Climate Sensitivity Experiments Research Articles

Flux coupling approach on an exchange grid for the IOW Earth System Model (version 1.04.00) of the Baltic Sea region

Abstract. In this article the development of a high-resolution Earth System Model (ESM) for the Baltic Sea region is described. In contrast to conventional coupling approaches, the presented model features an additional (technical) component, the flux calculator, which calculates fluxes between the model components on a common exchange grid. This approach naturally ensures conservation of exchanged quantities, a locally consistent treatment of the fluxes, and facilitates interchanging model components in a straightforward manner. The main purpose of this model is to downscale global reanalysis or climate model data to the Baltic Sea region as typically, global model grids are too coarse to resolve the region of interest sufficiently. The regional ESM consists of the Modular Ocean Model 5 (MOM5) for the ocean and the COSMO model in CLimate Mode (CCLM, version 5.0_clm3) for the atmosphere. The bi-directional ocean–atmosphere coupling allows for a realistic air–sea feedback that outperforms the traditional approach of using uncoupled standalone models, as typically pursued with the EURO-CORDEX protocol. In order to address marine environmental problems (e.g., eutrophication and oxygen depletion), the ocean model is internally coupled with the marine biogeochemistry model, ERGOM, set up for the Baltic Sea's hydrographic conditions. The regional ESM can be used for various scientific questions such as climate sensitivity experiments, reconstruction of ocean dynamics, study of past climates, and natural variability, as well as investigation of ocean–atmosphere interactions. Therefore, it can serve for a better understanding of natural processes via attribution experiments that relate observed changes to mechanistic causes.

Potential for bias in effective climate sensitivity from state-dependent energetic imbalance

Abstract. To estimate equilibrium climate sensitivity from a simulation where a step change in carbon dioxide concentrations is imposed, a common approach is to linearly extrapolate temperatures as a function of top-of-atmosphere energetic imbalance to estimate the equilibrium state (“effective climate sensitivity”). In this study, we find that this estimate may be biased in some models due to state-dependent energetic leaks. Using an ensemble of multi-millennial simulations of climate model response to a constant forcing, we estimate equilibrium climate sensitivity through Bayesian calibration of simple climate models which allow for responses from subdecadal to multi-millennial timescales. Results suggest potential biases in effective climate sensitivity in the case of particular models where radiative tendencies imply energetic imbalances which differ between pre-industrial and quadrupled CO2 states, whereas for other models even multi-thousand-year experiments are insufficient to predict the equilibrium state. These biases draw into question the utility of effective climate sensitivity as a metric of warming response to greenhouse gases and underline the requirement for operational climate sensitivity experiments on millennial timescales to better understand committed warming following a stabilization of greenhouse gases.

Spatial and temporal patterns of snowmelt refreezing in a Himalayan catchment

AbstractRecent progress has been made in quantifying snowmelt in the Himalaya. Although the conditions are favorable for refreezing, little is known about the spatial variability of meltwater refreezing, hindering a complete understanding of seasonal snowmelt dynamics. This study aims to improve our understanding about how refreezing varies in space and time. We simulated refreezing with the seNorge (v2.0) snow model for the Langtang catchment, Nepalese Himalaya, covering a 5-year period. Meteorological forcing data were derived from a unique elaborate network of meteorological stations and high-resolution meteorological simulations. The results show that the annual catchment average refreezing amounts to 122 mm w.e. (21% of the melt), and varies strongly in space depending on elevation and aspect. In addition, there is a seasonal altitudinal variability related to air temperature and snow depth, with most refreezing during the early melt season. Substantial intra-annual variability resulted from fluctuations in snowfall. Daily refreezing simulations decreased by 84% (annual catchment average of 19 mm w.e.) compared to hourly simulations, emphasizing the importance of using sub-daily time steps to capture melt–refreeze cycles. Climate sensitivity experiments revealed that refreezing is highly sensitive to changes in air temperature as a 2°C increase leads to a refreezing decrease of 35%.

Quantifying COVID-19 enforced global changes in atmospheric pollutants using cloud computing based remote sensing

Global lockdowns in response to the COVID-19 pandemic have led to changes in the anthropogenic activities resulting in perceivable air quality improvements. Although several recent studies have analyzed these changes over different regions of the globe, these analyses have been constrained due to the usage of station based data which is mostly limited up to the metropolitan cities. Also the quantifiable changes have been reported only for the developed and developing regions leaving the poor economies (e.g. Africa) due to the shortage of in-situ data. Using a comprehensive set of high spatiotemporal resolution satellites and merged products of air pollutants, we analyze the air quality across the globe and quantify the improvement resulting from the suppressed anthropogenic activity during the lockdowns. In particular, we focus on megacities, capitals and cities with high standards of living to make the quantitative assessment. Our results offer valuable insights into the spatial distribution of changes in the air pollutants due to COVID-19 enforced lockdowns. Statistically significant reductions are observed over megacities with mean reduction by 19.74%, 7.38% and 49.9% in nitrogen dioxide (NO2), aerosol optical depth (AOD) and PM2.5 concentrations. Google Earth Engine empowered cloud computing based remote sensing is used and the results provide a testbed for climate sensitivity experiments and validation of chemistry-climate models. Additionally, Google Earth Engine based apps have been developed to visualize the changes in a real-time fashion.

Responses of soil organic carbon to climate change in the Qilian Mountains and its future projection

Soil is the largest organic carbon pool in the terrestrial biosphere, and a small variation of soil organic carbon (SOC) can substantially affect the global carbon cycle and climate. Climate change is a major driver affecting the dynamics of SOC; however, our understanding about the responses of SOC in alpine ecosystems to climate change is quite limited. In particular, the differences of SOC dynamics at different depths were rarely reported. In this study, we investigated the impacts of precipitation variations and warming on SOC dynamics at both top and deep soils in the Qilian Mountains in Northwestern China using a machine learning approach and climate sensitivity experiments. Our simulation revealed the temporal inconsistency between topsoil SOC (in the top 20 cm, denoted as SOC20) and deeper soil SOC dynamics—SOC20 showed a decreasing trend since 2009 which is earlier than that (2012) in the top 100 cm soil (SOC100). We also found that SOC100 may be more sensitive to warming due to the strengthened microbial decomposition rate and additional carbon source through deepened active layer. On the contrary, SOC20 presented more intense responses to precipitation than SOC100, which was mainly attributed to the different responses of upland and lowland SOC to precipitation variations. Our projection indicated that SOC20 may not substantially change under future climate trajectories because the enriched SOC induced by increased precipitation may offset the carbon loss via warming. However, SOC100 was projected to decrease significantly due to the enhanced carbon emissions via warming induced strengthened decomposition rate, additional carbon source from the deepened active layer, and exposed soil carbon to the atmosphere caused by ground subsidence and disrupted soil horizons resulting from thawed frozen soil. In brief, this study deepened our understanding of the mechanism of climate effect on SOC dynamics and can be helpful for regional soil ecological security assessment and risk projection.

Atmospheric teleconnection processes linking winter air stagnation and haze extremes in China with regional Arctic sea ice decline

Abstract. Recent studies suggested significant impacts of boreal cryosphere changes on wintertime air stagnation and haze pollution extremes in China. However, the underlying mechanisms of such a teleconnection relationship remains unclear. Here we use the Whole Atmosphere Community Climate Model (WACCM) to investigate dynamic processes leading to atmospheric circulation and air stagnation responses to Arctic sea ice changes. We conduct four climate sensitivity experiments by perturbing sea ice concentrations (SIC) and corresponding sea surface temperature (SST) in autumn and early winter over the whole Arctic and three subregions in the climate model. The results indicate distinct responses in circulation patterns and regional ventilation to the region-specific Arctic changes, with the largest increase of both the probability (by 132 %) and the intensity (by 30 %) of monthly air stagnation extremes being found in the experiment driven by SIC and SST changes over the Pacific sector of the Arctic (the East Siberian and Chukchi seas). The increased air stagnation extremes are mainly driven by an amplified planetary-scale atmospheric teleconnection pattern that resembles the negative phase of the Eurasian (EU) pattern. Dynamical diagnostics suggest that convergence of transient eddy forcing in the vicinity of Scandinavia in winter is largely responsible for the amplification of the teleconnection pattern. Transient eddy vorticity fluxes dominate the transient eddy forcing and produce a barotropic anticyclonic anomaly near Scandinavia and wave train propagation across Eurasia to the downstream regions in East Asia. The piecewise potential vorticity inversion analysis reveals that this long-range atmospheric teleconnection of Arctic origin takes place primarily via the middle and upper troposphere. The anomalous ridge over East Asia in the middle and upper troposphere worsens regional ventilation conditions by weakening monsoon northwesterlies and enhancing temperature inversions near the surface, leading to more and stronger air stagnation and pollution extremes over eastern China in winter. Ensemble projections based on state-of-the-art climate models in the Coupled Model Intercomparison Project Phase 6 (CMIP6) corroborate this teleconnection relationship between high-latitude environmental changes and midlatitude weather extremes, though the tendency and magnitude vary considerably among each participating model.

Warming Alters Hydrologic Heterogeneity: Simulated Climate Sensitivity of Hydrology‐Based Microrefugia in the Snow‐to‐Rain Transition Zone

AbstractIn complex terrain, drifting snow contributes to ecohydrologic landscape heterogeneity and ecological refugia. In this study, we assessed the climate sensitivity of hydrological dynamics in a semiarid mountainous catchment in the snow‐to‐rain transition zone. This catchment includes a distinct snow drift‐subsidized refugium that comprises a small portion (14.5%) of the watershed but accounts for a disproportionate amount (modeled average 56%) of hydrological flux generation. We conducted climate sensitivity experiments using a physically based hydrologic model to assess responses of a suite of hydrologic metrics across the watershed. Experiments with an imposed 3.5 °C warming showed reductions in average maximum snow water equivalent of 58–68% and deep percolation by 72%. While relative decreases were similar across the watershed, much greater absolute decreases in snowpack occurred in the drift‐subsidized site than the surrounding landscape. In drift‐subsidized locations, warming caused a shift from a regime that included both energy‐ and water‐limited evapotranspiration conditions to exclusively water‐limited conditions. Warming also resulted in altered interannual variability of hydrologic metrics. The drift‐subsidized unit was more sensitive to warming than the surrounding landscape, with reduced potential for the effects of warming to be offset by increased precipitation. Despite spatially homogeneous changes in climate, the effects of climate change on the hydrological dynamics were spatially heterogeneous in this watershed due to the presence of lateral water transport in the form of drifting snow. These findings suggest an increase in hydrologic homogeneity across the landscape and relatively large changes in snow drift‐subsidized refugia.

Simulating melt, runoff and refreezing on Nordenskiöldbreen, Svalbard, using a coupled snow and energy balance model

Abstract. A distributed energy balance model is coupled to a multi-layer snow model in order to study the mass balance evolution and the impact of refreezing on the mass budget of Nordenskiöldbreen, Svalbard. The model is forced with output from the regional climate model RACMO and meteorological data from Svalbard Airport. Extensive calibration and initialisation are performed to increase the model accuracy. For the period 1989–2010, we find a mean net mass balance of −0.39 m w.e. a−1. Refreezing contributes on average 0.27 m w.e. a−1 to the mass budget and is most pronounced in the accumulation zone. The simulated mass balance, radiative fluxes and subsurface profiles are validated against observations and are generally in good agreement. Climate sensitivity experiments reveal a non-linear, seasonally dependent response of the mass balance, refreezing and runoff to changes in temperature and precipitation. It is shown that including seasonality in climate change, with less pronounced summer warming, reduces the sensitivity of the mass balance and equilibrium line altitude (ELA) estimates in a future climate. The amount of refreezing is shown to be rather insensitive to changes in climate.

Extinguishing a Permian World

At the end of the Permian, ca. 252 Ma ago, marine and terrestrial fauna were facing the most extensive mass extinction in Earth history (Raup and Sepkoski, 1982). 80%–95% of all species on Earth, on land and in the oceans, became extinct (Benton et al., 2004) within an estimated time interval of less than 200 k.y. to 700 k.y. (Huang et al., 2011; Shen et al., 2011). Among the prominent Paleozoic animal groups that vanished are fusulinid foraminifera, rugose and tabulate corals, and the arthropod class Trilobita. The numerous hypotheses about the causes of the mass extinction include various environmental changes, mostly related to the emplacement of the Siberian Traps large igneous province. The compilation of radiometric U/Pb ages for the mass extinction and the Siberian Traps demonstrate a temporal overlap of both events (Svensen et al., 2009). A prominent hypothesis for the mass extinction is an accentuated global climate change scenario induced by volcanic CO 2 degassing (e.g., Svensen et al., 2009) that triggered biotic responses in the sea and on land. But what do we know about the climate at this time in Earth history? The two main features of climate recorded in the geological archives are temperature and humidity i.e., the moisture that is available for plant growth, and to a certain extent rainfall patterns. Climate simulations model the general circulation patterns during Permian–Triassic times. The paleogeography was characterized by the supercontinent Pangea extending nearly from pole to pole, with large land masses in the mid-latitudes of the Northern and Southern Hemispheres, and with the Tethys Ocean in the tropics (e.g., Smith et al., 1994). Climate models and sensitivity experiments demonstrated that this paleogeography provided the preconditions for a monsoonal circulation with strong seasonality of temperatures and rainfall on the Tethyan coasts, and distinct northern and southern intertropical convergence zones (Fig. 1) (e.g., Kutzbach and Ziegler, 1993; Parrish, 1993). Moist conditions prevailed in middle and high latitudes and along the

Climate Impacts of Making Evapotranspiration in the Community Land Model (CLM3) Consistent with the Simple Biosphere Model (SiB)

Abstract In recent climate sensitivity experiments with the Community Climate System Model, version 3 (CCSM3), a wide range of studies have found that the Community Land Model, version 3 (CLM3), simulates mean global evapotranspiration with low contributions from transpiration (15%), and high contributions from soil and canopy evaporation (47% and 38%, respectively). This evapotranspiration partitioning is inconsistent with the consensus of other land surface models used in GCMs. To understand the high soil and canopy evaporation and the low transpiration observed in the CLM3, select individual components of the land surface parameterizations that control transpiration, canopy and soil evaporation, and soil hydrology are compared against the equivalent parameterizations used in the Simple Biosphere Model, versions 2 and 3 (SiB2 and SiB3), and against more recent developments with CLM. The findings of these investigations are used to develop new parameterizations for CLM3 that would reproduce the functional dynamics of land surface processes found in SiB and other alternative land surface parameterizations. Global climate sensitivity experiments are performed with the new land surface parameterizations to assess how the new SiB, consistent CLM land surface parameterizations, influence the surface energy balance, hydrology, and atmospheric fluxes in CLM3, and through that the larger-scale climate modeled in CCSM3. It is found that the new parameterizations enable CLM to simulate evapotranspiration partitioning consistently with the multimodel average of other land surface models used in GCMs, as evaluated by Dirmeyer et al. (2005). The changes in surface fluxes also resulted in a number of improvements in the simulation of precipitation and near-surface air temperature in CCSM3. The new model is fully coupled in the CCSM3 framework, allowing a wide range of climate modeling investigations without the surface hydrology issues found in the current CLM3 model. This provides a substantially more robust framework for performing climate modeling experiments investigating the influence of land cover change and surface hydrology in CLM and CCSM than the existing CLM3 parameterizations. The study also shows that changes in land surface hydrology have global scale impacts on model climatology.

A piecewise-integration method for simulating the influence of external forcing on climate

Climate drift occurs in most general circulation models (GCMs) as a result of incomplete physical and numerical representation of the complex climate system, which may cause large uncertainty in sensitivity experiments evaluating climate response to changes in external forcing. To solve this problem, we propose a piecewise-integration method to reduce the systematic error in climate sensitivity studies. The observations are firstly assimilated into a numerical model by using the dynamic relaxation technique to relax to the current state of atmosphere, and then the assimilated fields are continuously used to reinitialize the simulation to reduce the error of climate simulation. When the numerical model is integrated with changed external forcing, the results can be split into two parts, background and perturbation fields, and the background is the state before the external forcing is changed. The piecewise-integration method is used to continuously reinitialize the model with the assimilated field, instead of the background. Therefore, the simulation error of the model with the external forcing can be reduced. In this way, the accuracy of climate sensitivity experiments is greatly improved. Tests with a simple low-order spectral model show that this approach can significantly reduce the uncertainty of climate sensitivity experiments.

Modeling Northern Hemisphere Summer Heat Extreme Changes and Their Uncertainties Using a Physics Ensemble of Climate Sensitivity Experiments

Abstract Changes in extreme daily temperature events are examined using a perturbed physics ensemble of global model simulations under present-day and doubled CO2 climates where ensemble members differ in their representation of various physical processes. Modeling uncertainties are quantified by varying poorly constrained model parameters that control atmospheric processes and feedbacks and analyzing the ensemble spread of simulated changes. In general, uncertainty is up to 50% of projected changes in extreme heat events of the type that occur only once per year. Large changes are seen in distributions of daily maximum temperatures for June, July, and August with significant shifts to warmer conditions. Changes in extremely hot days are shown to be significantly larger than changes in mean values in some regions. The intensity, duration, and frequency of summer heat waves are expected to be substantially greater over all continents. The largest changes are found over Europe, North and South America, and East Asia. Reductions in soil moisture, number of wet days, and nocturnal cooling are identified as significant factors responsible for the changes. Although uncertainty associated with the magnitude of expected changes is large in places, it does not bring into question the sign or nature of the projected changes. Even with the most conservative simulations, hot extreme events are still expected to substantially increase in intensity, duration, and frequency. This ensemble, however, does not represent the full range of uncertainty associated with future projections; for example, the effects of multiple parameter perturbations are neglected, as are the effects of structural changes to the basic nature of the parameterization schemes in the model.

Evaluating the accuracy and generality of a hybrid patch model

Forest patch models have been used extensively to simulate vegetation development under current and changing environmental conditions. However, their physiological foundation is subject to criticism and recent validation experiments against long-term growth and yield data have shown major deficiencies in reproducing observed growth patterns of mixed-species forests. Here we describe the modified forest patch model PICUS Version 1.3, a model variant that couples the structurally detailed three-dimensional patch model PICUS Version 1.2 and the physiologically based stand-level production module of the 3-PG (Physiological Principles in Predicting Growth) model. The approach attempts to combine the ability of PICUS v1.2 to simulate forest dynamics on time scales relevant to forest succession with a simplified but successful production model based on the concept of radiation-use efficiency. We evaluated the hybrid model in a series of simulation experiments. Results indicated a realistic response to a climate sensitivity experiment: the response to environmental gradients was well captured both in terms of productivity on time scales of a rotation length and of forest succession over several hundreds of years. Testing against independent long- term growth and yield data revealed good correspondence between observed and predicted values of volume production and stand structure. Further model development should include a dynamic soil component to consider effects of nutrient cycling.

Simulating the Greenland atmospheric boundary layer Part II: Energy balance and climate sensitivity

A three-dimensional dynamic downscaling model of the Greenland atmospheric boundary layer is usedto investigate the surface energy balance of the Greenland ice sheet for the summer ablation season of1998. The model is forced by ECMWF analysis data and has a horizontal resolution of 20 km. Theresults show that the turbulent heat fluxes contribute significantly (32%) to the total surface energy fluxin the ablation region. A climate sensitivity experiment is carried out with the model by increasing thefree atmospheric forcing temperature by 1 K. The resulting change, known as climate sensitivity, in2 m temperature, 2 m wind speed and the surface energy balance components are discussed. The 2 mtemperature is shown to have a climate sensitivity significantly less than unity, as low as 0.3, in regionswhere melt prevails for most of the ablation season. The surface energy flux in the ablation region isfound to increase by 21% for a 1 K increase in atmospheric temperature. Wind speeds also increase, particularly near the margin of the ice sheet, as a result of enhanced katabatic forcing. Enhanced windspeeds are shown to account for roughly 12% of the climate sensitivity of the sensible heat flux whenaveraged over the ablation region. Albedo feedback is also shown to play an important role in the totalclimate sensitivity of the surface energy balance.

Troposphere‐stratosphere dynamic coupling under strong and weak polar vortex conditions

The relationship between Northern Hemisphere (NH) tropospheric and stratospheric wave‐like anomalies of spherical zonal wave number (ZWN) 1 is studied by applying Canonical Correlation Analysis (CCA). A lag‐correlation technique is used with 10‐day lowpass filtered daily time series of 50‐ and 500‐hPa geopotential heights. Generally stratospheric circulation is determined by ultralong tropospheric planetary waves. During winter seasons characterized either by an anomalously strong or weak polar winter vortex different propagation characteristics for waves of ZWN 1 are observed. The non‐linear perspective of the results have implications for medium range weather forecast and climate sensitivity experiments.

Variability and climate sensitivity of landfast Arctic sea ice

A one‐dimensional thermodynamic sea ice model is used to perform multiyear simulations of land‐fast sea ice and its snow cover. The model is applied to two locations in the Canadian high Arctic: Resolute and Alert, Northwest Territories. The sites are chosen because meteorological observations and ice and snow depth measurements spanning the period 1955–1990 are available. Bulk formulae are used to convert standard meteorological observations into surface fluxes, snowfall is specified from observed annual accumulation, and a crude oceanic mixed layer is included to allow heat storage during ice‐free summer months. The model is successful at reproducing seasonal and interannual variability in maximum ice thickness, snow depth, freeze‐up, and breakup dates. Sensitivity studies demonstrate the relative importance of variables such as air temperature and snowfall rate on interannual variability of maximum ice thickness and the duration of the ice‐free season. Multiple equilibria, corresponding to thin, seasonal ice and thick, perennial ice, are shown to exist in the model. Climate sensitivity experiments illustrate the potential effects of changes in air temperature and snowfall on ice thickness and open water duration in the Canadian Archipelago.

Atmospheric Variability on a Zonally Symmetric Land Planet

Abstract Atmospheric variability an a zonally symmetric planet in the absence of external forcing anomalies is studied. With idealized boundary conditions such as the absence of ocean and topography, and by using perpetual equinox solar forcing, a 15-year long stationary time series of the atmosphere is simulated with the NCAR Community Climate Model (CCM0). This provides sufficient time samples for realistic study of the properties of the atmosphere. Zonally averaged and space-time statistics for the surface air temperature field on this planet are presented. Such statistics can serve as noise climatologies for climate sensitivity experiments, allowing the effects of changes of external forcing on the atmosphere to be asssessed. In search of a simple statistical model for atmospheric variability, the space-time spectra obtained from the CCM simulation are fitted statistically with a stochastic energy balance model. The space-time spectra for three zonal wavenumbers are found to be fitted satisfactorily b...

Climatic effects of reduced Arctic sea ice limits in the Giss II General Circulation Model

In this paper we present results of an atmospheric general circulation model (GCM) experiment in which Arctic sea ice limits were substantially reduced in all months. March sea ice limits were set equivalent to modern September limits, and all sea ice was removed in September. Sea ice coverage for other months varied between these two extremes. This climate sensitivity experiment makes predictions about mean northern hemisphere atmospheric conditions (including temperature, pressure, wind patterns, and precipitation) consistent with these boundary constraints. The major effects of reduced sea ice limits are observed in winter. They include large regional warming of the circum‐Arctic region, northward migration of the Icelandic low pressure system, and strengthening of the Azores high. Changes in net heating over the North Atlantic Ocean suggest that increases in sea surface temperatures and salinities in this region would also accompany reductions in Arctic sea ice limits. In the wind field, a weakening of the polar easterlies and an intensification of cyclonic circulation over the Norwegian‐Greenland Sea suggest that surface water exchange between the Atlantic and Arctic would increase when sea ice limits are reduced. However, zonally averaged changes in strength of the westerlies or upper level jet stream are minimal. The late Pliocene cooling of the North Atlantic Ocean and North American Arctic margin may have been linked in part to the development of perennial sea ice cover.

Coupled ocean-atmosphere models with flux correction

A method is proposed for removing the drift of coupled atmosphere-ocean models, which in the past has often hindered the application of coupled models in climate response and sensitivity experiments. The ocean-atmosphere flux fields exhibit inconsistencies when evaluated separately for the individual sub-systems in independent, uncoupled mode equilibrium climate computations. In order to balance these inconsistencies a constant ocean-atmosphere flux correction field is introduced in the boundary conditions coupling the two sub-systems together. The method ensures that the coupled model operates at the reference climate state for which the individual model subsystems were designed without affecting the dynamical response of the coupled system in climate variability experiments. The method is illustrated for a simple two component box model and an ocean general circulation model coupled to a two layer diagnostic atmospheric model.

Recurrence Analysis of Climate Sensitivity Experiments

A difficulty with the statistical techniques which are ordinarily used in the analysis of climate sensitivity experiments is that they do not identify the stable, or recurrent, aspects of the experimental response. Therefore, a new concept called “recurrence” is proposed. With this concept it is possible to identify the parts of the response which are likely to recur with an a priori likelihood each time a new experimental realization is obtained. A variety of statistical tests which may be used to assess an a priori level of recurrence by means of limited samples is suggested. A recurrence analysis is performed with data simulated by the Canadian Climate Centre general circulation model forced with climatological sea surface temperatures (SSTs) and with several El Ninño SST anomalies. All considered SST anomalies, a positive and a negative doubled standard Rasmusson and Carpenter anomaly and the winter 1982/83 anomaly excite a globally significantly response in terms of height and temperature. However, only part of the significant response is also recurrent. In the cold SST anomaly experiment, recurrence is confined to a minor part of the tropics. In the warm SST anomaly runs, recurrence is found in most of the tropics and partly over the northeastern Pacific. These results indicate that equatorial Pacific SST anomalies are associated with a rather limited predictive value, even if the anomalies are very strong.