Initial Sea Ice Research Articles

Majority of Southern Ocean Seasonal Sea Ice Zone Bloom Net Community Production Precedes Total Ice Retreat

AbstractThe Southern Ocean seasonal sea ice zone (SIZ) spring is characterized by sea ice retreat and the development of phytoplankton blooms. Until recently, assessing SIZ blooms and associated net community production (bNCP) has been limited by a lack of under‐ice observations. We relate the timing of phytoplankton growth to the drawdown of surface nitrate and sea ice cover and estimate bNCP from biogeochemical profiling float observations. The onset of biological production closely follows initial sea ice breakup and, on average, 64% of bNCP occurs before total sea ice retreat. This indicates that satellite‐derived estimates largely miss under‐ice production and underestimate SIZ bNCP. Float bNCP estimates range from <1 to >4 mol C m−2 bloom−1, with higher bNCP when sea ice breakup occurs early in the year, and the highest bNCP near topographic features that may increase micronutrient supply. Our results suggest changes in Southern Ocean sea ice will influence future bNCP.

Causes and evolution of winter polynyas north of Greenland

Abstract. During the 42-year period (1979–2020) of satellite measurements, four major winter (December–March) polynyas have been observed north of Greenland: one in December 1986 and three in the last decade, i.e., February of 2011, 2017, and 2018. The 2018 polynya was unparalleled in its magnitude and duration compared to the three previous events. Given the apparent recent increase in the occurrence of these extreme events, this study aims to examine their evolution and causality, in terms of forced versus natural variability. The limited weather station and remotely sensed sea ice data are analyzed combining with output from the fully coupled Regional Arctic System Model (RASM), including one hindcast and two ensemble simulations. We found that neither the accompanying anomalous warm surface air intrusion nor the ocean below had an impact (i.e., no significant ice melting) on the evolution of the observed winter open-water episodes in the region. Instead, the extreme atmospheric wind forcing resulted in greater sea ice deformation and transport offshore, accounting for the majority of sea ice loss in all four polynyas. Our analysis suggests that strong southerly winds (i.e., northward wind with speeds greater than 10 m s−1) blowing persistently over the study region for at least 2 d or more were required over the study region to mechanically redistribute some of the thickest Arctic sea ice out of the region and thus to create open-water areas (i.e., a latent heat polynya). To assess the role of internal variability versus external forcing of such events, we carried out and examined results from the two RASM ensembles dynamically downscaled with output from the Community Earth System Model (CESM) Decadal Prediction Large Ensemble (DPLE) simulations. Out of 100 winters in each of the two ensembles (initialized 30 years apart: one in December 1985 and another in December 2015), 17 and 16 winter polynyas were produced north of Greenland, respectively. The frequency of polynya occurrence had no apparent sensitivity to the initial sea ice thickness in the study area pointing to internal variability of atmospheric forcing as a dominant cause of winter polynyas north of Greenland. We assert that dynamical downscaling using a high-resolution regional climate model offers a robust tool for process-level examination in space and time, synthesis with limited observations, and probabilistic forecasts of Arctic events, such as the ones being investigated here and elsewhere.

The Arctic Surface Heating Efficiency of Tropospheric Energy Flux Events

Abstract This paper examines the processes that drive Arctic anomalous surface warming and sea ice loss during winter-season tropospheric energy flux events, synoptic periods of increased tropospheric energy flux convergence (Ftrop), using the NASA MERRA-2 reanalysis. During an event, a poleward anomaly in Ftrop initially increases the sensible and latent energy of the Arctic troposphere; as the warm and moist troposphere loses heat, the anomalous energy source is balanced by a flux upward across the tropopause and a downward net surface flux. A new metric for the Arctic surface heating efficiency (Etrop) is defined, which measures the fraction of the energy source that reaches the surface. Composites of high-, medium-, and low-efficiency events help identify key physical factors, including the vertical structure of Ftrop and Arctic surface preconditioning. In high-efficiency events (Etrop ≥ 0.63), a bottom-heavy poleward Ftrop occurs in the presence of an anomalously warm and unstratified Arctic—a consequence of decreased sea ice—resulting in increased vertical mixing, enhanced near-surface warming and moistening, and further sea ice loss. Smaller Etrop, and thus weaker surface impacts, are found in events with anomalously large initial sea ice extent and more vertically uniform Ftrop. These differences in Etrop are manifested primarily through turbulent heat fluxes rather than downward longwave radiation. The frequency of high-efficiency events has increased from the period 1980–99 to the period 2000–19, contributing to Arctic surface warming and sea ice decline.

Benefits of sea ice initialization for the interannual-to-decadal climate prediction skill in the Arctic in EC-Earth3

Abstract. A substantial part of Arctic climate predictability at interannual timescales stems from the knowledge of the initial sea ice conditions. Among all sea ice properties, its volume, which is a product of sea ice concentration (SIC) and thickness (SIT), is the most responsive parameter to climate change. However, the majority of climate prediction systems are only assimilating the observed SIC due to lack of long-term reliable global observation of SIT. In this study, the EC-Earth3 Climate Prediction System with anomaly initialization to ocean, SIC and SIT states is developed. In order to evaluate the regional benefits of specific initialized variables, three sets of retrospective ensemble prediction experiments are performed with different initialization strategies: ocean only; ocean plus SIC; and ocean plus SIC and SIT initialization. In the Atlantic Arctic, the Greenland–Iceland–Norway (GIN) and Barents seas are the two most skilful regions in SIC prediction for up to 5–6 lead years with ocean initialization; there are re-emerging skills for SIC in the Barents and Kara seas in lead years 7–9 coinciding with improved skills of sea surface temperature (SST), reflecting the impact of SIC initialization on ocean–atmosphere interactions for interannual-to-decadal timescales. For the year 2–9 average, the region with significant skill for SIT is confined to the central Arctic Ocean, covered by multi-year sea ice (CAO-MYI). Winter preconditioning with SIT initialization increases the skill for September SIC in the eastern Arctic (e.g. Kara, Laptev and East Siberian seas) and in turn improve the skill of air surface temperature locally and further expanded over land. SIT initialization outperforms the other initialization methods in improving SIT prediction in the Pacific Arctic (e.g. East Siberian and Beaufort seas) in the first few lead years. Our results suggest that as the climate warming continues and the central Arctic Ocean might become seasonal ice free in the future, the controlling mechanism for decadal predictability may thus shift from sea ice volume to ocean-driven processes.

Summer Extreme Cyclone Impacts on Arctic Sea Ice

AbstractIn this study the impact of extreme cyclones on Arctic sea ice in summer is investigated. Examined in particular are relative thermodynamic and dynamic contributions to sea ice volume budgets in the vicinity of Arctic summer cyclones in 2012 and 2016. Results from this investigation illustrate that sea ice loss in the vicinity of the cyclone trajectories during each year was associated with different dominant processes: thermodynamic processes (melting) in the Pacific sector of the Arctic in 2012, and both thermodynamic and dynamic processes in the Pacific sector of the Arctic in 2016. Comparison of both years further suggests that the Arctic minimum sea ice extent is influenced by not only the strength of the cyclone, but also by the timing and location relative to the sea ice edge. Located near the sea ice edge in early August in 2012, and over the central Arctic later in August in 2016, extreme cyclones contributed to comparable sea ice area (SIA) loss, yet enhanced sea ice volume loss in 2012 relative to 2016. Central to a characterization of extreme cyclone impacts on Arctic sea ice from the perspective of thermodynamic and dynamic processes, we present an index describing relative thermodynamic and dynamic contributions to sea ice volume changes. This index helps to quantify and improve our understanding of initial sea ice state and dynamical responses to cyclones in a rapidly warming Arctic, with implications for seasonal ice forecasting, marine navigation, coastal community infrastructure, and designation of protected and ecologically sensitive marine zones.

Causes and impacts of sea ice variability in the sea of Okhotsk using CESM-LE

This study provides a holistic view of the coupled ocean-atmosphere-sea ice processes responsible for generating interannual variability in sea ice coverage in the Sea of Okhotsk as well as the atmospheric response to this variability. Simulations from the Community Earth System Model Large Ensemble project are analyzed, providing the ability to elucidate the time evolution of these relationships through weekly lead-lag composite analysis, while maintaining a large number of samples to provide robust conclusions. We find that thermodynamic processes involving anomalous ocean-atmosphere heat fluxes affect the timing of initial sea ice growth in the Sea of Okhotsk as early as November. Low-level wind anomalies in the winter affect the extent to which sea ice fully develops, both through advection of the sea ice itself and through changes in the transport of air masses over the Sea of Okhotsk. In this study, the results synthesize and support a diverse set of mechanisms identified in previous observational studies to be responsible for anomalous sea ice conditions, but in a coupled global climate model framework with a large sample size. We also find evidence that anomalous ocean-atmosphere heat fluxes in the winter can trigger an atmospheric response comprised of a local negative sea-level pressure anomaly and Rossby wave that extends over North America. The sign of the turbulent heat fluxes relative to the sea ice anomalies confirm that this is indeed a lagged response of the atmosphere forced by sea ice anomalies. This validates the Rossby wave train response identified in more idealized model simulations with prescribed sea ice and sea surface temperature by demonstrating that this process also occurs in a more realistic coupled model framework.

Seasonal Arctic Sea Ice Prediction Using a Newly Developed Fully Coupled Regional Model With the Assimilation of Satellite Sea Ice Observations

AbstractTo increase our capability to predict Arctic sea ice and climate, we have developed a coupled atmosphere‐sea ice‐ocean model configured for the pan‐Arctic with sufficient flexibility. The Los Alamos Sea Ice Model is coupled with the Weather Research and Forecasting Model and the Regional Ocean Modeling System in the Coupled Ocean‐Atmosphere‐Wave‐Sediment Transport modeling system. It is well known that dynamic models used to predict Arctic sea ice at short‐term periods strongly depend on model initial conditions. Parallel Data Assimilation Framework is implemented into the new modeling system to assimilate sea ice observations and generate skillful model initialization, which aid in the prediction procedures. The Special Sensor Microwave Imager/Sounder sea ice concentration, the CyroSat‐2, and Soil Moisture and Ocean Salinity sea ice thickness are assimilated with the localized error subspace transform ensemble Kalman filter. We conduct Arctic sea ice prediction for the melting seasons of 2017 and 2018. Predictions with improved initial sea ice conditions show reasonable sea ice evolution and small biases in the minimum sea ice extent, although the ice refreezing is delayed. Our prediction experiments suggest that the use of appropriate uncertainty for the observed sea ice thickness can lead to improved spatial distribution of the initial ice thickness and thus the predicted sea ice distribution. Our new modeling system initialized by the output of the National Centers for Environmental Prediction Climate Forecast System seasonal forecasts with data assimilation can significantly increase the sea ice prediction skills in sea ice extent for the entire Arctic as well as in the Northern Sea Route compared with the predictions by the National Centers for Environmental Prediction Climate Forecast System.

Climate models disagree on the sign of total radiative feedback in the Arctic

Climate feedbacks have been found to strongly impact the observed amplified Arctic warming. However, Arctic amplification is modeled with a wide spread which partly arises from intermodel differences of the various feedbacks. To explain the spread in modeled Arctic warming, feedback uncertainties and their origins are investigated in 13 climate models in an experiment with abruptly quadrupled CO2. While intermodel differences in the cloud feedback, being strongest in the Tropics, have been found to determine the spread of global mean effective climate sensitivity, we find that in the Arctic the cloud feedback is not responsible for the spread of Arctic warming as its contribution is too small. Instead, the spread of Arctic warming is explained by differing estimates of surface albedo and Planck feedbacks which show the largest intermodel differences. Our results indicate that these uncertainties not only arise from different degrees of simulated Arctic warming but also are partly related to the large differences in initial sea ice cover and surface temperatures which contribute to the increased spread in estimated warming compared to lower latitudes. Further investigations of feedback dependencies to the base state are needed to constrain the impact of initial uncertainties and to obtain robust results. The most significant distinction between models is the sign of the total feedback parameter. While all models investigated here simulate a negative global mean total feedback, only half of them also show negative Arctic feedbacks which implies that Arctic local feedbacks alone suffice to stably adjust Arctic surface temperatures in response to a radiative perturbation. The other half exhibits positive total Arctic feedbacks indicating local runaway systems which need to be balanced by decreased meridional heat transports. Whether or not a model features such a behaviour depends upon the strength of the simulated positive surface albedo versus the negative Planck feedback.

The potential of sea ice leads as a predictor for summer Arctic sea ice extent

Abstract. The Arctic sea ice extent throughout the melt season is closely associated with initial sea ice state in winter and spring. Sea ice leads are important sites of energy fluxes in the Arctic Ocean, which may play an important role in the evolution of Arctic sea ice. In this study, we examine the potential of sea ice leads as a predictor for summer Arctic sea ice extent forecast using a recently developed daily sea ice lead product retrieved from the Moderate-Resolution Imaging Spectroradiometer (MODIS). Our results show that July pan-Arctic sea ice extent can be predicted from the area of sea ice leads integrated from midwinter to late spring, with a prediction error of 0.28 million km2 that is smaller than the standard deviation of the observed interannual variability. However, the predictive skills for August and September pan-Arctic sea ice extent are very low. When the area of sea ice leads integrated in the Atlantic and central and west Siberian sector of the Arctic is used, it has a significantly strong relationship (high predictability) with both July and August sea ice extent in the Atlantic and central and west Siberian sector of the Arctic. Thus, the realistic representation of sea ice leads (e.g., the areal coverage) in numerical prediction systems might improve the skill of forecast in the Arctic region.

Study on Optimization of Sea Ice Concentration with Adjoint Method

Liu, X. and Zhang, L., 2018. Study on optimization of sea ice concentration with adjoint method. In: Wang, D. and Guido-Aldana, P.A. (eds.), Select Proceedings from the 3rd International Conference on Water Resource and Environment (WRE2017). Journal of Coastal Research, Special Issue No. 84, pp. 44–50. Coconut Creek (Florida), ISSN 0749-0208.Obtaining initial sea ice concentration (SIC) values with high accuracy that are consistent with other models have been a hot topic in both sea ice prediction and sea ice modeling studies. Here, an ocean-sea ice coupled numerical model and its adjoint code have been utilized to carry out numerical experiments to optimize the initial SIC values. In the experiments, the cost function was defined as the difference between the SIC values from the reanalysis dataset and the modeled results. The gradient of cost function, relative to SIC and other model variables, was computed by the adjoint model, and a linear search algorithm was employed to optimize the SIC values by minimizing the cost function. The influences of the weight coefficients of the cost function, the extent of the geographical region, and the seawater temperature and sea ice thickness initial values on the optimization results have been analyzed. The weight coefficients of the cost function had little effect on the SIC distribution pattern but substantial influence on the SIC values. The optimized SIC in the Greenland Sea, Okhotsk Sea, and the Arctic Ocean, with a constant weight coefficient, is better than that with variable weight coefficients. The errors in the initial model fields, other than SIC, may deteriorate the overall result, implying that optimizing multiple model fields simultaneously may improve the optimization effect. Decreasing the size of the geographical region for optimization does not improve the SIC optimization results substantially. Compared to the results from a global cost function, the Barents Sea SIC values from a northern hemispheric cost function are poorly optimized.

Impacts of Assimilating Satellite Sea Ice Concentration and Thickness on Arctic Sea Ice Prediction in the NCEP Climate Forecast System

Here sea ice concentration derived from the Special Sensor Microwave Imager/Sounder and thickness derived from the Soil Moisture and Ocean Salinity and CryoSat-2 satellites are assimilated in the National Centers for Environmental Prediction Climate Forecast System using a localized error subspace transform ensemble Kalman filter (LESTKF). Three ensemble-based hindcasts are conducted to examine impacts of the assimilation on Arctic sea ice prediction, including CTL (without any assimilation), LESTKF-1 (with initial sea ice assimilation only), and LESTKF-E5 (with every 5-day sea ice assimilation). Assessment with the assimilated satellite products and independent sea ice thickness datasets shows that assimilating sea ice concentration and thickness leads to improved Arctic sea ice prediction. LESTKF-1 improves sea ice forecast initially. The initial improvement gradually diminishes after ~3-week integration for sea ice extent but remains quite steady through the integration for sea ice thickness. Large biases in both the ice extent and thickness in CTL are remarkably reduced through the hindcast in LESTKF-E5. Additional numerical experiments suggest that the hindcast with sea ice thickness assimilation dramatically reduces systematic bias in the predicted ice thickness compared with sea ice concentration assimilation only or without any assimilation, which also benefits the prediction of sea ice extent and concentration due to their covariability. Hence, the corrected state of sea ice thickness would aid in the forecast procedure. Increasing the number of ensemble members or extending the integration period to generate estimates of initial model states and uncertainties seems to have small impacts on sea ice prediction relative to LESTKF-E5.

Impact of Model Physics on Seasonal Forecasts of Surface Air Temperature in the Arctic

The impacts of model physics and initial sea ice thickness on seasonal forecasts of surface energy budget and air temperature in the Arctic during summer were investigated based on Climate Forecast System, version 2 (CFSv2), simulations. The model physics changes include the enabling of a marine stratus cloud scheme and the removal of the artificial upper limit on the bottom heat flux from ocean to sea ice. The impact of initial sea ice thickness was examined by initializing the model with relatively realistic sea ice thickness generated by the Pan-Arctic Ice Ocean Modeling and Assimilation System (PIOMAS). Model outputs were compared to that from a control run that did not impose physics changes and used Climate Forecast System Reanalysis (CFSR) sea ice thickness. After applying the physics modification to either sea ice thickness initialization, the simulated total cloud cover more closely resembled the observed monthly variations of total cloud cover except for the midsummer reduction. Over the Chukchi–Bering Seas, the model physics modification reduced the seasonal forecast bias in surface air temperature by 24%. However, the use of initial PIOMAS sea ice thickness alone worsened the surface air temperature predictions. The experiment with physics modifications and initial PIOMAS sea ice thickness achieves the best surface air temperature improvement over the Chukchi–Bering Seas where the area-weighted forecast bias was reduced by 71% from 1.05 K down to −0.3 K compared with the control run. This study supports other results that surface temperatures and sea ice characteristics are highly sensitive to the Arctic cloud and radiation formulations in models and need priority in model formulation and validation.

Skillful seasonal forecasts of Arctic sea ice retreat and advance dates in a dynamical forecast system

AbstractThe need for skillful seasonal forecasts of Arctic sea ice is rapidly increasing. Technology to perform such forecasts with coupled atmosphere‐ocean‐sea ice systems has only recently become available, with previous skill evaluations mainly limited to area‐integrated quantities. Here we show, based on a large set of retrospective ensemble model forecasts, that a dynamical forecast system produces skillful seasonal forecasts of local sea ice retreat and advance dates—variables that are of great interest to a wide range of end users. Advance dates can generally be skillfully predicted at longer lead times (~5 months on average) than retreat dates (~3 months). The skill of retreat date forecasts mainly stems from persistence of initial sea ice anomalies, whereas advance date forecasts benefit from longer time scale and more predictable variability in ocean temperatures. These results suggest that further investments in the development of dynamical seasonal forecast systems may result in significant socioeconomic benefits.

Improving Arctic Sea Ice Prediction Using PIOMAS Initial Sea Ice Thickness in a Coupled Ocean–Atmosphere Model

Abstract Because sea ice thickness is known to influence future patterns of sea ice concentration, it is likely that an improved initialization of sea ice thickness in a coupled ocean–atmosphere model would improve Arctic sea ice cover forecasts. Here, two sea ice thickness datasets as possible candidates for forecast initialization were investigated: the Climate Forecast System Reanalysis (CFSR) and the Pan-Arctic Ice Ocean Modeling and Assimilation System (PIOMAS). Using Ice, Cloud, and Land Elevation Satellite (ICESat) data, it was shown that the PIOMAS dataset had a more realistic representation of sea ice thickness than CFSR. Subsequently, both March CFSR and PIOMAS sea ice thicknesses were used to initialize hindcasts using the Climate Forecast System, version 2 (CFSv2), model. A second set of model runs was also done in which the original model physics were modified to more physically reasonable settings—namely, increasing the number of marine stratus clouds in the Arctic region and enabling realistic representation of the ice–ocean heat flux. Hindcasts were evaluated using sea ice concentration observations from the National Aeronautics and Space Administration (NASA) Team and Bootstrap algorithms. Results show that using PIOMAS initial sea ice thickness in addition to the physics modifications yielded significant improvement in the prediction of September Arctic sea ice extent along with increased interannual predictive skill. Significant local improvements in sea ice concentration were also seen in distinct regions for the change to PIOMAS initial thickness or the physics adjustments, with the most improvement occurring when these changes were applied concurrently.

Model forecast skill and sensitivity to initial conditions in the seasonal Sea Ice Outlook

AbstractWe explore the skill of predictions of September Arctic sea ice extent from dynamical models participating in the Sea Ice Outlook (SIO). Forecasts submitted in August, at roughly 2 month lead times, are skillful. However, skill is lower in forecasts submitted to SIO, which began in 2008, than in hindcasts (retrospective forecasts) of the last few decades. The multimodel mean SIO predictions offer slightly higher skill than the single‐model SIO predictions, but neither beats a damped persistence forecast at longer than 2 month lead times. The models are largely unsuccessful at predicting each other, indicating a large difference in model physics and/or initial conditions. Motivated by this, we perform an initial condition sensitivity experiment with four SIO models, applying a fixed −1 m perturbation to the initial sea ice thickness. The significant range of the response among the models suggests that different model physics make a significant contribution to forecast uncertainty.

Model sensitivity of the Weddell and Ross seas, Antarctica, to vertical mixing and freshwater forcing

We examine the sensitivity of the Weddell and Ross seas to vertical mixing and surface freshwater forcing using an ocean–sea ice model. The high latitude Southern Ocean is very weakly stratified, with a winter salinity difference across the pycnocline of only ∼0.2 PSU. We find that insufficient vertical mixing, freshwater supply from the Antarctic Ice Sheet, or initial sea ice causes a high salinity bias in the mixed layer which erodes the stratification and causes excessive deep convection. This leads to vertical homogenisation of the Weddell and Ross seas, opening of polynyas in the sea ice and unrealistic spin-up of the subpolar gyres and Antarctic Circumpolar Current. The model freshwater budget shows that a ∼30% error in any component can destratify the ocean in about a decade. We find that freshwater forcing in the model should be sufficient along the Antarctic coastline to balance a salinity bias caused by dense coastal water that is unable to sink to the deep ocean. We also show that a low initial sea ice area introduces a salinity bias in the marginal ice zone. We demonstrate that vertical mixing, freshwater forcing and initial sea ice conditions need to be constrained simultaneously to reproduce the Southern Ocean hydrography, circulation and sea ice in a model. As an example, insufficient vertical mixing will cause excessive convection in the Weddell and Ross seas even in the presence of large surface freshwater forcing and initial sea ice cover.

전지구 해양·해빙예측시스템 NEMO-CICE/NEMOVAR의 북극 영역 해빙초기조건 특성 분석

s: In this study, the characteristics of sea ice initial conditions generated from a global ocean and sea ice prediction system, the Nucleus for European Modeling of the Ocean (NEMO) - Los Alamos Sea Ice Model (CICE)/ NEMOVAR were analyzed for the period June 2013 to May 2014 over the Arctic region. For the purpose, the observed and reanalyzed data were used to compare with the sea ice initial conditions. Results indicated that the variability of the monthly sea ice extent and thickness in model initial conditions were well represented as compared to the observation, while it was found that the extent and thickness of Arctic sea ice in initial data were narrower and thinner than those in reanalysis and observation for the period. The reason for the narrower sea ice extent in model initial conditions seems to be due to the fact that the initial sea ice concentration at the boundary area of sea ice was about 20 percent less than the reanalysis data. Also, the reason for the thinner sea-ice thickness in the Arctic region is due to the underestimation of Arctic sea ice thickness (about 60 cm) of the model initial conditions in the Arctic Ocean area adjacent to Greenland and Arctic archipelago where thick sea ice appears all the year round.

Prospects for improved seasonal Arctic sea ice predictions from multivariate data assimilation

Predicting the summer Arctic sea ice conditions a few months in advance has become a challenging priority. Seasonal prediction is partly an initial condition problem; therefore, a good knowledge of the initial sea ice state is necessary to hopefully produce reliable forecasts. Most of the intrinsic memory of sea ice lies in its thickness, but consistent and homogeneous observational networks of sea ice thickness are still limited in space and time. To overcome this problem, we constrain the ocean–sea ice model NEMO-LIM3 with gridded sea ice concentration retrievals from satellite observations using the ensemble Kalman filter. No sea ice thickness products are assimilated. However, thanks to the multivariate formalism of the data assimilation method used, sea ice thickness is globally updated in a consistent way whenever observations of concentration are available. We compare in this paper the skill of 27 pairs of initialized and uninitialized seasonal Arctic sea ice hindcasts spanning 1983–2009, driven by the same atmospheric forcing as to isolate the pure role of initial conditions on the prediction skill. The results exhibit the interest of multivariate sea ice initialization for the seasonal predictions of the September ice concentration and are particularly encouraging for hindcasts in the 2000s. In line with previous studies showing the interest of data assimilation for sea ice thickness reconstruction, our results thus show that sea ice data assimilation is also a promising tool for short-term prediction, and that current seasonal sea ice forecast systems could gain predictive skill from a more realistic sea ice initialization.

Ensemble of sea ice initial conditions for interannual climate predictions

Polar climate studies are severely hampered by the sparseness of the sea ice observations. We aim at filling this critical gap by producing two 5-member sea ice historical simulations strongly constrained by ocean and atmosphere observational data and covering the 1958–2006 and 1979–2012 periods. This is the first multi-member sea ice reconstruction covering more than 50 years. The obtained sea ice conditions are in reasonable agreement with the few available observations. These best estimates of sea ice conditions serve subsequently as initial sea ice conditions for a set of 28 3-year-long retrospective climate predictions. We compare it to a set in which the sea ice initial conditions are taken from a single-member sea ice historical simulation constrained by atmosphere observations only. We find an improved skill in predicting the Arctic sea ice area and Arctic near surface temperature but a slightly degraded skill in predicting the Antarctic sea ice area. We also obtain a larger spread between the members for the sea ice variables, thus more representative of the forecast error.

Seasonal Prediction of Arctic Sea Ice Extent from a Coupled Dynamical Forecast System

Abstract While fully coupled atmosphere–ocean models have been used to study the seasonal predictability of sea ice variations within the context of models’ own variability, their capability in predicting the observed sea ice at the seasonal time scales is not well assessed. In this study, sea ice predictions from the recently developed NCEP Climate Forecast System, version 2 (CFSv2), a fully coupled atmosphere–ocean model including an interactive dynamical sea ice component, are analyzed. The focus of the analysis is the performance of CFSv2 in reproducing observed Northern Hemisphere sea ice extent (SIE). The SIE climatology, long-term trend, interannual variability, and predictability are assessed. CFSv2 contains systematic biases that are dependent more on the forecast target month than the initial month, with a positive SIE bias for the forecast for January–September and a negative SIE bias for the forecast for October–December. A large source of seasonal prediction skill is from the long-term trend, which is underestimated in the CFSv2. Prediction skill of interannual SIE anomalies is found to be primarily within the first three target months and is largest in the summer and early fall. The performance of the prediction of sea ice interannual variations varies from year to year and is found to be related to initial sea ice thickness. Potential predictability based on the forecast ensemble, its dependence on model deficiencies, and implications of the results from this study for improvements in the seasonal sea ice prediction are discussed.