Sea Surface Temperature Prediction Skill Research Articles

On the Development of GFDL's Decadal Prediction System: Initialization Approaches and Retrospective Forecast Assessment

AbstractUsing GFDL's new coupled model SPEAR, we have developed a decadal coupled reanalysis/initialization system (DCIS) that does not use subsurface ocean observations. In DCIS, the winds and temperature in the atmosphere, along with sea surface temperature (SST), are restored to observations. Under this approach the ocean component of the coupled model experiences a sequence of surface heat and momentum fluxes that are similar to observations. DCIS offers two initialization approaches, called A1 and A2, which differ only in the atmospheric forcing from observations. In A1, the atmospheric winds/temperature are restored toward the JRA reanalysis; in A2, surface pressure observations are assimilated in the model. Two sets of coupled reanalyses have been completed during 1961–2019 using A1 and A2, and they show very similar multi‐decadal variations of the Atlantic Meridional Overturning Circulation (AMOC). Two sets of retrospective decadal forecasts were then conducted using initial conditions from the A1 and A2 reanalyses. In comparison with previous prediction system CM2.1, SPEAR‐A1/A2 shows comparable skill of predicting the North Atlantic subpolar gyre SST, which is highly correlated with initial values of AMOC at all lead years. SPEAR‐A1 significantly outperforms CM2.1 in predicting multi‐decadal SST trends in the Southern Ocean (SO). Both A1 and A2 have skillful prediction of Sahel precipitation and the associated ITCZ shift. The prediction skill of SST is generally lower in A2 than A1 especially over SO presumably due to the sparse surface pressure observations.

Evaluation of Sea Surface Temperature Prediction Skill around the Korean Peninsula in GloSea5 Hindcast: Improvement with Bias Correction

Evaluation of Sea Surface Temperature Prediction Skill around the Korean Peninsula in GloSea5 Hindcast: Improvement with Bias Correction

Ensemble estimates of the wave state related parameters in a sea spray parameterization scheme

Uncertainties of the wave state parameters in a sea spray parameterization scheme can be a major source of errors for air-sea turbulent (momentum, sensible and latent) fluxes parameterizations resulting in biases in numerical ocean simulations and forecasts. In this study, we explore applications of the ensemble adjustment Kalman filter (EAKF) data assimilation method to optimize the wave states parameters the 1-D POM ocean model for a full range of wind speed conditions. Thus, we assimilate sea surface temperature (SST) synthesized observations to improve our estimates of the SST analysis and prediction skill from low to high winds. Two types of experiments are conducted. In the first type, in a “twin” experiment framework, the SST “observations” generated by a “truth” model are assimilated into an imperfect, biased model to investigate the extent to which the parameters are able to be optimized, with respect to the “truth” values based on data from Station Papa and the Kuroshio Extension Observatory (KEO). In the second type, real SST observations from KEO are assimilated to obtain optimized parameters. With these optimized parameters, the SST analysis and prediction errors are significantly reduced, especially for high wind conditions.

Importance of convective parameterization in ENSO predictions

AbstractThis letter explored the influence of atmospheric convection scheme on El Niño–Southern Oscillation (ENSO) predictions using a set of hindcast experiments. Specifically, a low‐resolution version of the Climate Forecast System version 2 is used for 12 month hindcasts starting from each April during 1982–2011. The hindcast experiments are repeated with three atmospheric convection schemes. All three hindcasts apply the identical initialization with ocean initial conditions taken from the European Centre for Medium‐Range Weather Forecasts and atmosphere/land initial states from the National Centers for Environmental Prediction. Assessments indicate a substantial sensitivity of the sea surface temperature prediction skill to the different convection schemes, particularly over the eastern tropical Pacific. For the Niño 3.4 index, the anomaly correlation skill can differ by 0.1–0.2 at lead times longer than 2 months. Long‐term simulations are further conducted with the three convection schemes to understand the differences in prediction skill. By conducting heat budget analyses for the mixed‐layer temperature anomalies, it is suggested that the convection scheme having the highest skill simulates stronger and more realistic coupled feedbacks related to ENSO. Particularly, the strength of the Ekman pumping feedback is better represented, which is traced to more realistic simulation of surface wind stress. Our results imply that improving the mean state simulations in coupled (ocean‐atmosphere) general circulation model (e.g., ameliorating the Intertropical Convergence Zone simulation) might further improve our ENSO prediction capability.

The role of atmospheric internal variability on the prediction skill of interannual North Pacific sea-surface temperatures

The sensitivity of the sea-surface temperature (SST) prediction skill to the atmospheric internal variability (weather noise) in the North Pacific (20∘–60∘N;120∘E–80∘W) on decadal timescales is examined using state-of-the-art Climate Forecasting System model version 2 (CFS) and a variation of CFS in an Interactive Ensemble approach (CFSIE), wherein six copies of atmospheric components with different perturbed initial states of CFS are coupled with the same ocean model by exchanging heat, momentum and fresh water fluxes dynamically at the air-sea interface throughout the model integrations. The CFSIE experiments are designed to reduce weather noise and using a few ten-year long forecasts this study shows that reduction in weather noise leads to lower SST forecast skill. To understand the pathways that cause the reduced SST prediction skill, two twenty-year long forecasts produced with CFS and CFSIE for 1980-2000 are analyzed for the ocean subsurface characteristics that influence SST due to the reduction in weather noise in the North Pacific. The heat budget analysis in the oceanic mixed layer across the North Pacific reveals that weather noise significantly impacts the heat transport in the oceanic mixed layer. In the CFSIE forecasts, the reduced weather noise leads to increased variations in heat content due to shallower mixed layer, diminished heat storage and enhanced horizontal heat advection. The enhancement of the heat advection spans from the active Kuroshio regions of the east coast of Japan to the west coast of continental United States and significantly diffuses the basin-wide SST anomaly (SSTA) contrasts and leads to reduction in the SST prediction skill in decadal forecasts.

Seasonality in Prediction Skill and Predictable Pattern of Tropical Indian Ocean SST

Abstract Seasonality of sea surface temperature (SST) predictions in the tropical Indian Ocean (TIO) was investigated using hindcasts (1982–2009) made with the NCEP Climate Forecast System version 2 (CFSv2). CFSv2 produced useful predictions of the TIO SST with lead times up to several months. A substantial component of this skill was attributable to signals other than the Indian Ocean dipole (IOD). The prediction skill of the IOD index, defined as the difference between the SST anomaly (SSTA) averaged over 10°S–0°, 90°–110°E and 10°S–10°N, 50°–70°E, had strong seasonality, with high scores in the boreal autumn. In spite of skill in predicting its two poles with longer leads, CFSv2 did not have skill significantly better than persistence in predicting IOD. This was partly because the seasonal nature of IOD intrinsically limits its predictability. The seasonality of the predictable patterns of the TIO SST was further explored by applying the maximum signal-to-noise (MSN) empirical orthogonal function (EOF) method to the predicted SSTA in March and October, respectively. The most predictable pattern in spring was the TIO basin warming, which is closely associated with El Niño. The basin mode, including its associated atmospheric anomalies, can be predicted at least 9 months ahead, even though some biases were evident. On the other hand, the most predictable pattern in fall was characterized by the IOD mode. This mode and its associated atmospheric variations can be skillfully predicted only 1–2 seasons ahead. Statistically, the predictable IOD mode coexists with El Niño; however, the 1994 event in a non-ENSO year (at least not a canonical ENSO year) can also be predicted at least 3 months ahead by CFSv2.

An Analysis of the Temporal Evolution of ENSO Prediction Skill in the Context of the Equatorial Pacific Ocean Observing System

Abstract Subsurface ocean observations in the equatorial tropical Pacific Ocean dramatically increased after the 1990s because of the completion of the TAO moored array and a steady increase in Argo floats. In this analysis the question explored is whether a steady increase in ocean observations can be discerned in improvements in skill of predicting sea surface temperature (SST) variability associated with El Niño–Southern Oscillation (ENSO)? The analysis is based on the time evolution of skill of sea surface temperatures in the equatorial tropical Pacific since 1982 based on a seasonal prediction system. It is found that for forecasts up to a 6-month lead time, a clear fingerprint of increases in subsurface ocean observations is not readily apparent in the time evolution of prediction skill that is dominated much more by the signal-to-noise consideration of SSTs to be predicted. Finding no clear relationship between an increase in ocean observations and prediction skill of SSTs, various possibilities for why it may be so are discussed. This discussion is to motivate further exploration on the question of the tropical Pacific observing system, its influence on the skill of ENSO prediction, and the capabilities of the current generation of coupled models and ocean data assimilation systems to take advantage of ocean observations.

Assimilating operational SST and sea ice analysis data into an operational circulation model for the coastal seas of China

The prediction of sea surface temperature (SST) is an essential task for an operational ocean circulation model. A sea surface heat flux, an initial temperature field, and boundary conditions directly affect the accuracy of a SST simulation. Here two quick and convenient data assimilation methods are employed to improve the SST simulation in the domain of the Bohai Sea, the Yellow Sea and the East China Sea (BYECS). One is based on a surface net heat flux correction, named as Qcorrection (QC), which nudges the flux correction to the model equation; the other is ensemble optimal interpolation (EnOI), which optimizes the model initial field. Based on such two methods, the SST data obtained from the operational SST and sea ice analysis (OSTIA) system are assimilated into an operational circulation model for the coastal seas of China. The results of the simulated SST based on four experiments, in 2011, have been analyzed. By comparing with the OSTIA SST, the domain averaged root mean square error (RMSE) of the four experiments is 1.74, 1.16, 1.30 and 0.91°C, respectively; the improvements of assimilation experiments Exps 2, 3 and 4 are about 33.3%, 25.3%, and 47.7%, respectively. Although both two methods are effective in assimilating the SST, the EnOI shows more advantages than the QC, and the best result is achieved when the two methods are combined. Comparing with the observational data from coastal buoy stations, show that assimilating the high-resolution satellite SST products can effectively improve the SST prediction skill in coastal regions.

An evaluation of experimental decadal predictions using CCSM4

This study assesses retrospective decadal prediction skill of sea surface temperature (SST) variability in initialized climate prediction experiments with the Community Climate System Model version 4 (CCSM4). Ensemble forecasts initialized with two different historical ocean and sea-ice states are evaluated and compared to an ensemble of uninitialized coupled simulations. Both experiments are subject to identical twentieth century historical radiative forcings. Each forecast consists of a 10-member ensemble integrated over a 10-year period. One set of historical ocean and sea-ice conditions used for initialization comes from a forced ocean-ice simulation driven by the Coordinated Ocean-ice Reference Experiments interannually varying atmospheric dataset. Following the Coordinated Model Intercomparison Project version 5 (CMIP5) protocol, these forecasts are initialized every 5 years from 1961 to 1996, and every year from 2000 to 2006. A second set of initial conditions comes from historical ocean state estimates obtained through the assimilation of in-situ temperature and salinity data into the CCSM4 ocean model. These forecasts are only available over a limited subset of the CMIP5 recommended start dates. Both methods result in retrospective SST prediction skill over broad regions of the Indian Ocean, western Pacific Ocean and North Atlantic Ocean that are significantly better than reference skill levels from a spatio-temporal auto-regressive statistical model of SST. However the subpolar gyre region of the North Atlantic stands out as the only region where the CCSM4 initialized predictions outperform uninitialized simulations. Some features of the ocean state estimates used for initialization and their impact on the forecasts are discussed.

Improvement of Initialized Decadal Predictions over the North Pacific Ocean by Systematic Anomaly Pattern Correction

The prediction skill and errors in surface temperature anomalies in initialized decadal hindcasts from phase 5 of the Coupled Model Intercomparison Project (CMIP5) are assessed using six ocean–atmosphere coupled models initialized every year from 1961 to 2008. The initialized hindcasts show relatively high prediction skill over the regions where external forcing dominates, indicating that a large portion of the prediction skill is due to the long-term trend. After removing the linear trend, high prediction skill is shown near the centers of action of the dominant decadal climate oscillations, such as the Pacific decadal oscillation (PDO) and Atlantic multidecadal oscillation (AMO). Low prediction skill appears over the tropical and eastern North Pacific Ocean where the predicted anomaly patterns associated with the PDO are systematically different in model and observations. By statistically correcting those systematic errors using a stepwise pattern projection method (SPPM) based on the data in an independent training period, the prediction skill of sea surface temperature (SST) is greatly enhanced over the North Pacific Ocean. The SST prediction skill over the North Pacific Ocean after the SPPM error correction is as high as that over the North Atlantic Ocean. In addition, the prediction skill in a single model after correction exceeds the skill of the multimodel ensemble (MME) mean before correction, implying that the MME method is not as effective in addressing systematic errors as the SPPM correction.

A study of the predictability of sea surface temperature over the tropics

In this study, reasons for spatial variability in prediction skill of seasonal mean sea surface temperature (SST) are investigated. The analysis explores the connection between the spatial variation of SST prediction skill in tropical latitudes and local simultaneous correlation between observed SST and precipitation (SST-P). The results show that high (low) SST prediction skills and slow (fast) decay in skill with lead time are generally collocated with large positive (weakly positive or negative) SST-P correlations. The reasons for spatial variation in SST-P correlation can be explained by whether the primary forcing is from the ocean to the atmosphere or vice versa. Over regions where the ocean is generally known to force the atmosphere, it is found that SST-P correlation has large positive values. Over regions where the atmosphere forces ocean, SST-P correlation is weak due to intrinsically unpredictable nature of atmospheric variability. The physical explanation for spatial variation in SST-P correlation, and apparent link between the spatial variations in SST-P correlation and spatial variations in skill of SST predictions also establishes a physical basis for the latter. As a corollary, low SST prediction skill in coupled forecast models over certain geographical regions may be due to the inherent limits on predictability in addition to the contribution from model biases or initialization errors.

Prediction skill of monthly SST in the North Atlantic Ocean in NCEP Climate Forecast System version 2

This work evaluates the skill of retrospective predictions of the second version of the NCEP Climate Forecast System (CFSv2) for the North Atlantic sea surface temperature (SST) and investigates the influence of El Nino-Southern Oscillation (ENSO) and North Atlantic Oscillation (NAO) on the prediction skill over this region. It is shown that the CFSv2 prediction skill with 0–8 month lead displays a “tripole”-like pattern with areas of higher skills in the high latitude and tropical North Atlantic, surrounding the area of lower skills in the mid-latitude western North Atlantic. This “tripole”-like prediction skill pattern is mainly due to the persistency of SST anomalies (SSTAs), which is related to the influence of ENSO and NAO over the North Atlantic. The influences of ENSO and NAO, and their seasonality, result in the prediction skill in the tropical North Atlantic the highest in spring and the lowest in summer. In CFSv2, the ENSO influence over the North Atlantic is overestimated but the impact of NAO over the North Atlantic is not well simulated. However, compared with CFSv1, the overall skills of CFSv2 are slightly higher over the whole North Atlantic, particularly in the high latitudes and the northwest North Atlantic. The model prediction skill beyond the persistency initially presents in the mid-latitudes of the North Atlantic and extends to the low latitudes with time. That might suggest that the model captures the associated air-sea interaction in the North Atlantic. The CFSv2 prediction is less skillful than that of SSTA persistency in the high latitudes, implying that over this region the persistency is even better than CFSv2 predictions. Also, both persistent and CFSv2 predictions have relatively low skills along the Gulf Stream.

Retrospective El Niño Forecasts Using an Improved Intermediate Coupled Model

AbstractA new intermediate coupled model (ICM) is presented and employed to make retrospective predictions of tropical Pacific sea surface temperature (SST) anomalies. The ocean dynamics is an extension of the McCreary baroclinic modal model to include varying stratification and certain nonlinear effects. A standard configuration is chosen with 10 baroclinic modes plus two surface layers, which are governed by Ekman dynamics and simulate the combined effects of the higher baroclinic modes from 11 to 30. A nonlinear correction associated with vertical advection of zonal momentum is incorporated and applied (diagnostically) only within the two surface layers, forced by the linear part through nonlinear advection terms. As a result of these improvements, the model realistically simulates the mean equatorial circulation and its variability. The ocean thermodynamics include an SST anomaly model with an empirical parameterization for the temperature of subsurface water entrained into the mixed layer (Te), which is optimally calculated in terms of sea surface height (SSH) anomalies using an empirical orthogonal function (EOF) analysis technique from historical data. The ocean model is then coupled to a statistical atmospheric model that estimates wind stress (τ) anomalies based on a singular value decomposition (SVD) analysis between SST anomalies observed and τ anomalies simulated from ECHAM4.5 (24-member ensemble mean). The coupled system exhibits realistic interannual variability associated with El Niño, including a predominant standing pattern of SST anomalies along the equator and coherent phase relationships among different atmosphere–ocean anomaly fields with a dominant 3-yr oscillation period.Twelve-month hindcasts/forecasts are made during the period 1963–2002, starting each month. Only observed SST anomalies are used to initialize the coupled predictions. As compared to other prediction systems, this coupled model has relatively small systematic errors in the predicted SST anomalies, and its SST prediction skill is apparently competitive with that of most advanced coupled systems incorporating sophisticated ocean data assimilation. One striking feature is that the model skill surpasses that of persistence at all lead times over the central equatorial Pacific. Prediction skill is strongly dependent on the season, with the correlations attaining a minimum in spring and a maximum in fall. Cross-validation experiments are performed to examine the sensitivity of the prediction skill to the data periods selected for training the empirical Te model. It is demonstrated that the artificial skill introduced by using a dependently constructed Te model is not significant. Independent forecasts are made for the period 1997–2002 when no dependent data are included in constructing the two empirical models (Te and τ). The coupled model has reasonable success in predicting transition to warm phase and to cold phase in the spring of 1997 and 1998, respectively. Potential problems and further improvements are discussed with the new intermediate prediction system.

The BMRC Coupled General Circulation Model ENSO Forecast System

An El Nino-Southern Oscillation (ENSO) prediction system with a coupled general circulation model and an ocean data assimilation scheme has been developed at the Australian Bureau of Meteorology Research Centre (BMRC). The coupled model consists of an R21L9 version of the BMRC climate model and a global version of the Geophysical Fluid Dynamics Laboratory modular ocean general circulation model with resolution focused in the tropical region and 25 vertical levels. A univariate statistical interpolation method, with 10-day data ingestion windows, is used to assimilate ocean temperature data and initialize the coupled model. The coupling procedure does not use any flux corrections. Hindcasts have been carried out for the period 1981-95 for each season (60 in all), for up to a lead time of 12 months. This paper will describe these initial experiments and show that the skill of sea surface temperature (SST) hindcasts in the tropical Pacific is comparable to other published coupled models. The skill of the model is strongest in the central Pacific. SST skill tends to be lower during the earlier 1990s than during 1980s in the eastern Pacific but not in the central Pacific. Since the ENSO SST anomaly in the central Pacific is the most important forcing of regional and global climate anomalies, the high SST prediction skill and its insensitivity over the hindcast period in this region in this model give grounds for optimism in the use of coupled general circulation models.