Over the period of the oscillation the poleward heat transport changes by as much as a factor of three at certain latitudes. The anomalies are initially formed by the upwelling of warm, saline waters which are being transported polewards by a western boundary current that has separated from the coast. The observed decadal variability is robust in that it is present in all numerical experiments (12 and 33 vertical level models; one and two hemisphere models; synchronous and asynchronous integrations).

An atmospheric energy moisture-balance model: climatology, interpentadal climate change and coupling to an OGCM.

Augustus F. Fanning and Andrew J. Weaver

An atmospheric model incorporating energy and moisture balance equations is developed for use in process studies of the climate system. Given the sea surface temperature and specified surface wind field, the atmospheric model calculates the surface fields of: air temperature, specific humidity, as well as heat and freshwater fluxes. The inclusion of the moisture balance in the atmospheric model allows the effects of latent heat transport to be included explicitly in the model.

Under fixed climatological sea surface temperature (SST) and surface wind conditions, surface air temperatures, specific humidities and surface fluxes are comparable to direct estimates. Precipitation compares less favorably with observations. As an extension to the climatological forcing case, we conduct a simple perturbation experiment in which the 1955-59 pentad is compared to the 1970-74 pentad by riving the model under the respective SST fields. The model exhibits a global air temperature decrease in the latter pentad of 0.27 Deg C (comparable to direct estimates) with cooling in the northern hemisphere, and warming in the southern hemisphere. Such large scale cooling in our atmospheric model is driven by equivalent local changes in the prescribed SST fields, subsequently smoothed by atmospheric diffusion of heat. The interpentadal modeled differences are shown to be quite robust through model experiments using parameters representative of several different unrealistic climatologies. The resulting interpentadal difference fields change remarkably little even when the background state has changed dramatically. This emphasizes the almost linear response of the atmospheric model to the imposed SST changes.

The atmospheric model is also coupled to an ocean general circulation model without the need for flux adjustments. This coupled climate model faithfully represents deepwater formation in the North Atlantic and Southern Ocean, with upwelling throughout the Pacific and Indian oceans. Water mass characteristics in the vertical compare very favorably with direct observations.

JEBAR, bottom pressure torque and Gulf Stream separation

Paul G. Myers, Augustus F. Fanning and Andrew J. Weaver

A diagnostic finite element barotropic ocean model has been used to simulate the circulation in the North Atlantic. With the inclusion of the JEBAR term (the Joint Effect of Baroclinicity And Relief), the Gulf Stream is found to separate at the correct latitude, about 35 degrees N, off Cape Hatteras. Results suggest that the JEBAR term in three key regions (offshore of the separation point in the path of the main jet, along the slope region of the North Atlantic Bight and in the central Irminger Sea) is crucial in determining the separation point. The transport driven by the bottom pressure torque component of JEBAR dominates the solution, and is also responsible for the separation of the Gulf Stream. Excluding high latitudes (in the deep water formation regions) density variations in the upper 1000 m of the water column govern the generation of the necessary bottom pressure torque in our model. Examination of results from the WOCE-CME (World Ocean Circulation Experiment - Community Modelling Effort) indicates that the bottom pressure torque effects are underestimated in the CME by almost an order of magnitude. Removing this term in each of the three key areas forces our modelled Gulf Stream to overshoot its observed separation point as in the Community Model. The reason for the behaviour of the CME is unclear, but may be associated with the diffuse nature of the modelled thermocline as suggested by our model's sensitivity to the density field above 1000 m.

A diagnostic barotropic finite element ocean circulation model.

Paul G. Myers and Andrew J. Weaver

The finite element method possesses many advantages over more traditional numerical techniques used to solve systems of differential equations. These advantages include a number of conservation properties and a natural treatment of boundary conditions. The method's piecewise nature makes it useful when dealing with irregular domains, and similarly when using variable horizontal resolution. To take advantage of these properties, a finite element representation of the linearized, steady-state, barotropic potential vorticity equation, is developed. The Stommel problem is used as an initial test for the model. A fourth order eddy viscosity term is then added and the resulting problem is solved in both simply- and multiply-connected domains under both slip and no-slip boundary conditions. The beta plane assumption is then relaxed and the model reformulated in spherical co-ordinates. A realistic geography and topography version of this model is also used to examine the barotropic circulation in the North Atlantic Ocean. Results are found to agree very well with those of previous diagnostic calculations. In particular, the Gulf Stream separates at the correct latitude with the inclusion of the JEBAR term.

Interdecadal Climate Variability in the Subpolar North Atlantic

Trudy M.H. Wohlleben and Andrew J. Weaver

The statistical relationships between various components of the subpolar North Atlantic air-sea-ice climate system are reexamined in order to investigate potential processes involved in interdecadal climate variability. It is found that sea surface temperature anomalies concentrated in the Labrador Sea region have a strong impact upon atmospheric sea level pressure anomalies over Greenland, which in turn influence the transport of freshwater and ice anomalies out of the Arctic Ocean, via Fram Strait. These freshwater and ice anomalies are advected around the subpolar gyre into the Labrador Sea affecting convection and the formation of Labrador Sea Water. This has an impact upon the transport of North Atlantic Current water into the subpolar gyre and thus, also upon sea surface temperatures in the region. An interdecadal climate loop is therefore proposed as an internal source of climate variability within the subpolar North Atlantic. Through the lags associated with the correlations between different climatic components, observed horizontal advection timescales, and the use of Boolean Delay Equation models, the timescale for one cycle of this loop is determined to have a period of about 21 years.

Semi-Lagrangian Advection Algorithms for Ocean Circulation Models

Salil Kumar Das and Andrew J. Weaver

The semi-Lagrangian advection scheme is applied to the meridional-plane model of the thermohaline circulation, developed by Marotzke et al (1988), whose governing equations are solved under a variety of boundary conditions. To determine the extent to which the accuracy and efficiency of the calculations depends on the numerical integration scheme, the test problem is solved independently using an explicit finite difference (leap-frog in time, centered difference in space) method and three implicit methods: a finite difference, a finite element and an upwind scheme. Integrations of the model to several equilibria are performed to determine the accuracy, efficiency and stability of each integration scheme as a function of time step. For the same level of accuracy the time step used in the semi-Lagrangian scheme is found to be at least five times greater than that employed in the case of the implicit methods. The time step used in the implicit methods in turn are at least six times greater than those needed in the explicit integration of the governing equations. It is further shown that Dirichlet, Neumann and mixed boundary conditions can be handled efficiently with the semi-Lagrangian method. The semi-Lagrangian method is applied in the usual three-time level and two-time level interpolating versions as well as in a noninterpolating, three-time level version. The two-time level scheme further doubles the speed of the time integration step for the same level of accuracy, beyond that which is achieved using the three-time level scheme. The noninterpolating scheme does not eliminate the damping introduced by the interpolation, as pointed out by Ritchie (1986). Hence, the two-time level, semi-Lagrangian advection method stands out as a viable time integration scheme for climate models which are normally run for hundreds of years and is best suited for ocean climate studies.

On the numerical implementation of mixing parameterizations in the GFDL ocean model

Andrew J. Weaver and Michael Eby

The results from ocean model experiments conducted with isopycnal and isopycnal thickness diffusion parameterizations for subgrid scale mixing associated with mesoscale eddies are examined from a numerical standpoint. It is shown that when the mixing tensor is rotated, so that mixing is primarily along isopycnals, numerical problems may occur and non-monotonic solutions which violate the second law of thermodynamics may arise. These numerical problems can be reduced or eliminated if sufficient explicit background horizontal diffusion is added to the mixing scheme. A more appropriate solution is the use of more sophisticated numerical advection algorithms, such as the flux-corrected transport algorithm. This choice of advection scheme adds additional mixing only where it is needed to preserve montonicty and so retains the physically-desirable aspects of the isopycnal and isopycnal thickness diffusion parameterizations, while removing the undesirable numerical noise. The price for this improvement is a computational increase.

Paleoclimatic response of the closing of the Isthmus of Panama in a coupled ocean-atmosphere model

Trevor Q. Murdock, Andrew J. Weaver and Augustus F. Fanning

The paleoclimatic effects of the closure of the Isthmus of Panama ~3 million years ago are investigated using an ocean general circulation model coupled to a energy-moisture balance model and a thermodynamic ice model. Consistent with earlier ocean-only modelling studies, it is shown that prior to the closing of the Isthmus of Panama, the Atlantic behaved more similar to the present-day Pacific Ocean with a conspicuous absence of deep water formation. Associated with the absence of North Atlantic deep water formation is a significant reduction in both the Atlantic and global oceanic heat transports. This reduction in oceanic heat transport is largely compensated for by an increase in the total atmospheric heat transport, with the result that only small changes in planetary heat transport occur. Model results suggest that the present-day climate of the North Atlantic is significantly warmer, together with a general cooler trend in the southern hemisphere and the region surrounding the Pacific Ocean, than before the closure of the Isthmus. Finally, possible relationships to glaciation and initiation of glacial cycles in the Northern Hemisphere, are discussed.

Temporal-geographical meltwater influences on the North Atlantic Conveyor: Implications for the Younger Dryas.

Augustus F. Fanning and Andrew J. Weaver

The temporal and geographical roles of meltwater discharge (from the Laurentide ice sheet) on North Atlantic Deep Water (NADW) production are investigated utilizing a global, realistic geometry, coupled climate model which does not require the use of flux adjustments. Model results suggest that preconditioning by meltwater discharge (to the Mississippi) prior to the Younger Dryas (YD) is capable of pushing NADW beyond the limit of its sustainability. The diversion of meltwater to the St. Lawrence then merely serves to completely inhibit NADW production. The modeled change in surface air temperature generally agrees with the global pattern and magnitude of temperature change seen in paleoclimatic reconstructions of the YD and is intimately linked to changes in NADW formation. The global thermohaline circulation provides an interhemispheric teleconnection with the Southern Oceans, while changes in the atmospheric heat transport (reacting to a global redistribution of oceanic heat transport) provide a mechanism for interbasin teleconnection. Although the primary thermodynamic and hydrological cycle feedback processes are included within the atmospheric model, in the absence of additional feedbacks an equilibrium without the presence of NADW is possible. The inclusion of the wind stress/speed feedback is found to significantly contribute to the resumption of NADW production, as suggested by previous studies. Contrary to these same studies, however, the coupled model indicates an advective spin-up timescale is required for resumption of NADW production and hence the termination of the modeled YD-like climate event (as opposed to a decadal-century timescale). The reason for the discrepancy is unclear but may be associated with the use of fixed salt flux fields applied in previous studies, or the duration, strength, and geographical location of the imposed meltwater applied.

A horizontal resolution and parameter sensitivity study of heat transport in an idealized coupled climate model.

Augustus F. Fanning and Andrew J. Weaver

An idealized coupled ocean-atmosphere model is utilized to study the influence of horizontal resolution and parameterized eddy processes on the poleward heat transport in the climate system. A series of experiments ranging from 4 Deg to 1/4 Deg resolution, with appropriate horizontal viscosities and diffusivities in each case are performed. The coupled atmosphere-ocean model results contradict earlier studies which showed that the heat transport associated with time varying circulations counteracts increases in the time mean so that the total remained unchanged as resolution was increased. Even though the total oceanic heat transport has not converged, the net planetary heat transport has essentially converged owing to the strong constraint of energy balance at the top of the atmosphere. Consequently, the atmospheric heat transport is reduced to offset the increasing oceanic heat transport.

To interpret these results, the oceanic heat transport is decomposed into its baroclinic overturning (related to the meridional overturning and Ekman transports), barotropic gyre (that in the horizontal plane) and baroclinic gyre (associated with the jet core within the western boundary current) components. The increase in heat transport occurs in the steady currents. In particular the baroclinic gyre transport increases by a factor of 5 from the coarsest to the finest resolution case, equaling the baroclinic overturning transport at mid to high latitudes.

To further assess the results, a parallel series of experiments under restoring conditions are performed to elucidate the differences between heat transport in coupled versus uncoupled models, and models driven by temperature and salinity or equivalent buoyancy. Although heat transport is more strongly constrained in the restoring experiments, results are similar to those in the coupled model. Again, the total heat transport is increased due to an increasing baroclinic gyre component.

These results point to the importance of higher resolution in the oceanic component of current coupled climate models. These results also stress the need to adequately represent the heat transport associated with the `Warm Core' region of the Gulf Stream (the baroclinic gyre transport) in order to adequately represent oceanic poleward heat transport.

On the influence of the parameterization of lateral boundary layers on the thermohaline circulation in coarse-resolution ocean models

Thierry Huck, Andrew J. Weaver and A. Colin De Verdiere

Because of the first order geostrophic balance in the ocean interior, the parameterization of lateral boundary layers is more important than the parameterization of viscosity for shaping up the thermohaline overturning and the deep water properties in coarse-resolution ocean circulation models. Different formulations of momentum dissipation and associated boundary conditions are implemented within a planetary-geostrophic ocean circulation model for a Cartesian coordinate, flat-bottomed, b-plane, with restoring boundary conditions for the surface density and zero wind stress. Traditional Laplacian friction with a no-slip boundary condition produces an interior circulation in good agreement with geostrophy and the Sverdrup balance, but generates very large vertical (diapycnal) transports at lateral boundaries, especially upwelling in the western boundary current and downwelling in the north-east corner. The meridional and zonal overturning are thus enhanced, but drive to depth surface waters that are not as cold as the ones in the deep convection regions. Rayleigh friction with various frictional closures for the along-shore velocities within a no-normal-flow boundary condition framework efficiently reduce the diapycnal vertical transports along the boundaries, by allowing horizontal recirculation of geostrophic currents impinging into coasts. Hence, these parameterizations induce weaker overturnings, with colder deep water and a sharper thermocline resulting in higher poleward heat transports. We suggest that the upwelling along the boundaries is a consequence of the coarse- resolution dynamics and not horizontal diffusion (termed the `Veronis effect', horizontal diffusion produces large diapycnal fluxes once the isopycnals are tilted by coastal upwellings). Alternative parameterizations for the lateral boundary layers reduces this effect without the need for rotating the mixing tensor along isopycnals. This model inter- comparison proves the need to clearly assess the extent of the diapycnal upwelling in the western boundary currents and to develop physically-based parameterizations of lateral boundary layers in order to improve coarse-resolution OGCMs.

On the role of flux adjustments in an idealized coupled climate model

Augustus F. Fanning and Andrew J. Weaver

The effect of employing flux adjustments on the climatic response of an idealized coupled model to an imposed radiative forcing is investigated with two coupled models, one of which employs flux adjustments. A linear reduction (to the planetary longwave flux) of 4W/m2 is applied over a 70 year period and held constant thereafter. Similar model responses are found (during the initial 70 year period) for global-scale diagnostics of hemispheric air temperature due to the nearly linear surface air temperature response to the radiative forcing. Significant regional scale differences do exist, however. As the perturbation away from the present climate grows, basin-scale diagnostics (such as meridional overturning rates) begin to diverge between flux adjusted and non-flux adjusted models. Once the imposed radiative forcing is held constant, however, differences in global mean air temperature of up to 0.5 Deg are found, with large regional-scale differences in air temperature and overturning rates within the North Atlantic and Southern Ocean. Two additional experiments with the flux adjusted model (beginning from points further along the control integration) suggest that the elimination of much of the coupling shock before the radiative forcing is applied leads to results slightly closer to the non-flux adjusted case, although large differences still persist. In particular a dipole structure indicating an enhanced warming within the Pacific sector of the Southern Ocean, and cooling within the Atlantic sector is not reproduced by the flux adjusted models. This disparity is intimately linked to the Southern Ocean overturning cell along with the flux adjustments employed as well as the drift arising from coupling shock. If a similar form of sensitivity exists in more realistic coupled models, our results suggest: 1) perturbation experiments should not be undertaken until after the coupled model control experiment is carried out for several hundred years (thereby minimizing the coupling shock); 2) care should be exercised in the interpretation of regional-scale results (over the ocean) in coupled models which employ flux adjustments; 3) care should also be taken in interpreting even global-scale diagnostics in flux adjusted models for large perturbations about the present climate.

Thermohaline variability: The effects of horizontal resolution and diffusion

Augustus F. Fanning and Andrew J. Weaver

An idealized coupled ocean-atmosphere model is utilized without flux adjustments to study the influence of horizontal resolution and parameterized eddy processes on the thermohaline circulation. A series of experiments ranging from 4 Deg x 4 Deg to 0.25 Deg x 0.25 Deg resolution, with appropriate horizontal viscosities and diffusivities in each case are performed for both coupled and ocean-only models. Spontaneous decadal-intradecadal variability (whose period varies slightly between cases) is found to exist in the higher resolution cases (with the exception of one of the restoring experiments). The oscillation is described as an advective-convective mechanism which is thermally driven, and linked to the value of the horizontal diffusivity utilized in the model. Increasing the diffusivity in our high resolution cases is enough to destroy the variability, while decreasing the diffusivity in the moderately coarse resolution cases is capable of inducing the variability. As the resolution is increased still further, baroclinic instability within the western boundary current adds a more stochastic component to the solution such that the variability is less regular and more chaotic. These results point to the importance of higher resolution in the ocean component of coupled models, revealing the existence of richer decadal-intradecadal scale variability in models which require less parameterized diffusion.

On the variability of the thermohaline circulation in the GFDL coupled model

Sophie Valcke and Andrew J. Weaver

The ocean component of the GFDL coupled model is used to investigate whether or not the interdecadal variability found in Delworth et al. (1993) is an ocean-only mode or a mode of the full coupled system. In particular, it has been previously suggested that the variability in the full coupled model is either: 1) an ocean-only mode which is excited by atmospheric noise; 2) an internal ocean mode driven by fixed atmospheric fluxes (flux adjustment plus annual mean) which is made less regular through forcing from atmospheric noise; 3) a consequence of the use of flux adjustments. Through a series of experiments conducted under fixed flux boundary conditions we show that none of these three hypotheses holds and therefore conclude that the interdecadal variability found in Delworth et al. (1993) is a mode of the full coupled system.

Decadal variability of the thermohaline circulation in ocean models

Thierry Huck, Andrew J. Weaver and A. Colin De Verdiere

Intrinsic modes of decadal variability driven by constant surface buoyancy fluxes are analysed using a box-geometry ocean model. A complete parameter sensitivity analysis of the oscillatory behavior is carried out with respect to the spherical Cartesian geometry, the beta-effect, the Coriolis parameter, the parameterization of momentum dissipation and associated boundary conditions and viscosities, the vertical and horizontal diffusivities, the convective adjustment parameterization and the horizontal and vertical model resolution. The oscillation stands out as a robust geostrophic feature whose amplitude is mainly controlled by the horizontal diffusivity. Since the beta-effect has no influence, Rossby waves play no role in the mechanism (Winton 1996). Various experiments with different geometry and forcing are conducted to test the importance of numerical boundary waves in sustaining the oscillation. The results suggest that only the western boundary is crucial (Greatbatch and Peterson 1996). The analysis of the variability patterns differentiates two types of oscillatory behavior: temperature anomalies traveling westward in an eastward jet (northern part of the basin) inducing geostrophically an opposite anomaly in their wake; temperature anomalies in the north-west corner which respond to the western boundary current transport changes, but reinforce geostrophically this change and build the opposite temperature anomaly in the east, which finally reverse the meridional overturning anomaly (and thus the anomalous western boundary current transport). The analysis of the transition from steady to oscillatory states suggests, in agreement with a one layer and a half model, that the variability is triggered in the regions of strongest cooling. Finally, we propose a simple box-model analogy that captures the observed phase-shift between meridional overturning and north-south density gradient anomalies.

Simulated influence of carbon dioxide, orbital forcing and ice-sheets on the climate of the last glacial maximum.

Andrew J. Weaver, Michael Eby, Augustus F. Fanning and Edward C. Wiebe

A fully coupled atmosphere-ocean-sea ice model is used to investigate the climate of the Last Glacial Maximum (21 KBP) and the competing effects of CO₂, orbital forcing and continental ice sheet albedo. The equilibrium ocean circulation within the coupled model yields tropical sea surface temperature differences from the present climate which are colder than CLIMAP estimations. Specifically, model results reveal annual and global mean surface air temperature cooling of ~3.2 Deg C and sea surface temperature cooling of ~2.3 Deg C, the latter with 1.9-3.1 Deg C cooling in the tropical Pacific and 2.0 Deg C cooling in the tropical Indian Oceans. In the Atlantic Ocean, tropical annual mean sea surface temperature differences range from 2.0-3.4 Deg C cooling with an amplification of the signal in the North Atlantic to as high as 6.3 Deg C. This North Atlantic amplification is not as large as that suggested by CLIMAP and is associated with a weakening and shallowing of the conveyor there. A sensitivity study to perturbations in model river drainage basins, and hence indirectly the strength of the conveyor in the present climate, yields remarkable results. First, when the initial present-day conveyor is close to an instability threshold of ~12 Sv, perturbations to the radiative forcing via changes in orbital forcing and CO₂ cause the conveyor to shutdown. The stronger the conveyor is in the present climate, the less it changes and hence the smaller the change in ocean heat transport under 21 KBP forcing. In all cases, regardless of the strength of the present day conveyor, the global mean sea surface temperature and surface air temperature changes are the same between 21KBP and the present, although regional differences do exist.

Thermohaline circulation: high latitude phenomena and the difference bewteen the Pacific and Atlantic.

Andrew J. Weaver, Cecilia M. Bitz, Augustus F. Fanning and Marika M. Holland

Deepwater formation, the process whereby surface water is actively converted to deep water through heat and freshwater exchange at the air-sea interface, is known to occur in the North Atlantic but not to occur in the North Pacific. As such, the thermohaline circulation is fundamentally different in these two regions. In this review we provide a description of this circulation and ouline a number of reasons as to why the North Atlantic forms deepwater but not the North Pacific. Special emphasis is given to the role of interactions with the Arctic Ocean. We extend our analysis to discuss observational evidence and current theories for decadal-interdecadal climate variability in each region, with particular focus on the role of the ocean. Differences between the North Atlantic and North Pacific are once more highlighted.

Millennial timescale variability in ocean/climate models

Andrew J. Weaver

A review of mechanisms for millennial timescale variability from a hierarchy of ocean models is presented together with a comparison with observed variability in the last glaciation (Dansgaard-Oeschger oscillations, Heinrich events and Bond Cycles). Special attention is also given to a review of modelling efforts aimed at unraveling the causes and consequences of the Younger Dryas cooling event (12,700 - 11,650 years BP) as well as potential mechanisms for interglacial millennial timescale variability. Finally, some recent experiments are included to examine the influence of Heinrich event runoff (obtained from a continental ice sheet model) on the global ocean circulation in a coupled atmosphere- ocean-sea ice model of intermediate complexity.

Data-model comparison of the Younger Dryas

Nathaniel W. Rutter, Andrew J. Weaver, Dean Rokosh, Augustus F. Fanning, and Daniel G. Wright

The Younger Dryas cooling event is well-established in the North Atlantic region through numerous climate proxy records. Although the climatological controls vary from site to site, it is considered to have taken place between about 10 000 and 11 000 radiocarbon years BP(�11 000 - 13 000 calendar years BP). Outside the North Atlantic region, climate proxy records and chronology commonly became problematic because of weaker signals and less dating control. In addition to this evidence, oceanic records reveal conflicting evidence for YD forcing mechanisms and the timing of events. We compare proxy evidence with the results from an ocean general circulation model coupled to the energy/moisture balance atmospheric model. The model results reveal a global pattern and regional magnitude which generally agree with temperature changes interpreted from paleoclimate reconstructions. The model also supports the general duration of global cooling of the Younger Dryas.

Although proxy data is controversial outside of the North Atlantic region, the authors believe that there is enough evidence to support the YD cooling event on a global scale. They also recognize, however, that more concrete evidence is needed before the question can be unequivocally answered.

On the sensitivity of global warming experiments to the parametrisation of sub-grid scale ocean mixing

Edward C. Wiebe and Andrew J. Weaver

An ocean general circulation model coupled to an energy-moisture balance atmosphere model is used to investigate the sensitivity of global warming experiments to parameterisations of sub-grid scale ocean mixing. The climate sensitivity of the coupled model using three different parameterisations of sub-grid scale mixing is 3C for a doubling of CO₂ (6C for a quadrupling of CO₂). This suggests that the ocean has only a weak feedback on global mean surface air temperature although significant regional differences, notably at high latitudes, exist with different sub-grid scale parameterisations. In the experiment using the Gent and McWilliams parameterisation for mixing associated with mesoscale eddies, an enhancement of the surface response in the Southern Ocean is found. This enhancement is largely due to the existence of more realistic sea-ice in the climatological control integration and the subsequent enhanced ice albedo feedback upon warming. In accordance with earlier analyses, the Gent and McWilliams scheme decreases the global efficiency of ocean heat uptake. During the transient phase of all experiments, the North Atlantic overturning initially weakened but ultimately recovered, surpassing its former strength. This suggests that in the region around the North Atlantic the ocean acts as a negative feedback on local warming during the transient phase but a positive feedback at equilibrium. During the transient phase of the experiments with a more sophistic ated and realistic parameterisation of sub-grid scale mixing, warmed Atlantic water was found to penetrate at depth into the Arctic, consistent with recent observations in the region.

The Canadian Centre for Climate Modeling and Analysis global coupled model and its climate

Greg M. Flato, George J. Boer, Warren G. Lee, Norm A. McFarlane, Dave Ramsden, M. Cathy Reader and Andrew J. Weaver

A global, three-dimensional climate model, developed by coupling the CCCma second-generation atmospheric general circulation model (GCM2) to a version of the GFDL modular ocean model (MOM1), forms the basis for extended simulations of past, current and projected future climate. The spin-up and coupling procedures are described, as is the resulting climate based on a 200-year model simulation with constant atmospheric composition and external forcing. Comparison of this control simulation to observations demonstrates a generally successful reproduction of mean climate quantities. Variability is also generally well-simulated over land, but somewhat underestimated over the tropical ocean. The modelled climate state shows only small trends, indicating a reasonable level of balance at the surface, which is achieved in part by the use of heat and freshwater flux adjustments. The control simulation provides a basis against which to compare simulated climate change due to historical and projected greenhouse gas and aerosol forcing as described in companion publications.

Extratropical Subduction and Decadal Modulation of El Nino

Andrew J. Weaver

An extension of the Battisti and Hirst delayed oscillator model is developed in an attempt to understand the potential effects of extratropical subduction on El Nino. This extension involves the inclusion of a meridional delay term parameterizing the effects of estratropical subduction on eastern equatorial Pacific pycnocline anomalies. The magnitude of the eastern equatorial pycnocline displacement, and the time delay for its response to an extratropical sea surface temperature anomaly, are inferred from experiments with both ocean and coupled atmosphere-ocean GCMs. The inclusion of the delayed extratropical subduction term causes El Nino to modulate on decadal-interdecadal timescales, consistent with observations and the expectations of Gu and Philander.

Late Ordovician glaciation and high atmospheric CO₂: A coupled model analysis

Pascale F. Poussart, Andrew J. Weaver and Christopher R. Barnes

The analysis of the geologic record has revealed a question concerning how Late Ordovician glaciation could have occurred simultaneously with high atmospheric CO₂ levels (10-18x). Sensitivity studies using a coupled atmosphere-ocean-sea ice model show that it is possible to maintain a permanent snow cover (which corresponds to 60% of all the glacial deposits found on Gondwana) under 10x CO₂ levels, warm fall/cool spring orbital parameters, a 4.5% reduction in solar luminosity, a length of day of 21.5 hours and an enhanced snow/sea ice albedo of 0.3. A cold summer orbit experiment with 10x CO₂ and a reduced snow/sea ice albedo of 0.1 also sustains a permanent (albeit less extensive) snow cover. Results from all experiments consistently reveal the presence of a peramanent sea ice cover extending from the north pole to 60-70 deg N. The geographic configuration of the late Ordovician results in an up to 42% increase in the global ocean poleward heat transport in the Southern Hemisphere relative to present-day and a significant asymmetry relative to the equator. The single most important internal parameter, which determines whether a permanent snow cover existed or not, is the magnitude of the ice/snow albedo feedback.

Effects of sinking of salt rejected during formation of sea ice on results of a global ocean-atmosphere-sea ice climate model

Phil B. Duffy, Michael Eby and Andrew J. Weaver

Simulations performed with an ocean-atmosphere-sea-ice model show significant sensitivities to the treatment of salt rejected during sea-ice formation. In our Control simulation, we place rejected salt in the top ocean-model level. In the Plume simulation, we instantaneously mix rejected salt into the subsurface ocean, to a maximum depth which depends on local density gradients. This mimics the effects of subgrid-scale vertical mixing of rejected salt. Compared to that of the Control simulation, the steady-state climatology of the Plume simulation is more realistic: the spatial pattern of simulated salinities is more realistic, especially in the Southern Ocean; deep-ocean temperatures are warmer and more realistic, again especially in the Southern Ocean; simulated sea-ice extents and surface air temperatures also agree better with observations. A similar pair of simulations using horizontal tracer diffusion instead of the Gent-McWilliams eddy parameterization show similar changes due to instantaneous mixing of rejected salt.

Evaluation of ocean and climate models using present-day observations and forcing

Andrew J. Weaver, Phil B. Duffy, Michael Eby and Edward C. Wiebe

The most common method used to evaluate climate models involves spinning them up under perpetual present-day forcing and comparing the model results with present-day observations. This approach clearly ignores any potential long term memory of the model ocean to past climatic conditions. Here we examine the validity of this approach through the 6000 year integration of a coupled atmosphere-ocean-sea ice model. The coupled model is initially spun-up with atmospheric CO₂ concentrations and orbital parameters applicable for 6KBP. The model is then integrated forward in time through to 1998. Results from this transient coupled model simulation are compared with the results from two additional simulations, in which the model is spun up with perpetual 1850 (preindustrial) and 1998 (present-day) atmospheric CO₂ concentrations and orbital parameters. This comparison leads to substantial differences between the equilibrium climatologies and the transient simulation, even at 1850 (prior to any significant changes in atmospheric CO₂). When compared to the present-day equilibrium climatology, differences are very large: the global mean surface air and sea surface temperatures are ~0.5. Deg C and ~0.4 Deg C colder, respectively, deep ocean temperatures are substantially cooler, southern hemisphere sea ice cover is 38% larger, and the North Atlantic conveyor 3 Sv weaker in the transient case. These differences are due to the long timescale memory of the deep ocean to climatic conditions which prevailed throughout the late Holocene, as well as to its large thermal inertia. Our results question the validity of current techniques for climate model evaluation and underline the importance of using paleoclimatic simulations in parallel with present-day simulations in this evaluation process. In addition, they point out that the 'cold start' problem, whereby global warming experiments which do not take into account prior buildup of greenhouse gases are thought to display an anomalous low warming rate, is actually a `warm start' problem.

The impact of rising atmospheric CO₂ levels on low frequency North Atlantic climate variability

Marika M. Holland, Aaron J. Brasket and Andrew J. Weaver

Observations show that the North Atlantic climate system possesses pronounced interdecadal variability in its sea-ice, ocean and atmosphere components. The long timescale associated with this variability suggests that the ocean, and in particular its thermohaline component, may play an integral role in its mechanism. A question which naturally arises concerns how such variability will change under increased levels of atmospheric CO₂. Several studies have examined the impact of increased atmospheric CO₂ on climate variability through the use of atmospheric/ocean mixed layer model or full coupled models, although they have used fairly short integrations (20-80 years) and largely focused on changes in atmospheric variability. Here we use a coupled ice/ocean/atmosphere model in an attempt to quantify potential changes in North Atlantic interdecadal climate variability under increased atmospheric CO₂. We focus our attention on the variability of the thermohaline circulation induced by fluctuations in Arctic Ocean ice export to the North Atlantic. Under 2XCO₂ conditions, the variance of the thermohaline circulation is reduced to 7% of its simulated value under present day forcing. This decrease is caused by relatively low ice export variability and changes in the primary ice melt location in the northern North Atlantic.

The role of ice ocean interactions in the variability of the North Atlantic thermohaline circulation

Marika M. Holland, Cecelia M. Bitz, Michael Eby and Andrew J. Weaver

The simulated influence of Arctic sea ice on the variability of the North Atlantic climate is discussed in the context of a global coupled ice/ocean/atmosphere model. This coupled system incorporates a general circulation ocean model, an atmospheric energy moisture balance model and a dynamic/thermodynamic sea ice model. Under steady seasonal forcing, an equilibrium solution is obtained with very little variability. To induce variability in the model, daily varying stochastic anomalies are applied to the wind forcing of the northern hemisphere sea ice cover. These stochastic anomalies have observed spatial patterns but are random in time. Model simulations are run for 1000 years from an equilibrium state and the variability in the system is analyzed. The sensitivity of the system to the ice/ocean coupling of both heat and fresh water is also examined.

Under the stochastic forcing conditions, low amplitude (approximately 10% of the mean) variability in the thermohaline circulation (THC) occurs. This variability has significant spectral power at interdecadal timescales which are concentrated at approximately 20 years. It is forced by fluctuations in the export of ice from the Arctic into the North Atlantic. Large changes in sea-surface temperature and salinity are related to changes in the overturning circulation and the sea ice coverage in the northern North Atlantic. Additionally, the THC variability influences the Arctic basin through heat transport under the ice pack.

Results from sensitivity studies suggest that the fresh water exchange from the variable ice cover is the dominant process for forcing variability in the overturning. The simulated Arctic ice export provides a stochastic forcing to the northern North Atlantic which excites a damped oscillatory ocean-only mode. The insulating capacity of the variable sea ice has a negligible effect on the THC. Ice/ocean thermal coupling acts to preferentially damp THC variability with periods greater than 30 years, but has little influence on variability at higher frequencies.

Projections of climate change onto modes of atmospheric variability

Daithi A. Stone, Andrew J. Weaver and Ron J. Stouffer

Two possible interpretations of forced climate change view it as projecting, either linearly or nonlinearly, onto the dominant modes of variability of the climate system. An evaluation of these two interpretations is performed using sea level pressure (SLP) and surface air temperature (SAT) fields obtained from integrations of the Geophysical Fluid Dynamics Laboratory coupled general circulation model forced with varying concentrations of greenhouse gases.

The dominant modes of SLP both represent much of the total variability and remain important in warmer climates. With SAT, however, the dominant modes are often related to variations in the sea ice edge and so do not remain important since the ice retreats as the climate warms; those unrelated to sea ice remain dominant in the warmer climates, but represent smaller fractions of the total variability.

The change in SLP projects partially onto the AO-like mode in the Northern Hemisphere. In the Southern Hemisphere the change projects negligibly onto the dominant modes between equilibrium climates, but almost entirely onto the AAO-like mode in the transient integration. This difference between the transient and equilibrium responses arises from the substantial retreat of Antarctic sea ice and subsequent ocean warming. Unlike SLP, the changes in SAT do not project substantially onto any of the dominant patterns of variability. Changes appear to project strongly onto the ENSO-like mode, but in fact ~90% of this projection relates to the mean global warming associated with the mode, while only ~10% relates to the actual pattern.

In all cases examined the projection of climate change overwhelmingly manifests itself as a linear trend in the mode, with no important alteration of its behaviour. These results also demonstrate that recent observational studies supporting the nonlinear projection interpretation may instead be indicative of a linear projection of climate change onto the dominant modes.

On the sensitivity of projected oceanic thermal expansion to the parameterisation of sub-grid scale ocean mixing

Andrew J. Weaver and Edward C. Wiebe

A coupled model of intermediate complexity is used to examine the importance of the parameterisation of sub-grid scale ocean mixing on the global mean steric sea level rise in global warming simulations. It is shown that when mixing associated with meoscale eddies is treated in a more physically realistic way than the commonly used horizontal/vertical scheme, quasi-equilibrium projected steric sea level rise is more than two times lower in both 2 x CO₂ and 4 x CO₂ climates. This occurs despite the invariance of the coupled model climate sensitivity to the particular sub-grid scale mixing scheme employed. During the early phase of the transient integrations thermal expansion differences are smaller, although experiments using the Gent and McWilliams parameterisation for mixing associated with mesoscale eddies approach equilibrium more rapidly once the radiative forcing is held fixed. This reduced expansion commitment reflects a greater decoupling of the surface ocean from the deep ocean, due to a reduction in spurious high latitude convection that occurs when a horizontal/vertical mixing scheme is used.

Trends in Canadian precipitation intensity.

Daithi A. Stone, Andrew J. Weaver and Francis W. Zwiers

Past research has unveiled important variations in mean precipitation, often related to large scale shifts in atmospheric circulation, and consistent with projected responses to enhanced global warming. More recently, however, it has been realised that important and influential changes in the variability of daily precipitation events have also occurred in the past, often unrelated to changes in mean accumulation.

This study aims to uncover variations in precipitation event intensity over Canada and to compare the observed variations with those in mean accumulation and two dominant modes of atmospheric variability, namely the North Atlantic Oscillation (NAO) and the Pacific/North America teleconnection pattern (PNA). Results are examined on both annual and seasonal bases, and with regions defined by similarities in monthly variability.

Seasonally increasing trends are found in southern areas of Canada that result from increases in all levels of event intensity during the 20th century. During the latter half of the century increases are concentrated in heavy and intermediate events, with the largest changes occurring in Arctic areas. Variations in precipitation intensity can, however, be unrelated to variations in the mean accumulation. Consistent with these differences, the precipitation responses to the NAO and PNA are found to often occur in northeastern regions in summer and winter with the intensity affected in both seasons. The PNA strongly influences precipitation in many regions of the country during autumn and winter. In particular, it strongly influences variations in southern British Columbia and the Prairies, affecting the intensity in only some areas. However, it only influences the frequency of heavier events in autumn and winter in Ontario and southern Quebec, where this response is actually more robust than the response in total accumulation. During these seasons a negative PNA generally leads to more extreme precipitation events.

Simulating the ice-thickness distribution in a coupled climate model.

Cecilia M. Bitz, Marika M. Holland, Andrew J. Weaver and Mike Eby

Climate simulations in a global coupled model are investigated using a dynamic-thermodynamic sea ice and snow model with sophisticated thermodynamics and a sub-grid scale parameterization for multiple ice thicknesses. In addition to the sea ice component, the model includes a full primitive-equation ocean and a simple energy-moisture balance atmosphere. We introduce a new formulation of the ice thickness distribution that is Lagrangian in thickness-space. The method is designed to use fewer thickness categories because it adjusts to place resolution where it is needed most, and it is free of diffusive effects that tend to smooth Eulerian distributions. Simulations demonstrate that the model does reasonably well in simulating the mean arctic climate. Compared to simulations without an ice-thickness distribution, we find widespread changes in the Arctic and northern North Atlantic climate. The ice-thickness distribution causes ice export through Fram Strait to be more variable and more strongly linked to meridional overturning in the North Atlantic Ocean.

The Langrangian formulation of the ice thickness distribution allows for the inclusion of a vertical temperature profile with relative ease compared to an Eulerian method. We find ice growth rates and ocean surface salinity differ in our model with a well resolved vertical temperature profile in the ice and snow and an explicit brine-pocket parameterization compared to a simulation with Semtner zero-layer thermodynamics. Although these differences are important for the climate of the Arctic, the effect of an ice thickness distribution are more dramatic and have further reaching consequences for the Northern Hemisphere. Sensitivity experiments indicate that five ice thickness categories with ~ 50 cm vertical temperature resolution captures the effects of the ice thickness distribution on the heat and freshwater exchange across the surface in the presence of sea ice in climate simulations

Glacial termination: Sensitivity to orbital and CO₂ forcing in a coupled climate system model.

Masakazu Yoshimori, Andrew J. Weaver, Shawn J. Marshall and Garry K. C. Clarke

To study glacial terminations and related feedback mechanisms, a continental ice dynamics model was globally and asynchronously coupled to a physical climate (atmosphere-ocean-sea ice) model. The model performs reasonably well under present-day, 11kaBP (thousand years before present) and 21kaBP perpetual forcing. To address the ice sheet response under the effects of both orbital and CO₂ forcing, sensitivity experiments are conducted with two different orbital configurations (11kaBP and 21kaBP) and two different CO₂ concentrations (200ppmv and 280ppmv). This study reveals that, although both orbital and CO₂ forcing have an impact on ice sheet maintenance and deglacial processes, and although neither acting alone is sufficient to lead to complete deglaciation, orbital forcing seems to be more important. The lowered CO₂ has a large effect on climate, not uniformly or zonally over the globe, but concentrated over the continents adjacent to the North Atlantic. The impact of CO₂ forcing on surface air temperature there has its peak in winter associated with changes in sea ice cover and thickness in the northern North Atlantic. This is accompanied by the annual mean reduction in the intensity of the meridional overturning and poleward ocean heat transport in the North Atlantic. On the other hand, the impact of orbital forcing has its peak in summer at 11kaBP and 21kaBP. Since the summer temperature, rather than winter temperature, is found to be dominant for the ice sheet mass balance, orbital forcing (the difference between 11kaBP and 21kaBP) has a larger effect than CO₂ forcing (the difference between 200ppmv and 280ppmv) in deglaciation. Also, warm winter sea surface temperature arising from increased CO₂ during the deglaciation contributes to ice sheet nourishment through slightly enhanced precipitation.

The Canada Basin 1989-1995: Upstream events and far-field effects of the Barents Sea branch

Fiona McLaughlin, Ed Carmack, Robbie Macdonald, Andrew J. Weaver and J. Smith

Physical and geochemical tracer measurements were collected at one oceanographic station (Station A: 72 N 143W) in the southern Canada Basin from 1989 to 1995, along sections from the Beaufort Shelf to this station in 1993 and 1995, and along a section westward of Banks Island in 1995. These measurements were examined to see how recent events in three upstream Arctic Ocean sub-basins impacted upon Canada Basin waters. Upstream events included Atlantic layer warming, relocation of the Atlantic/Pacific water mass boundary, and increased ventilation of boundary current waters. Early signals of change appeared first in the Canada Basin in 1993 along the continental margin and, by 1995, were evident at Station A in the basin interior and farther downstream. Differences in physical and geochemical properties (nutrients, oxygen, 129l and CFCs) were observed throughout much of the water column to depths greater than 1600 m. In particular, the boundary distinguishing Pacific from Atlantic-origin water was found to be shallower and Atlantic-origin water occupied more of the Canada Basin water column. By 1995, Atlantic-origin water in the lower halocline at Station A was found to be colder and more ventilated. Likewise, within the Atlantic layer, Fram Strait Branch (FSB) water was colder, fresher, and more ventilated, and Barents Sea Branch (BSB) water was warmer, fresher, and more ventilated than during previous years. By comparing observations at Station A with eastern Nansen Basin observations, the main source of these changes was traced to dense water outflow from the Barents Sea. Studies indicated that in early 1989 Barents Sea waters were 2 Deg C warmer and that between 1988 and 1989, a large volume of dense water had left the shelf. These events coincided with an atmospheric shift to increased cyclonic circulation in 1989, a transition unprecedented in its magnitude, geographic reach, and apparent oceanographic impact. The effects of a large outflow of dense Barents Sea water were observed some 5000 km away downstream in the Canada Basin where the BSB component of the Atlantic layer had increased 20% by 1995.

Climate model simulations of effects of increased atmospheric CO₂ and loss of sea ice on ocean salinity and tracer uptake

Phil B. Duffy, Michael Eby and Andrew J. Weaver

Recent observations show a decrease in the extent of Northern Hemisphere sea ice; this decrease has been attributed to human activities. We present climate model simulations which examine how loss of sea ice affects the ocean salinity and density structure, and rates of uptake of an idealized transient tracer. The latter results are indicative of how loss of sea ice might affect the ocean's rate of uptake of anthropogenic carbon from the atmosphere. In simulations in which there is no fresh water forcing due to sea ice forming or melting, the salinity minimum associated with Antarctic Intermediate Water (AAIW) is much weaker than in simulations of the present-day ocean. This suggests that this salinity minimum is maintained in part by a steady supply of fresh water from melting of Antarctic sea ice. In addition, in our simulations with no fresh water forcing due to sea ice, vertical salinity and density gradients in the Southern and Arctic Oceans are weaker than in simulations of the present-day ocean. This supports the notion that that these gradients are maintained in part by fresh water forcing due to the seasonal cycle of formation and melting of sea ice. As a result, loss of sea ice due to global warming would tend to decrease the stability in parts of the ocean; this opposes the well-known tendency of global warming to increase ocean stability by warming and freshening the upper ocean. We perform simulations of ocean uptake of an idealized transient tracer in both constant-CO₂ and increasing-CO₂ environments to investigate effects of physical changes in ocean and sea ice on transient tracer uptake. In the Southern Ocean, we find that physical changes to the ocean and sea ice result in slower transient tracer accumulation in most locations. When averaged over the entire Southern Ocean, however, these reductions are small, because changes in convective activity due to increased atmospheric CO₂ are relatively small, and because transient tracer uptake is relatively insensitive to changes in convective activity. These results suggest that Southern Ocean uptake of anthropogenic CO₂ may decrease less than previously supposed as global warming progresses.

Gender differences in introductory atmospheric and oceanic science exams: Multiple choice versus constructed response questions

Andrew J. Weaver and Helen Raptis

An analysis of 295 male and 194 female examinations from introductory atmospheric and oceanic science courses, offered at the University of Victoria since 1994, is conducted to determine whether or not their exists gender differences in the performance on multiple choice versus constructed response sections of the exams. The difference in the mean performance of males and females on constructed response relative to multiple choice question sections of final exams, even in years where the females performed better than or worse than the males on both sections, is on average 5% which is significant at the 0.1% level. Gender differences on time-limited midterm exams are not significant. It is further shown that final exam performance is not significantly related to whether or not the exam starts with a multiple choice versus constructed response set of questions. While our analysis is unable to differentiate between the possibilities that multiple choice questions favour male students and the competing hypothesis that constructed response questions favour female students, existing literature is reviewed to suggest that a combination of both is possible. Nevertheless, from the analysis of our examination results, we can conclude that an exam of introductory atmospheric or oceanic science curricula, which is made up of 60% multiple choice questions and 40% constructed response questions, would not be skewed to favour any particular gender.

Dependence of multiple climate states on ocean mixing parameters

Andreas Schmittner and Andrew J. Weaver

Multiple equilibria of the climate system, inferred from paleo reconstructions, have also been observed in both ocean-only and coupled ocean-atmosphere general circulation models. These multiple states are thought to be associated with different modes of operation of the conveyor in the North Atlantic. It has recently been suggested that the stability of these states depends on the amount of vertical mixing in ocean models. Here we investigate the dependence of the hysteresis behaviour of the thermohaline circulation to sub-gridscale mixing processes in the ocean. Using a simplified coupled ocean-atmosphere model we find that both vertical and horizontal diffusivities have considerable influence on the stability of the different circulation modes. They also change the transition points from one circulation pattern to another. Larger vertical diffusivities lead to higher values of additional precipitation into the North Atlantic being necessary to stop the formation of deep water. However, for slightly increased evaporation, the state without deep water formation becomes increasingly unstable for stronger vertical diffusion. Larger values of horizontal mixing lead to a narrowing of the phase space for which two equilibria are stable. These results suggest that it is currently not possible, given the large uncertainty in ocean mixing, to quantitatively determine possible thresholds for the transition of the North Atlantic thermohaline circulation between on and off modes. This presents a policy predicament as it makes it extremely difficult to assign a probability to such a potentially important future nonlinear climate transition.

The influence of sea ice physics on simulations of climate change

Marika M. Holland, Cecilia M. Bitz, and Andrew J. Weaver

Sea ice cover is an important factor in the climate system due to feedback mechanisms associated with its influence on the surface albedo and ice-ocean-atmosphere exchange. However, sea ice models in GCMs typically used relatively crude physics. Single column and basin scale ice models have attempted to assess the importance of different physical parameterizations, however, this has often been done in uncoupled systems which means that various coupled feedback mechanisms are missing.

In this study, we examine the sensitivity of climate change simulations in a global coupled ice-ocean-atmosphere model to different sea ice physics. In particular, the influences of ice dynamics and a sub-gridscale ice thickness distribution are addressed. The importance of these parameterizations for the simulation of present-day climate conditions and the climate response to increasing atmospheric CO₂ levels is discussed. Additionally, we examine the influence of the albedo feedback mechanism in climate change experiments.

As in several previous studies, we find that the sea-ice parameterizations have a significant influence on present-day climate simulations, modifying both the annual mean ice-ocean-atmosphere conditions and the seasonal variation of these properties. For example, in models with motionless sea ice (i.e., thermodynamic-only models) the ice volume increases significantly and undergoes a smaller seasonal cycle. Resolving the ice thickness distribution also increases the ice thickness, but acts to enhance the seasonal cycle. Additionally, the ocean circulation is modified due to different ice/ocean buoyancy fluxes, leading to different Antarctic Bottom Water formation rates.

The presence of ice dynamics and the sub-gridscale ice thickness distribution also influences the response of the system to climate perturbations. Under increased atmospheric CO₂ forcing, simulating ice dynamics and the ice thickness distribution enhances the ice area response. However, the ice volume response is diminished when ice dynamics in included and enhances when the ice thickness distribution is resolved. The ocean response to global warming is also modified due to the changes in ice physics and the thermohaline circulation is less sensitive to climate change scenarios in models that resolve ice dynamics and the ice thickness distribution.

Additional simulations were performed to quantify the influence of the albedo feedback mechanism on climate change simulations. In increasing CO₂ simulations, which neglected the influence of a changing surface albedo, amplified warming was still present (although reduced) at high latitudes due to the poleward retreat of the ice cover and larger ocean-atmosphere heat exchange. In these simulations, the albedo feedback has a considerable influence on the climate response to global warming, accounting for 17% of the global air temperature increase, 37% of the Northern Hemisphere ice area decrease and 31% of the Northern Hemisphere ice volume decrease.

Earth System Models of Intermediate Complexity: Closing the Gap in the Spectrum of Climate System Models

Claussen, M., L. A. Mysak, A. J. Weaver, M. Crucifix, T. Fichefet, M.-F. Loutre, S.L. Weber, J. Alcamo, V.A. Alexeev, A. Berger, R. Calov, A. Ganopolsky, H. Goosse, G. Lohman, F. Lunkeit, I.I. Mohkov, V. Petoukhov, P. Stone, and Z. Wang

We propose a new perspective on the hierarchy of climate models which goes beyond the "classical" climate modeling pyramid that is restricted mainly to atmospheric processes. Most notably, we introduce a new indicator, called "integration", which characterizes the number of interacting components of the climate system being explicitly described in a model. The location of several model types, from conceptual to comprehensive, is presented in a new spectrum of climate system models. In particular, the location of the so-called Earth system Models of Intermediate Complexity (EMICS) in this spectrum is discussed in some detail and examples are given, which indicate that there is currently a broad range of EMICs in use. In some EMICS, the number of processes and/or the detail of description is reduced for the sake of simulating the feedbacks between as many components of the climate system as feasible. Others, with a lesser degree of interaction, are used for long-term ensemble simulations to study specific aspects of climate variability. EMICs appear to be closer to comprehensive coupled models of atmospheric and oceanic circulation (CGCMS) than to "conceptual" or "box" models. We advocate that EMICs be considered as complementary to CGCMs and conceptual models, because we believe that there is an advantage of having a spectrum of climate system models which are designed to tackle specific aspects of climate and which together provide the proper tool for climate system modeling.

Forcing of the deep ocean circulation in simulations of the Last Glacial Maximum

A. Schmittner, K. J. Meissner, M. Eby and A. J. Weaver

From the interpretation of different proxy data it is widely believed that the North Atlantic thermohaline circulation during the maximum of the last ice age about 21,000 years ago was considerably weaker than today. Recent equilibrium simulations with a coupled ocean - atmosphere - sea ice model successfully simulated a reduction in North Atlantic Deep Water (NADW) formation consistent with reconstructions. Here we examine the influence of different air - sea fluxes on simulated changes in the deep ocean circulation between the Last Glacial Maximum and present day. We find that changes in the oceanic surface freshwater fluxes are the dominant forcing mechanism for the reduced Atlantic overturning. Diminished export of freshwater out of the Atlantic drainage basin through the atmosphere, decreases surface salinities in the North Atlantic, leading to less NADW formation in the colder climate. Changes in heat fluxes, which lead to increased sea surface densities in the North Atlantic and therefore to an enhanced overturning, are of secondary importance. Wind stress variations seem to play a negligible role.

The degree to which the Atlantic freshwater export and hence the NADW formation is reduced depends on the formulation of the atmospheric hydrological cycle and on the strength of the overturning in the present day simulation. Simulated changes in sea surface properties for a large variety of overturning strengths are compared with different reconstruction datasets. The results depend strongly on the dataset used. Sea surface temperature reconstructions from CLIMAP and earlier salinity reconstructions based on planktonic foraminifera are most consistent with a significant reduction of the circulation, while recent reconstructions using dinocyst assemblages allow no unequivocal conclusion.