Kinetic energy analysis of the response of the Atlantic meridional overturning circulation to CO2-forced climate change

Atmosphere?Cocean general circulation models (AOGCMs) predict a weakening of the Atlantic meridional overturning circulation (AMOC) in response to anthropogenic forcing of climate, but there is a large model uncertainty in the magnitude of the predicted change. The weakening of the AMOC is generally understood to be the result of increased buoyancy input to the north Atlantic in a warmer climate, leading to reduced convection and deep water formation. Consistent with this idea, model analyses have shown empirical relationships between the AMOC and the meridional density gradient, but this link is not direct because the large-scale ocean circulation is essentially geostrophic, making currents and pressure gradients orthogonal. Analysis of the budget of kinetic energy (KE) instead of momentum has the advantage of excluding the dominant geostrophic balance. Diagnosis of the KE balance of the HadCM3 AOGCM and its low-resolution version FAMOUS shows that KE is supplied to the ocean by the wind and dissipated by viscous forces in the global mean of the steady-state control climate, and the circulation does work against the pressure-gradient force, mainly in the Southern Ocean. In the Atlantic Ocean, however, the pressure-gradient force does work on the circulation, especially in the high-latitude regions of deep water formation. During CO₂-forced climate change, we demonstrate a very good temporal correlation between the AMOC strength and the rate of KE generation by the pressure-gradient force in 50?C70°N of the Atlantic Ocean in each of nine contemporary AOGCMs, supporting a buoyancy-driven interpretation of AMOC changes. To account for this, we describe a conceptual model, which offers an explanation of why AOGCMs with stronger overturning in the control climate tend to have a larger weakening under CO₂ increase.     

Downscaling of South America present climate driven by 4-member HadCM3 runs

The objective of this work is to evaluate climate simulations over South America using the regional Eta Model driven by four members of an ensemble of the UK Met Office Hadley Centre HadCM3 global model. The Eta Model has been modified with the purpose of performing long-term decadal integrations and has shown to reproduce “present climate”—the period 1961–1990—reasonably well when forced by HadCM3. The global model lateral conditions with a resolution of 2.5° latitude?×?3.75° longitude were provided at a frequency of 6?h. Each member of the global model ensemble has a different climate sensitivity, and the four members were selected to span the range of uncertainty encompassed by the ensemble. The Eta Model nested in the HadCM3 global model was configured with 40-km horizontal resolution and 38 layers in the vertical. No large-scale internal nudging was applied. Results are shown for austral summer and winter at present climate defined as 1961–90. The upper and low-level circulation patterns produced by the Eta-CPTEC/HadCM3 experiment set-up show good agreement with reanalysis data and the mean precipitation and temperature with CRU observation data. The spread in the downscaled mean precipitation and temperature is small when compared against model errors. On the other hand, the benefits in using an ensemble is clear in the improved representation of the seasonal cycle by the ensemble mean over any one realization. El Ni?o and La Ni?a years were identified in the HadCM3 member runs based on the NOAA Climate Prediction Center criterion of sea surface temperature anomalies in the Ni?o 3.4 area. The frequency of the El Ni?o and La Ni?a events in the studied period is underestimated by HadCM3. The precipitation and temperature anomalies typical of these events are reproduced by most of the Eta-CPTEC/HadCM3 ensemble, although small displacements of the positions of the anomalies occur. This experiment configuration is the first step on the implementation of Eta-CPTEC/HadCM3 upcoming experiments on climate change studies that are discussed in a companion paper.     

A method for obtaining pre-twentieth century initial conditions for use in climate change studies

A method is proposed to initialise coupled atmosphere-ocean general circulation models (AOGCMs) developed to study climate change on multi-century time scales. The method assumes that current generation AOGCMs are developed and evaluated using present-day radiative forcing and near present day oceanic initial conditions. To find pre-twentieth century initial conditions, we propose that the radiative forcing be run backwards in time from the present to the desired starting date. The model should then be run for 3–5 centuries with the radiative forcing held constant at the desired date. In our tests, instantaneously switching to pre-twentieth century radiative forcing did not save computational time. When a sufficiently stable pre-twentieth century condition is achieved, the coupled system can be integrated forward to the present and into the future. This method is a first step toward the standardization of AOGCM initialization and suggests a framework for AOGCM initialization for the first time. It provides an internally consistent set of pre-twentieth century initial conditions, although they will vary from model to model. Furthermore, it is likely that this method will yield a fairly realistic present-day climate in transient climate change experiments of the twentieth century, if the model biases are not too large. The main disadvantage of the method is that it is fairly computationally expensive in that it requires an additional 4–6 centuries of model integration before starting historical twentieth century integrations. However, the relative cost of this technique diminishes as more simulations are conducted using the oceanic initial condition obtained using our method.     

Downscaled simulations of the ECHAM5, CCSM3 and HadCM3 global models for the eastern Mediterranean–Black Sea region: evaluation of the reference period

The outputs of three GCMs, ECHAM5, CCSM3 and HadCM3, are downscaled for the eastern Mediterranean–Black Sea region for the period 1961–1990 using a regional climate model, RegCM3, to assess the capability of these models in simulating the climatology of the region. In addition, the NCEP/NCAR Reanalysis data are also downscaled for the same period to display the performance of the regional climate model for the same region, which constitutes a relatively complex terrain and rich variety of climates. The gridded observational dataset of CRU is primarily used in the evaluation of the models, however, a regional dataset, which is based on a relatively dense gauging network, is also used to see how it affects the performance measures of the models. The reanalysis simulation indicates that RegCM3 is able to simulate the precipitation and surface temperature as well as the upper level fields reasonably well. However, it tends to overestimate the precipitation over the mountainous areas. All three GCM models are found to be highly skilled in simulating the winter precipitation and temperature in the region. The two models, ECHAM5 and HadCM3, are also good at simulating the summer precipitation and temperature, but the CCSM3 simulation generates dryer and warmer conditions than the observations for the whole region, which are most likely a result of the dryness in the upper levels of the original outputs. The use of the regional observational dataset does not necessarily improve the pattern correlations, but it yields better match between the modeled and observed precipitation in terms of variability and root-mean-square difference. It could be said that the outputs of these GCMs can be used in the climate change downscaling and impact assessment studies for the region, given that their strengths and weaknesses that are displayed in the present study are considered.     

The variation of ENSO characteristics associated with atmospheric parameter perturbations in a coupled model

We analyse the differences in the properties of the El Niño Southern Oscillation (ENSO) in a set of 17 coupled integrations with the flux-adjusted, 19-level HadCM3 model with perturbed atmospheric parameters. Within this ensemble, the standard deviation of the NINO3.4 deseasonalised SSTs ranges from 0.6 to 1.3 K. The systematic changes in the properties of the ENSO with increasing amplitude confirm that ENSO in HadCM3 is prevalently a surface (or SST) mode. The tropical-Pacific SST variability in the ensemble of coupled integrations correlates positively with the SST variability in the corresponding ensemble of atmosphere models coupled with a static mixed-layer ocean (“slab” models) perturbed with the same changes in atmospheric parameters. Comparison with the respective coupled ENSO-neutral climatologies and with the slab-model climatologies indicates low-cloud cover to be an important controlling factor of the strength of the ENSO within the ensemble. Our analysis suggests that, in the HadCM3 model, increased SST variability localised in the south-east tropical Pacific, not originating from ENSO and associated with increased amounts of tropical stratocumulus cloud, causes increased ENSO variability via an atmospheric bridge mechanism. The relationship with cloud cover also results in a negative correlation between the ENSO activity and the model’s climate sensitivity to doubling CO₂.     

Using the Twentieth Century Reanalysis to assess climate variability for the European wind industry

We characterise the long-term variability of European near-surface wind speeds using 142 years of data from the Twentieth Century Reanalysis (20CR), and consider the potential of such long-baseline climate data sets for wind energy applications. The low resolution of the 20CR would severely restrict its use on its own for wind farm site-screening. We therefore perform a simple statistical calibration to link it to the higher-resolution ERA-Interim data set (ERAI), such that the adjusted 20CR data has the same wind speed distribution at each location as ERAI during their common period. Using this corrected 20CR data set, wind speeds and variability are characterised in terms of the long-term mean, standard deviation and corresponding trends. Many regions of interest show extremely weak trends on century timescales, but contain large multidecadal variability. Since reanalyses such as ERAI are often used to provide the background climatology for wind farm site assessments, but contain only a few decades of data, our results can be used as a way of incorporating decadal-scale wind climate variability into such studies, allowing investment risks for wind farms to be reduced.     

Simulating the diurnal cycle of rainfall in global climate models: resolution versus parameterization

The effects of horizontal resolution and the treatment of convection on simulation of the diurnal cycle of precipitation during boreal summer are analyzed in several innovative weather and climate model integrations. The simulations include: season-long integrations of the Non-hydrostatic Icosahedral Atmospheric Model (NICAM) with explicit clouds and convection; year-long integrations of the operational Integrated Forecast System (IFS) from the European Centre for Medium-range Weather Forecasts at three resolutions (125, 39 and 16 km); seasonal simulations of the same model at 10 km resolution; and seasonal simulations of the National Center for Atmospheric Research (NCAR) low-resolution climate model with and without an embedded two-dimensional cloud-resolving model in each grid box. NICAM with explicit convection simulates best the phase of the diurnal cycle, as well as many regional features such as rainfall triggered by advancing sea breezes or high topography. However, NICAM greatly overestimates mean rainfall and the magnitude of the diurnal cycle. Introduction of an embedded cloud model within the NCAR model significantly improves global statistics of the seasonal mean and diurnal cycle of rainfall, as well as many regional features. However, errors often remain larger than for the other higher-resolution models. Increasing resolution alone has little impact on the timing of daily rainfall in IFS with parameterized convection, yet the amplitude of the diurnal cycle does improve along with the representation of mean rainfall. Variations during the day in atmospheric prognostic fields appear quite similar among models, suggesting that the distinctive treatments of model physics account for the differences in representing the diurnal cycle of precipitation.     

An efficient climate forecasting method using an intermediate complexity Earth System Model and the ensemble Kalman filter

We present the implementation and results of a model tuning and ensemble forecasting experiment using an ensemble Kalman filter for the simultaneous estimation of 12 parameters in a low resolution coupled atmosphere-ocean Earth System Model by tuning it to realistic data sets consisting of Levitus ocean temperature/salinity climatology, and NCEP/NCAR atmospheric temperature/humidity reanalysis data. The resulting ensemble of tuned model states is validated by comparing various diagnostics, such as mass and heat transports, to observational estimates and other model results. We show that this ensemble has a very reasonable climatology, with the 3-D ocean in particular having comparable realism to much more expensive coupled numerical models, at least in respect of these averaged indicators. A simple global warming experiment is performed to investigate the response and predictability of the climate to a change in radiative forcing, due to 100 years of 1% per annum atmospheric CO₂ increase. The equilibrium surface air temperature rise for this CO₂ increase is 4.2±0.1°C, which is approached on a time scale of 1,000 years. The simple atmosphere in this version of the model is missing several factors which, if included, would substantially increase the uncertainty of this estimate. However, even within this ensemble, there is substantial regional variability due to the possibility of collapse of the North Atlantic thermohaline circulation (THC), which switches off in more than one third of the ensemble members. For these cases, the regional temperature is not only 3–5°C colder than in the warmed worlds where the THC remains switched on, but is also 1–2°C colder than the current climate. Our results, which illustrate how objective probabilistic projections of future climate change can be efficiently generated, indicate a substantial uncertainty in the long-term future of the THC, and therefore the regional climate of western Europe. However, this uncertainty is only apparent in long-term integrations, with the initial transient response being similar across the entire ensemble. Application of this ensemble Kalman filtering technique to more complete climate models would improve the objectivity of probabilistic forecasts and hence should lead to significantly increased understanding of the uncertainty of our future climate.     

High-resolution simulations of West African climate using regional climate model (RegCM3) with different lateral boundary conditions

To downscale climate change scenarios, long-term regional climatologies employing global model forcing are needed for West Africa. As a first step, this work examines present-day integrations (1981–2000) with a regional climate model (RCM) over West Africa nested in both reanalysis data and output from a coupled atmospheric–ocean general circulation model (AOGCM). Precipitation and temperature from both simulations are compared to the Climate Research Unit observations. Their spatial distributions are shown to be realistic. Annual cycles are considerably correlated. Simulations are also evaluated with respect to the driving large-scale fields. RCM offers some improvements compared to the AOGCM driving field. Evaluation of seasonal precipitation biases reveals that RCM dry biases are highest on June–August around mountains. They are associated to cold biases in temperature which, in turn, are connected to wet biases in precipitation outside orographic zones. Biases brought through AOGCM forcing are relatively low. Despite these errors, the simulations produce encouraging results and show the ability of the AOGCM to drive the RCM for future projections.     

Frequency distributions of transient regional climate change from perturbed physics ensembles of general circulation model simulations

Large ensembles of coupled atmosphere–ocean general circulation model (AOGCM) simulations are required to explore modelling uncertainty and make probabilistic predictions of future transient climate change at regional scales. These are not yet computationally feasible so we have developed a technique to emulate the response of such an ensemble by scaling equilibrium patterns of climate change derived from much cheaper “slab” model ensembles in which the atmospheric component of an AOGCM is coupled to a mixed-layer ocean. Climate feedback parameters are diagnosed for each member of a slab model ensemble and used to drive an energy balance model (EBM) to predict the time-dependent response of global surface temperature expected for different combinations of uncertain AOGCM parameters affecting atmospheric, land and sea-ice processes. The EBM projections are then used to scale normalised patterns of change derived for each slab member, and hence emulate the response of the relevant atmospheric model version when coupled to a dynamic ocean, in response to a 1% per annum increase in CO₂. The emulated responses are validated by comparison with predictions from a 17 member ensemble of AOGCM simulations, constructed from variants of HadCM3 using the same parameter combinations as 17 members of the slab model ensemble. Cross-validation permits estimation of the spatial and temporal dependence of emulation error, and also allows estimation of a correction field to correct discrepancies between the scaled equilibrium patterns and the transient response, reducing the emulation error. Emulated transient responses and their associated errors are obtained from the slab ensemble for 129 pseudo-HadCM3 versions containing multiple atmospheric parameter perturbations. These are combined to produce regional frequency distributions for the transient response of annual surface temperature change and boreal winter precipitation change. The technique can be extended to any surface climate variable demonstrating a scaleable, approximately linear response to forcing.     

Climate entropy budget of the HadCM3 atmosphere–ocean general circulation model and of FAMOUS, its low-resolution version

The entropy budget is calculated of the coupled atmosphere–ocean general circulation model HadCM3. Estimates of the different entropy sources and sinks of the climate system are obtained directly from the diabatic heating terms, and an approximate estimate of the planetary entropy production is also provided. The rate of material entropy production of the climate system is found to be ～50 mW m^?2 K^?1, a value intermediate in the range 30–70 mW m^?2 K^?1 previously reported from different models. The largest part of this is due to sensible and latent heat transport (～38 mW m^?2 K^?1). Another 13 mW m^?2 K^?1 is due to dissipation of kinetic energy in the atmosphere by friction and Reynolds stresses. Numerical entropy production in the atmosphere dynamical core is found to be about 0.7 mW m^?2 K^?1. The material entropy production within the ocean due to turbulent mixing is ～1 mW m^?2 K^?1, a very small contribution to the material entropy production of the climate system. The rate of change of entropy of the model climate system is about 1 mW m^?2 K^?1 or less, which is comparable with the typical size of the fluctuations of the entropy sources due to interannual variability, and a more accurate closure of the budget than achieved by previous analyses. Results are similar for FAMOUS, which has a lower spatial resolution but similar formulation to HadCM3, while more substantial differences are found with respect to other models, suggesting that the formulation of the model has an important influence on the climate entropy budget. Since this is the first diagnosis of the entropy budget in a climate model of the type and complexity used for projection of twenty-first century climate change, it would be valuable if similar analyses were carried out for other such models.     

统计降尺度法对华北地区未来区域气温变化情景的预估

迄今为止,大部分海气耦合气候模式（AOGCM）的空间分辨率还较低,很难对区域尺度的气候变化情景做合理的预测。降尺度法已广泛用于弥补AOGCM在这方面的不足。作者采用统计降尺度方法对1月和7月华北地区49个气象观测站的未来月平均温度变化情景进行预估。采用的统计降尺度方法是主分量分析与逐步回归分析相结合的多元线性回归模型。首先,采用1961～2000年的 NCEP再分析资料和49个台站的观测资料建立月平均温度的统计降尺度模型,然后把建立的统计降尺度模型应用于HadCM3 SRES A2 和 B2 两种排放情景, 从而生成各个台站1950～2099年1月份和7月份温度变化情景。结果表明:在当前气候条件下,无论1月还是7月,统计降尺度方法模拟的温度与观测的温度有很好的一致性,而且在大多数台站,统计降尺度模拟气温与观测值相比略微偏低。对于未来气候情景的预估方面,无论1月还是7月,也无论是HadCM3 SRES A2 还是B2排放情景驱动统计模型,结果表明大多数的站点都存在温度的明显上升趋势,同时7月的上升趋势与1月相比偏低。     

Climate change scenarios for global impacts studies

We describe a set of global climate change scenarios that have been used in a series of studies investigating the global impacts of climate change on several environmental systems and resources — ecosystems, food security, water resources, malaria and coastal flooding. These scenarios derive from modelling experiments completed by the Hadley Centre over the last four years using successive versions of their coupled ocean–atmosphere global climate model. The scenarios benefit from ensemble simulations (made using HadCM2) and from an un-flux-corrected experiment (made using HadCM3), but consider only the effects of increasing greenhouse gas concentrations. The effects of associated changes in sulphate aerosol concentrations are not considered. The scenarios are presented for three future time periods — 30-year means centred on the 2020s, the 2050s and the 2080s — and are expressed with respect to the mean 1961–1990 climate. A global land observed climatology at 0.5° latitude/longitude resolution is used to describe current climate. Other scenario variables — atmospheric CO₂ concentrations, global-mean sea-level rise and non-climatic assumptions relating to population and economy — are also provided. We discuss the limitations of the created scenarios and in particular draw attention to sources of uncertainty that we have not fully sampled.     

Seasonal variability of the observed and the projected daily temperatures in northern Saudi Arabia

Variability in the observed daily temperature for the 31-year period (1978–2008) is studied for northern Saudi Arabia (nSA) by computing the probability distribution functions (PDFs) on a seasonal basis. The 31-year base period is divided into three decades and the results for the first (1978–1987) and the last decade (1999–2008) are presented. When averaged over all seasons, mean values of the observed decadal PDFs depict a positive shift from the first to last decade in the minimum, mean, and maximum temperature of 0.81 °C, 1.03 °C, and 1.25 °C, respectively. The daily temperature datasets from a regional climate model (RCM) and two versions of a coupled atmosphere-ocean general circulation model (AOGCM) are compared with the observed daily temperature datasets. The RCM is driven by re-analysis data for the historical period and by the HadCM3 model for the future, while the AOGCMs used are the GFDL CM2.0 and 2.1 models, with both HadCM3 and the GFDL simulations corresponding to the SRES A1B scenario. The average shifts from 1978–1987 to 1999–2008 in the mean value of the PDFs for the minimum, mean and maximum temperature are 0.63 °C, 0.54 °C and 0.45 °C, respectively, for the RCM, and 0.97 °C, 0.97 °C and 0.96 °C, respectively, for the AOGCM. Thus, the RCM shows a smaller shift in the mean of PDF for maximum temperature than for mean or minimum temperature, the AOGCM shows a comparable shift for all three, and the observations show a greater shift in the PDF for maximum temperature. For the period 2070–2099 relative to 1978–2008, the three average shifts are 4.11 °C, 3.87 °C and 3.44 °C for the RCM and 3.63 °C, 3.74 °C and 3.84 °C for the AOGCM.     

The ARPEGE/IFS atmosphere model: a contribution to the French community climate modelling

A new atmospheric model has been developed jointly by Météo-France, and the European Centre for Medium-range Weather Forecasts (ECMWF) under the acronyms ARPEGE (action de recherche petite echelle grande echelle, which means research project on small and large scales) and IFS (integrated forecast system). This model includes, inter alia, an atmospheric general circulation model (GCM) which is intended by the French climate modelling community to be used for studying the anthropogenic climate impact. A preliminary version of this model has been available since 1992. This paper describes its main characteristics. Three 10-year integrations of this model having spectral horizontal resolutions of T21, T42, and T79 have been performed using prescribed monthly mean sea surface temperatures (SST) observed from 1979 until 1988. The results of these integrations are presented and compared with the observed climatology. The comparison is made for the winter (DJF) and summer (JJA) periods. It is shown that the model is capable of reproducing the observed climatology in a generally successful manner.     

Evaluation of East Asian climatology as simulated by seven coupled models

Using observation and reanaiysis data throughout 1961-1990, the East Asian surface air temperature, precipitation and sea level pressure climatology as simulated by seven fully coupled atmosphere-ocean models, namely CCSR/NIES, CGCM2, CSIRO-Mk2, ECHAM4/OPYC3, GFDL-R30, HadCM3, and NCARPCM, axe systematically evaluated in this study. It is indicated that the above models can successfully reproduce the annual and seasonal surface air temperature and precipitation climatology in East Asia, with relatively good performance for boreal autumn and annual mean. The models‘ ability to simulate surface air temperature is more reliable than precipitation. In addition, the models can dependably capture the geographical distribution pattern of annual, boreal winter, spring and autumn sea level pressure in East Asia. In contrast, relatively large simulation errors axe displayed when simulated boreal summer sea level pressure is compaxed with reanalysis data in East Asia. It is revealed that the simulation errors for surface air temperature, precipitation and sea level pressure axe generally large over and around the Tibetan Plateau. No individual model is best in every aspect. As a whole, the ECHAM4/OPYC3 and HadCM3 performances axe much better, whereas the CGCM2 is relatively poorer in East Asia. Additionally, the seven-model ensemble mean usually shows a relatively high reliability.     

Climate of Russia in the 21st Century. Part 2. Verification of atmosphere-ocean general circulation models CMIP3 for projections of future climate changes

Results of simulation of radiation, cloud cover, surface air temperature, sea-level pressure, and hydrological regime components for Russia with the help of an ensemble of CMIP3 global climate models is analyzed. Despite a large spread among the models, the CMIP3 AOGCM ensemble simulations of the key characteristics of the observed surface climate agree well with observations, anyway in averaging over areas of vast regions, from watersheds of large rivers to the whole of Russia. These means (ensemble-and area-averaged values) often fall into the range of estimates derived from observations. This suggests the existence of uncertainty in the estimates obtained from simulations as well as from observational data. Comparison of different-generation models demonstrates a gradual improvement of the AOGCM simulation of surface climate characteristics. In general, the averaging over the CMIP3 AOGCM ensemble allows us to state that the ensemble is suitable for estimates of future climate changes.     

Multidecadal simulation of coastal fog with a regional climate model

In order to model stratocumulus clouds and coastal fog, we have coupled the University of Washington boundary layer model to the regional climate model, RegCM (RegCM-UW). By comparing fog occurrences observed at various coastal airports in the western United States, we show that RegCM-UW has success at modeling the spatial and temporal (diurnal, seasonal, and interannual) climatology of northern California coastal fog. The quality of the modeled fog estimate depends on whether coast-adjacent ocean or land grid cells are used; for the model runs shown here, the oceanic grid cells seem to be most appropriate. The interannual variability of oceanic northern California summertime fog, from a multi-decadal simulation, has a high and statistically significant correlation with the observed interannual variability (r = 0.72), which indicates that RegCM-UW is capable of investigating the response of fog to long-term climatological forcing. While RegCM-UW has a number of aspects that would benefit from further investigation and development, RegCM-UW is a new tool for investigating the climatology of coastal fog and the physical processes that govern it. We expect that with appropriate physical parameterizations and moderate horizontal resolution, other climate models should be capable of simulating coastal fog. The source code for RegCM-UW is publicly available, under the GNU license, through the International Centre for Theoretical Physics.     

Modelling the sensitivity of Mediterranean Outflow to anthropogenically forced climate change

 The possible future impact of anthropogenic forcing upon the circulation of the Mediterranean, and the exchange through the Strait of Gibraltar is investigated using a Cox-type model of the Mediterranean at 0.25° × 0.25° resolution, forced by “control” and “greenhouse” scenarios provided by the HadCM2 coupled climate model. The current structure of the Mediterranean forced by the “control” climate is compared with observations: certain aspects of the present circulation are reproduced, but others are absent or incorrectly represented. Deficiencies are most probably due to weaknesses in the forcing climatology generated by the climate model, so some caution must be exercised in interpreting the enhanced greenhouse simulation. Comparison of the control and greenhouse scenarios suggests that deep-water production in the Mediterranean may be reduced or cease in the relatively near future. The results also suggest that the Mediterranean outflow, may become warmer and more saline, but less dense, and hence shallower. The volume of the exchange at the Strait of Gibraltar seems to be relatively insensitive to future climate change, however. Our results indicate that a parameterisation of Gibraltar exchange and Mediterranean Outflow Water (MOW) production may be able to provide adequate representation of the changes we observe for the purposes of the current generation of climate models. Received: 10 August 1998 / Accepted: 11 October 1999