Sea Surface Temperature Forecasts Research Articles

The dependence of medium range northern Atlantic Ocean predictability on atmospheric forecasts

The Mercator-Océan system provides a weekly analysis and a 14-day forecast of the ocean using a 1/3° North Atlantic model assimilating satellite altimetry and in situ temperature and salinity profiles. The model system makes use of the European Centre for Medium Range Weather Forecasting (ECMWF) 10-day operational forecast for atmospheric forcing. The ECMWF monthly ensemble atmospheric forecast allows an extension to one month ocean forecasts. This paper examines the impact of forcing with these two different atmospheric models on ocean forecast skill. Results show an equivalence of the 7-day ocean forecast with the two different atmospheric data sets and an improvement of the 14-day ocean forecast using the ECMWF ensemble mean output. 28-day ocean forecasts using the ECMWF ensemble output are always better than persistence with significant correlation; especially between forecast sea surface temperature (SST) and best estimate SST.

Potential value of GCM-based seasonal rainfall forecasts for maize management in semi-arid Kenya

We estimate the potential value of general circulation model (GCM)-based seasonal precipitation forecasts for maize planting and fertilizer management decisions at two semi-arid locations (Katumani and Makindu) in Southern Kenya. Analyses combine downscaled rainfall forecasts, crop yield simulation, stochastic enterprise budgeting and identification of profit-maximizing fertilizer N rates and stand densities. October–February rainfall predictions were downscaled from a GCM, run with both observed and forecast sea surface temperature boundary conditions – representing upper and lower bounds of predictability – and stochastically disaggregated into daily crop model inputs. Simulated interactive effects of rainfall, N supply and stand density on yield and profit are consistent with literature. Perfect foreknowledge of daily weather for the growing season would be worth an estimated 15–30% of the average gross value of production and 24–69% of average gross margin, depending on location and on whether household labor is included in cost calculations. GCM predictions based on observed sea surface temperatures increased average gross margins 24% at Katumani and 9% at Makindu when labor cost was included. At the lead time used, forecasts using forecast sea surface temperatures are not skillful and showed near-zero value. Forecast value was much more sensitive to grain price than to input costs. Stochastic dominance analysis shows that farmers at any level of risk aversion would prefer the forecast-based management strategy over management optimized for climatology under the study’s assumptions, despite high probability (25% at Katumani, 34% at Makindu) of lower returns in individual years. Results contribute to knowledge of seasonal forecast value in a relatively high-risk, high-predictability context; utility and value of forecasts derived from a GCM; and risk implications of smallholder farmers responding to forecasts.

Two Extra Components in the Brier Score Decomposition

Abstract The Brier score is widely used for the verification of probability forecasts. It also forms the basis of other frequently used probability scores such as the rank probability score. By conditioning (stratifying) on the issued forecast probabilities, the Brier score can be decomposed into the sum of three components: uncertainty, reliability, and resolution. This Brier score decomposition can provide useful information to the forecast provider about how the forecasts can be improved. Rather than stratify on all values of issued probability, it is common practice to calculate the Brier score components by first partitioning the issued probabilities into a small set of bins. This note shows that for such a procedure, an additional two within-bin components are needed in addition to the three traditional components of the Brier score. The two new components can be combined with the resolution component to make a generalized resolution component that is less sensitive to choice of bin width than is the traditional resolution component. The difference between the generalized resolution term and the conventional resolution term also quantifies how forecast skill is degraded when issuing categorized probabilities to users. The ideas are illustrated using an example of multimodel ensemble seasonal forecasts of equatorial sea surface temperatures.

Characteristics of tropical Pacific SST predictability in coupled GCM forecasts using the NCEP CFS

The limits of predictability of El Nino and the Southern Oscillation (ENSO) in coupled models are investigated based on retrospective forecasts of sea surface temperature (SST) made with the National Centers for Environmental Prediction (NCEP) coupled forecast system (CFS). The influence of initial uncertainties and model errors associated with coupled ENSO dynamics on forecast error growth are discussed. The total forecast error has maximum values in the equatorial Pacific and its growth is a strong function of season irrespective of lead time. The largest growth of systematic error of SST occurs mainly over the equatorial central and eastern Pacific and near the southeastern coast of the Americas associated with ENSO events. After subtracting the systematic error, the root-mean-square error of the retrospective forecast SST anomaly also shows a clear seasonal dependency associated with what is called spring barrier. The predictability with respect to ENSO phase shows that the phase locking of ENSO to the mean annual cycle has an influence on the seasonal dependence of skill, since the growth phase of ENSO events is more predictable than the decay phase. The overall characteristics of predictability in the coupled system are assessed by comparing the forecast error growth and the error growth between two model forecasts whose initial conditions are 1 month apart. For the ensemble mean, there is fast growth of error associated with initial uncertainties, becoming saturated within 2 months. The subsequent error growth follows the slow coupled mode related the model’s incorrect ENSO dynamics. As a result, the Lorenz curve of the ensemble mean NINO3 index does not grow, because the systematic error is identical to the same target month. In contrast, the errors of individual members grow as fast as forecast error due to the large instability of the coupled system. Because the model errors are so systematic, their influence on the forecast skill is investigated by analyzing the erroneous features in a long simulation. For the ENSO forecasts in CFS, a constant phase shift with respect to lead month is clear, using monthly forecast composite data. This feature is related to the typical ENSO behavior produced by the model that, unlike the observations, has a long life cycle with a JJA peak. Therefore, the systematic errors in the long run are reflected in the forecast skill as a major factor limiting predictability after the impact of initial uncertainties fades out.

Simulation of 2006 monsoon using T170L42 AGCM: sensitivity to convective parameterization schemes

AbstractIn this study we used the National Center for Environmental Prediction (NCEP) T170L42 Global Forecast System model for simulating the monsoon of 2006 using two different cumulus parameterization schemes viz. the Simple Arakawa–Schubert (SAS) scheme and the Relaxed Arakawa–Schubert (RAS) scheme. The NCEP Coupled Forecast System (CFS) monthly forecast sea surface temperatures (SST) were used for simulating the monsoon season. It is seen that both the schemes forecast the mean seasonal rainfall close to the observed. However, there were differences in the spatial and temporal distribution of the rainfall. The spatial distribution of the precipitation simulated by the RAS scheme was comparable to the observed. The SAS scheme simulated realistic distribution of the rainfall in the months of June and July. In the months of August and September, the SAS scheme could not simulate the convective regions over the south and central Bay of Bengal, Arabian Sea and the Indian landmass. The SAS scheme simulated withdrawal‐like symptoms of the monsoon in the month of August itself. However, the same SAS scheme simulated more realistic rainfall distribution over the Bay of Bengal when the model was run with climatological SST and also when the model was run at a resolution of T62 and with CFS‐forecast SST. The onset of the monsoon over the Kerala coast of India simulated with both the schemes was within one pentad of the observed date of onset. Copyright © 2008 Royal Meteorological Society

Robust nonlinear canonical correlation analysis: application to seasonal climate forecasting

Abstract. Robust variants of nonlinear canonical correlation analysis (NLCCA) are introduced to improve performance on datasets with low signal-to-noise ratios, for example those encountered when making seasonal climate forecasts. The neural network model architecture of standard NLCCA is kept intact, but the cost functions used to set the model parameters are replaced with more robust variants. The Pearson product-moment correlation in the double-barreled network is replaced by the biweight midcorrelation, and the mean squared error (mse) in the inverse mapping networks can be replaced by the mean absolute error (mae). Robust variants of NLCCA are demonstrated on a synthetic dataset and are used to forecast sea surface temperatures in the tropical Pacific Ocean based on the sea level pressure field. Results suggest that adoption of the biweight midcorrelation can lead to improved performance, especially when a strong, common event exists in both predictor/predictand datasets. Replacing the mse by the mae leads to improved performance on the synthetic dataset, but not on the climate dataset except at the longest lead time, which suggests that the appropriate cost function for the inverse mapping networks is more problem dependent.

Sea Surface Temperature Feedback Extends the Predictability of Tropical Intraseasonal Oscillation

Abstract The possible impacts of different sea surface temperature (SST) configurations on the predictability of the boreal summer tropical intraseasonal oscillation (TISO) are assessed with a series of ensemble forecasts. The five different lower boundary conditions examined in this study are, respectively, (i) the fully interactive ocean–atmosphere coupling, (ii) “smoothed” SST, which excludes the intraseasonal signal from sea surface forcing, (iii) damped persistent SST, (iv) coupling to a slab mixed-layer ocean, and (v) daily SST from the coupled forecast. The full atmosphere–ocean coupling generates an interactive SST that results in the highest TISO predictability of about 30 days over Southeast Asia. The atmosphere-only model is capable of reaching this predictability if the ensemble mean daily SST forecast by the coupled model is used as the lower boundary condition, which suggests that, in principle, the so-called tier-one and tier-two systems have the same predictability for the boreal summer TISO. The atmosphere-only model driven by either smoothed or damped persistent SSTs, however, has the lowest predictability (∼20 days). The atmospheric model coupled to a slab mixed-layer ocean achieves a predictability of 25 days. The positive SST anomalies in the northern Indo–western Pacific Oceans trigger convective disturbances by moistening and warming up the atmospheric boundary layer. The seasonal mean easterly shear intensifies the anomalous convection by enhancing the surface convergence. An overturning meridional circulation driven by the off-equatorial anomalous convection suppresses the near-equatorial convection and enhances the northward flows, which further intensify the off-equatorial surface convergence and the TISO-related convection. Thus, the boreal summer mean easterly shear and the overturning meridional circulation in the northern Indo–western Pacific sector act as “amplifiers” for the SST feedback to the convection of the TISO.

Impact of Argo on analyses of the global ocean

The impact of Argo on ECMWF operational ocean analyses has been assessed by conducting a set of observing system experiments spanning the period 2001–2006. The experiments evaluate the information content of Argo temperature and salinity data, gauged in terms of influence on the ocean state and the skill of seasonal forecasts. The salinity data from Argo is instrumental in correcting the salinity of the ECMWF analysis on a basin‐wide scale. The effect of Argo temperature is noticeable in the Indian, Atlantic and Southern Oceans. In the Pacific, the impact is modest, being confined to the South American coast. The information content of Argo combines well with the altimeter information in most regions. The impact of Argo on seasonal forecasts of sea surface temperature has been assessed for the period 2001–2006. The assimilation of Argo improves seasonal forecast skill in most areas, although the reduction of the error is small.

Bayesian merging of multiple climate model forecasts for seasonal hydrological predictions

This study uses a Bayesian approach to merge ensemble seasonal climate forecasts generated by multiple climate models for better probabilistic and deterministic forecasting. Within the Bayesian framework, the climatological distribution of the variable of interest serves as the prior, and the likelihood function is developed with a weighted linear regression between the climate model hindcasts and the corresponding observations. The resulting posterior distribution is the merged forecast, which represents our best estimate of the variable, including its mean and variance, given the current model forecast and knowledge about the model’s performance. The handling of multimodel climate forecasts and nonnormal distributed variables, such as precipitation, are two important extensions toward the application of the Bayesian merging approach for seasonal hydrological predictions. Two examples are presented as follows: seasonal forecast of sea surface temperature over equatorial Pacific and precipitation forecast over the Ohio River basin. Cross validation of these forecasts shows smaller root mean square error and smaller ranked probability score for the merged forecast as compared with raw forecasts from climate models and the climatological forecast, indicating an improvement in both deterministic and probabilistic forecast skills. Therefore there is great potential to apply this method to seasonal hydrological forecasting.

SST sensitivity of a global ocean–atmosphere coupled system to the parameterization of boundary layer clouds

In this note we report on the impact of a more realistic low cloud cover parameterization on the Sea Surface Temperature (SST) predicted by an ocean–atmosphere coupled system. In particular, it is shown that the atmospheric model with more realistic boundary layer clouds leads to a more realistic distribution of the SST in the ocean's tropics, subtropics and mid-latitudes. Typical SST biases of up to 5 °C in regions under frequently observed stratus and stratocumulus are substantially reduced to unbiased values at places. This represents a substantial SST forecast improvement in terms of operational coupled systems used in the context of Maritime Rapid Environmental Assessment (MREA).

Ensemble Simulation of Indian Summer Monsoon Rainfall by an Atmospheric General Circulation Model

This study examines the fidelity of the Meteorological Research Institute (MRI) atmospheric general circulation model (AGCM), ensemble runs forced with observed sea surface temperature (SST), in simulating Indian summer monsoon rainfall (ISMR), and its interannual variation. Despite the simple ensemble mean (SEM) capturing essential features of climatological ISMR pattern and its extreme ISMR anomalies, it still shows certain systematic bias in simulating mean seasonal variation of rainfall over the Asia-Pacific region. Concurrently, the ISMR interannual variability throughout the analysis period is not adequately represented.A bias-correction is applied to remove this bias by deriving weights for the member simulations for each Julian day separately at every grid point through multiple linear regression of their daily rainfall, against corresponding observation over a 23-year ‘training phase’ (out of the total 24-year analysis period). Thereafter, at every grid point, for each Julian day of the remaining 1-year ‘forecast phase’, the bias-removed ensemble mean (BREM) is computed as an optimal linear combination of weighted member simulations. In cross validation, each year in the analysis period is treated successively as the ‘forecast phase’, with the remaining 23 years included in the corresponding training phase. The methodology minimises the systematic bias in mean seasonal variation of rainfall, and BREM consequently improves upon SEM in simulating the mean ISMR pattern, and its interannual variability over the entire analysis period.The skill of BREM in forecasting the ISMR, and its intraseasonal variability is validated, for the severe monsoon drought of 2002. The effective removal of climatological bias brings out the realistic precipitation response in BREM to fluctuations in SST boundary forcing. As a result, BREM captures the seasonal rainfall anomaly pattern of 2002, and markedly improves its intraseasonal evolution. Apart from the basic skill of the AGCM ensemble system, pronounced equatorial SST impact in modulating the monsoon circulation during 2002, played a seminal role in the success of the methodology. The analysis also underlines the importance of mean seasonal variation, not only for capturing ISMR climatology, and its interannual variation but for improving its intraseasonal variability as well. With the aid of realistic SST forecasts, this methodology has the application potential for dynamical prediction of ISMR.

Seasonal Prediction of Sea Surface Temperature Anomalies Using a Suite of 13 Coupled Atmosphere–Ocean Models

Abstract Improved seasonal prediction of sea surface temperature (SST) anomalies over the global oceans is the theme of this paper. Using 13 state-of-the-art coupled global atmosphere–ocean models and 13 yr of seasonal forecasts, the performance of individual models, the ensemble mean, the bias-removed ensemble mean, and the Florida State University (FSU) superensemble are compared. A total of 23 400 seasonal forecasts based on 1-month lead times were available for this study. Evaluation metrics include both deterministic and probabilistic skill measures, such as verification of anomalies based on model and observed climatology, time series of specific climate indices, standard deterministic ensemble mean scores including anomaly correlations, root-mean-square (RMS) errors, and probabilistic skill measures such as equitable threat scores for seasonal SST forecasts. This study also illustrates the Niño-3.4 SST forecast skill for the equatorial Pacific Ocean and for the dipole index for the Indian Ocean. The relative skills of total SST fields and of the SST anomalies from the 13 coupled atmosphere–ocean models are presented. Comparisons of superensemble-based seasonal forecasts with recent studies on SST anomaly forecasts are also shown. Overall it is found that the multimodel superensemble forecasts are characterized by considerable RMS error reductions and increased accuracy in the spatial distribution of SST. Superensemble SST skill also persists for El Niño and La Niña forecasts since the large comparative skill of the superensemble is retained across such years. Real-time forecasts of seasonal sea surface temperature anomalies appear to be possible.

Neural network forecasts of the tropical Pacific sea surface temperatures

A nonlinear forecast system for the sea surface temperature (SST) anomalies over the whole tropical Pacific has been developed using a multi-layer perceptron neural network approach, where sea level pressure and SST anomalies were used as predictors to predict the five leading SST principal components at lead times from 3 to 15 months. Relative to the linear regression (LR) models, the nonlinear (NL) models showed higher correlation skills and lower root mean square errors over most areas of the domain, especially over the far western Pacific (west of 155°E) and the eastern equatorial Pacific off Peru at lead times longer than 3 months, with correlation skills enhanced by 0.10–0.14. Seasonal and decadal changes in the prediction skills in the NL and LR models were also studied.

Simulating the value of El Niño forecasts for the Panama Canal

The Panama Canal relies on rain-fed streamflow into Gatun Lake, the canal’s primary storage facility, for operations—principally ship passage and hydropower generation. Precipitation in much of Panama has a strong negative relationship with eastern tropical Pacific sea surface temperature (SST) and this relationship is reflected in Gatun Lake inflows. For example, the correlation coefficient between wet season (July–December) inflow and NINO3 SST is −0.53 over the period 1914–1997. Operational capabilities to predict tropical Pacific SSTs have been demonstrated by several forecast systems during the past decade, and (as we show) such SST forecasts can be used to reduce the uncertainty of estimates of future inflows (compared with climatological expectations). Because substantial reductions in lake inflow negatively impact canal operations, we wondered whether these forecasts of future inflows, coupled with a method for translating that information into effective operational policy, might result in more efficient canal management. A combined simulation/optimization/assessment “virtual” canal system was implemented and exercised using operational El Niño forecasts over the period 1981–1998. The results show the following main points: (i) At current demand levels, the canal system is relatively robust (insensitive to flow forecasts) unless flows are substantially reduced (i.e., during El Niño episodes) or forecasts are extremely accurate. (ii) The inclusion of accurately specified levels of forecast uncertainty is critical in developing economically beneficial policies. (iii) The situations in which imperfect forecast information can be useful lie between those where storage and future inflows are relatively high, and those where storage and inflows are relatively low. In the former case, demands can be met without the benefit of forecast information, and in the latter case even perfect forecast information cannot prevent operational curtailments. (iv) For a nominally configured canal system, the use of operational El Niño forecasts with appropriately specified uncertainty resulted in approximately $US 20,000,000 (about 3%) in increased annual average income compared with the use of deterministic climatological forecasts.

Impact of Sea Surface Temperature and Soil Moisture on Seasonal Rainfall Prediction over the Sahel

Abstract The ensemble rainfall forecasts over the Sahel for July–September (JAS) from the NCEP Coupled Forecast System (CFS) were evaluated for the period 1981–2002. The comparison with the gauge-based precipitation analysis indicates that the predicted Sahel rainfall is light and exhibits little interannual variability. The rain belt is shifted about 4° southward. One major source of rainfall errors comes from the erroneous sea surface temperature (SST) forecasts. The systematic SST error pattern has positive errors in the North Pacific and the North Atlantic and negative errors in the tropical Pacific and the southern oceans. It resembles the decadal SST mode, which has a significant influence on rainfall over the Sahel. Because the systematic SST errors were not corrected during the forecasts, persistent errors serve as an additional forcing to the atmosphere. The second source of error is from the soil moisture feedback, which contributes to the southward shift of rainfall and dryness over West Africa. This was demonstrated by the comparison between simulations (SIMs) and the Atmospheric Model Intercomparison Project (AMIP) run. Both are forced with observed SSTs. The SIMs initialized at the end of June have realistic soil moisture and do not show the southward shift of rainfall. The AMIP, which predicts soil moisture, maintains the dryness through the summer over the Sahel. For AMIP, the decreased rainfall is contributed by the decreased evaporation (E) due to the dry soil and the shift of the large temperature gradients southward. In response, the African easterly jet (AEJ) shifts southward. Since this jet is the primary source of energy for the African waves and their associated mesoscale convective systems, these too shift southward. This negative feedback contributes to increased dryness over the Sahel.

Statistical Correction of Tropical Pacific Sea Surface Temperature Forecasts

Abstract This paper is about the statistical correction of systematic errors in dynamical sea surface temperature (SST) prediction systems using linear regression approaches. The typically short histories of model forecasts create difficulties in developing regression-based corrections. The roles of sample size, predictive skill, and systematic error are examined in evaluating the benefit of a linear correction. It is found that with the typical 20 yr of available model SST forecast data, corrections are worth performing when there are substantial deviations in forecast amplitude from that determined by correlation with observations. The closer the amplitude of the uncorrected forecasts is to the optimum squared error-minimizing amplitude, the less likely is a correction to improve skill. In addition to there being less “room for improvement,” this rule is related to the expected degradation in out-of-sample skill caused by sampling error in the estimate of the regression coefficient underlying the correction. Application of multivariate [canonical correlation analysis (CCA)] correction to three dynamical SST prediction models having 20 yr of data demonstrates improvement in the cross-validated skills of tropical Pacific SST forecasts through reduction of systematic errors in pattern structure. Additional beneficial correction of errors orthogonal to the CCA modes is achieved on a per-gridpoint basis for features having smaller spatial scale. Until such time that dynamical models become freer of systematic errors, statistical corrections such as those shown here can make dynamical SST predictions more skillful, retaining their nonlinear physics while also calibrating their outputs to more closely match observations.

Retrospective Forecasts of Interannual Sea Surface Temperature Anomalies from 1982 to Present Using a Directly Coupled Atmosphere–Ocean General Circulation Model

Abstract A large number of ensemble hindcasts (or retrospective forecasts) of tropical Pacific sea surface temperature (SST) have been made with a coupled atmosphere–ocean general circulation model (CGCM) that does not employ flux correction in order to evaluate the potential skill of the model as a seasonal forecasting tool. Oceanic initial conditions are provided by an ocean data assimilation system. Ensembles of seven forecasts of 6-month length are made starting each month in the 1982 to 2002 period. Skill of the coupled model is evaluated from both a deterministic and a probabilistic perspective. The skill metrics are calculated using both the bulk method, which includes all initial condition months together, and as a function of initial condition month. The latter method allows a more objective evaluation of how the model has performed in the context in which forecasts are actually made and applied. The deterministic metrics used are the anomaly correlation and the root-mean-square error. The coupled model deterministic skill metrics are compared with those from persistence and damped persistence reference forecasts. Despite the fact that the coupled model has a large cold bias in the central and eastern equatorial Pacific this coupled model is shown to have forecast skill that is competitive with other state-of-the-art forecasting techniques. Potential skill from probabilistic forecasts made using the coupled model ensemble members are evaluated using the relative operating characteristics method. This analysis indicates that for most initial condition months this coupled model has more skill at forecasting cold events than warm or neutral events in the central Pacific. In common with other forecasting systems, the coupled model forecast skill is found to be lowest for forecasts passing through the Northern Hemisphere (NH) spring. Diagnostics of this so-called spring predictability barrier in the context of this coupled model indicate that two factors likely contribute to this predictability barrier. First, the coupled model shows a too-weak coupling of the surface and subsurface temperature anomalies during NH spring. Second, the coupled-model-simulated signal-to-noise ratio for SST anomalies is much lower during NH spring than at other times of the year, indicating that the model’s potential predictability is low at this time.

Calibrated Probabilistic Forecasting Using Ensemble Model Output Statistics and Minimum CRPS Estimation

Abstract Ensemble prediction systems typically show positive spread-error correlation, but they are subject to forecast bias and dispersion errors, and are therefore uncalibrated. This work proposes the use of ensemble model output statistics (EMOS), an easy-to-implement postprocessing technique that addresses both forecast bias and underdispersion and takes into account the spread-skill relationship. The technique is based on multiple linear regression and is akin to the superensemble approach that has traditionally been used for deterministic-style forecasts. The EMOS technique yields probabilistic forecasts that take the form of Gaussian predictive probability density functions (PDFs) for continuous weather variables and can be applied to gridded model output. The EMOS predictive mean is a bias-corrected weighted average of the ensemble member forecasts, with coefficients that can be interpreted in terms of the relative contributions of the member models to the ensemble, and provides a highly competitive deterministic-style forecast. The EMOS predictive variance is a linear function of the ensemble variance. For fitting the EMOS coefficients, the method of minimum continuous ranked probability score (CRPS) estimation is introduced. This technique finds the coefficient values that optimize the CRPS for the training data. The EMOS technique was applied to 48-h forecasts of sea level pressure and surface temperature over the North American Pacific Northwest in spring 2000, using the University of Washington mesoscale ensemble. When compared to the bias-corrected ensemble, deterministic-style EMOS forecasts of sea level pressure had root-mean-square error 9% less and mean absolute error 7% less. The EMOS predictive PDFs were sharp, and much better calibrated than the raw ensemble or the bias-corrected ensemble.

Actual and potential skill of seasonal predictions using the CNRM contribution to DEMETER: coupled versus uncoupled model

The skill of seasonal forecasts carried out with the CNRM general circulation model (GCM) is assessed over the winter and summer seasons during the period 1958–2001. Three types of forecasts are considered. Two of them, compatible with real-time forecasting, make use of the ocean’atmosphere coupled model and the sea surface temperature (SST) statistical model, while the third provides an upper limit of the skill taking the observed SSTs into account. The ninemember ensemble skill is evaluated with the help of anomaly correlation coefficients (deterministic skill) and economical values (probabilistic skill). The coupled model skill is larger than that of the statistical SST forced model, from both deterministic and probabilistic point of views. Moreover, the coupled forecast skill is not far from that of the observed SST forecasts. Over the tropics, the correlations (above 0.3) and economical values are significant. Over Europe, only observed SST forecasts reach the level of significance in both seasons. For the years of the largest observed normalized anomalies (tropical precipitation and temperature at 850 hPa over Europe), composites of the atmospheric response are constructed in a trial of understanding the nature of predictability in the GCM, in comparison with the observation. The teleconnection patterns between the tropics and the Northern Hemisphere are not well reproduced by the GCM. However, it shows a rather good potential skill in the sense that the years characterized by the largest internal ensemble consistency give an atmospheric response very similar to that obtained during the years for which the observed anomalies are the largest.

Long-Lead Seasonal Temperature and Precipitation Prediction Using Tropical Pacific SST Consolidation Forecasts

Abstract Objective seasonal forecasts of temperature and precipitation for the conterminous United States are produced using tropical Pacific sea surface temperature forecasts for the Nino-3.4 region in conjunction with composites of observed temperature and precipitation keyed to phases of the ENSO cycle. The objective seasonal forecasts are validated against observations for the period February–March–April (FMA) 1995 to September–October–November (SON) 2002, and compared to NOAA's Official Seasonal Forecasts issued by the Climate Prediction Center (CPC) for the same period. The objective forecasts are shown to produce skill that is comparable to (and even exceeding) that achieved by the Official Seasonal Forecasts at all leads out to 12.5 months. The forecasts are divided into high-frequency (HF) and trend-adjusted (TA) components in order to show that seasonal forecasters could achieve higher skill in both temperature and precipitation forecasts by taking full advantage of trend information, especially...