The flexible global ocean-atmosphere-land system model, Grid-point Version 2: FGOALS-g2

This study mainly introduces the development of the Flexible Global Ocean-Atmosphere-Land System Model: Grid-point Version 2 (FGOALS-g2) and the preliminary evaluations of its performances based on results from the pre-industrial control run and four members of historical runs according to the fifth phase of the Coupled Model Intercomparison Project (CMIP5) experiment design. The results suggest that many obvious improvements have been achieved by the FGOALS-g2 compared with the previous version,FGOALS-g1, including its climatological mean states, climate variability, and 20th century surface temperature evolution. For example,FGOALS-g2 better simulates the frequency of tropical land precipitation, East Asian Monsoon precipitation and its seasonal cycle, MJO and ENSO, which are closely related to the updated cumulus parameterization scheme, as well as the alleviation of uncertainties in some key parameters in shallow and deep convection schemes, cloud fraction, cloud macro/microphysical processes and the boundary layer scheme in its atmospheric model. The annual cycle of sea surface temperature along the equator in the Pacific is significantly improved in the new version. The sea ice salinity simulation is one of the unique characteristics of FGOALS-g2, although it is somehow inconsistent with empirical observations in the Antarctic.     

FGOALS_gg1.1极地气候模拟

对中国科学院大气物理研究所大气科学和地球流体力学数值模拟国家重点实验室发展的气候系统模式FGOALS_g1.1的极地气候模拟现状进行了较为全面的评估.结果表明,FGOALS_g1.1对南北极海冰的主要分布特征、季节变化和年代际变化趋势具有一定的模拟能力.但也注意到,与观测相比,模式存在以下几方面的问题:(1)模拟的海冰总面积北极偏多,而南极偏少.北极,北大西洋海冰全年明显偏多;夏季,西伯利亚沿海海冰偏多,而波弗特海海冰偏少.南极,威德尔海和罗斯海冬季海冰偏少.南北极海冰边缘都存在异常的较大范围密集度很小的碎冰区,夏季尤为显著.(2)海冰流速在南北极海冰边缘和南极大陆沿岸附近较大.北极,模式没能模拟出波弗特涡流,并且由于模式网格中北极点的处理问题,造成其附近错误的海冰流场及厚度分布.这些海冰偏差与模式模拟的大气和海洋状况有着密切的联系.进一步分析表明,FGOALS_g1.1模拟的冰岛低压和南极绕极西风带明显偏弱,其通过大气环流和海表面风应力影响向极地的热量输送,在很大程度上导致上述的海冰偏差.此外,耦合模式中大气-海冰-海洋的相互作用可以放大子模式中的偏差.     

BCC-CSM2-MR模式对北极海冰和气候的模拟及预估

基于第六次耦合模式比较计划（CMIP6）,使用新一代全球模式BCC-CSM2-MR的历史试验和未来共享社会经济路径（SSPs）数据,依据Hadley中心的海表面温度和海冰密集度数据及NCEP/NCAR I再分析资料,评估了BCC-CSM2-MR模式对北极海冰及北极气候的模拟能力,并对未来变化进行了预估。结果表明：BCC-CSM2-MR模式可以较好再现北极海冰密集度、近地层大气平均温度和海表温度的多年平均空间分布特征。但模式对北极局地大气平均温度模拟存在一定偏差,可能在一定程度上导致相应地区海冰的模拟存在差异。21世纪,北极海冰范围持续减少,9月减少趋势显著,3月减少趋势相对较弱。3月北极大部地区表现为一致的增温,仅在北大西洋局部出现一定程度的降温,9月北极大气增温幅度弱于3月。与地表平均温度不同,3月和9月的北极大部地区海表温度均出现增加,且9月海表温度的增幅大于3月,仅拉布拉多海海温出现下降。     

Evaluation of Arctic Sea Ice Drift and its Relationship with Near-surface Wind and Ocean Current in Nine CMIP6 Models from China

The simulated Arctic sea ice drift and its relationship with the near-surface wind and surface ocean current during 1979-2014 in nine models from China that participated in the sixth phase of the Coupled Model Intercomparison Project(CMIP6)are examined by comparison with observational and reanalysis datasets.Most of the models reasonably represent the Beaufort Gyre(BG)and Transpolar Drift Stream(TDS)in the spatial patterns of their long-term mean sea ice drift,while the detailed location,extent,and strength of the BG and TDS vary among the models.About two-thirds of the models agree with the observation/reanalysis in the sense that the sea ice drift pattern is consistent with the near-surface wind pattern.About the same proportion of models shows that the sea ice drift pattern is consistent with the surface ocean current pattern.In the observation/reanalysis,however,the sea ice drift pattern does not match well with the surface ocean current pattern.All nine models missed the observational widespread sea ice drift speed acceleration across the Arctic.For the Arctic basin-wide spatial average,five of the nine models overestimate the Arctic long-term(1979-2014)mean sea ice drift speed in all months.Only FGOALS-g3 captures a significant sea ice drift speed increase from 1979 to 2014 both in spring and autumn.The increases are weaker than those in the observation.This evaluation helps assess the performance of the Arctic sea ice drift simulations in these CMIP6 models from China.     

FGOALS-g模式及其参与CMIP6的方案

为参加第六次国际耦合模式比较计划（CMIP6）和进一步提高模式的模拟能力,大气科学和地球流体力学数值模拟国家重点实验室（LASG）模式团队发展了新一代的格点大气版本的FGOALS-g耦合模式。新版本模式在大气分辨率、海洋网格,以及各分量模式的物理过程等方面都有一定的改进,并正在参与CMIP6最核心的试验以及多个CMIP6模式比较子计划试验。给定CMIP6外强迫,模式在工业革命前参照试验（piControl）和大气模式比较计划（AMIP）试验中模拟的初步结果都比较合理。     

Simulation and Improvements of Oceanic Circulation and Sea Ice by the Coupled Climate System Model FGOALS-f3-L

This study documents simulated oceanic circulations and sea ice by the coupled climate system model FGOALS-f3-L developed at the State Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics, Institute of Atmospheric Physics, Chinese Academy of Sciences, under historical forcing from phase 6 of the Coupled Model Intercomparison Project (CMIP6). FGOALS-f3-L reproduces the fundamental features of global oceanic circulations, such as sea surface temperature (SST), sea surface salinity (SSS), mixed layer depth (MLD), vertical temperature and salinity, and meridional overturning circulations. There are notable improvements compared with the previous version, FGOALS-s2, such as a reduction in warm SST biases near the western and eastern boundaries of oceans and salty SSS biases in the tropical western Atlantic and eastern boundaries, and a mitigation of deep MLD biases at high latitudes. However, several obvious biases remain. The most significant biases include cold SST biases in the northwestern Pacific (over 4°C), freshwater SSS biases and deep MLD biases in the subtropics, and temperature and salinity biases in deep ocean at high latitudes. The simulated sea ice shows a reasonable distribution but stronger seasonal cycle than observed. The spatial patterns of sea ice are more realistic in FGOALS-f3-L than its previous version because the latitude–longitude grid is replaced with a tripolar grid in the ocean and sea ice model. The most significant biases are the overestimated sea ice and underestimated SSS in the Labrador Sea and Barents Sea, which are related to the shallower MLD and weaker vertical mixing.     

Higher precision estimates of regional polar warming by ensemble regression of climate model projections

This study presents projections of twenty-first century wintertime surface temperature changes over the high-latitude regions based on the third Coupled Model Inter-comparison Project (CMIP3) multi-model ensemble. The state-dependence of the climate change response on the present day mean state is captured using a simple yet robust ensemble linear regression model. The ensemble regression approach gives different and more precise estimated mean responses compared to the ensemble mean approach. Over the Arctic in January, ensemble regression gives less warming than the ensemble mean along the boundary between sea ice and open ocean (sea ice edge). Most notably, the results show 3?°C less warming over the Barents Sea (~7?°C compared to ~10?°C). In addition, the ensemble regression method gives projections that are 30?% more precise over the Sea of Okhostk, Bering Sea and Labrador Sea. For the Antarctic in winter (July) the ensemble regression method gives 2?°C more warming over the Southern Ocean close to the Greenwich Meridian (~7?°C compared to ~5?°C). Projection uncertainty was almost half that of the ensemble mean uncertainty over the Southern Ocean between 30° W to 90° E and 30?% less over the northern Antarctic Peninsula. The ensemble regression model avoids the need for explicit ad hoc weighting of models and exploits the whole ensemble to objectively identify overly influential outlier models. Bootstrap resampling shows that maximum precision over the Southern Ocean can be obtained with ensembles having as few as only six climate models.     

Identifying uncertainties in Arctic climate change projections

Wide ranging climate changes are expected in the Arctic by the end of the 21st century, but projections of the size of these changes vary widely across current global climate models. This variation represents a large source of uncertainty in our understanding of the evolution of Arctic climate. Here we systematically quantify and assess the model uncertainty in Arctic climate changes in two CO₂ doubling experiments: a multimodel ensemble (CMIP3) and an ensemble constructed using a single model (HadCM3) with multiple parameter perturbations (THC-QUMP). These two ensembles allow us to assess the contribution that both structural and parameter variations across models make to the total uncertainty and to begin to attribute sources of uncertainty in projected changes. We find that parameter uncertainty is an major source of uncertainty in certain aspects of Arctic climate. But also that uncertainties in the mean climate state in the 20th century, most notably in the northward Atlantic ocean heat transport and Arctic sea ice volume, are a significant source of uncertainty for projections of future Arctic change. We suggest that better observational constraints on these quantities will lead to significant improvements in the precision of projections of future Arctic climate change.     

Sea-ice and its response to CO2 forcing as simulated by global climate models

The simulation of sea-ice in global climate models participating in the Coupled Model Intercomparison Project (CMIP1 and CMIP2) is analyzed. CMIP1 simulations are of the unpertubed control climate whereas in CMIP2, all models have been forced with the same 1% yr^–1 increase in CO₂ concentration, starting from a near equilibrium initial condition. These simulations are not intended as forecasts of climate change, but rather provide a means of evaluating the response of current climate models to the same forcing. The difference in modeled response therefore indicates the range (or uncertainty) in model sensitivity to greenhouse gas and other climatic perturbations. The results illustrate a wide range in the ability of climate models to reproduce contemporary sea-ice extent and thickness; however, the errors are not obviously related to the manner in which sea-ice processes are represented in the models (e.g. the inclusion or neglect of sea-ice motion). The implication is that errors in the ocean and atmosphere components of the climate model are at least as important. There is also a large range in the simulated sea-ice response to CO₂ change, again with no obvious stratification in terms of model attributes. In contrast to results obtained earlier with a particular model, the CMIP ensemble yields rather mixed results in terms of the dependence of high-latitude warming on sea-ice initial conditions. There is an indication that, in the Arctic, models that produce thick ice in their control integration exhibit less warming than those with thin ice. The opposite tendency appears in the Antarctic (albeit with low statistical significance). There is a tendency for models with more extensive ice coverage in the Southern Hemisphere to exhibit greater Antarctic warming. Results for the Arctic indicate the opposite tendency (though with low statistical significance).A list of the CMIP modeling groups is included in the Acknowledgements section.     

Evaluation of pan-Arctic melt-freeze onset in CMIP5 climate models and reanalyses using surface observations

The seasonal melt-freeze transitions are fundamental features of the Arctic climate system. The representation of the pan-Arctic melt and freeze onset (north of 60°N) is assessed in two reanalyses and eleven CMIP5 global circulation models (GCMs). The seasonal melt-freeze transitions are retrieved from surface air temperature (SAT) across the land and sea-ice domains and evaluated against surface observations. While monthly averages of SAT are reasonably well represented in models, large model-observation and model–model disparities of timing of melt and freeze onset are evident. The evaluation against surface observations reveals that the ERA-Interim reanalysis performs the best, closely followed by some of the climate models. GCMs and reanalyses capture the seasonal melt-freeze transitions better in the central Arctic than in the marginal seas and across the land areas. The GCMs project that during the 21st century, the summer length—the period between melt and freeze onset—will increase over land by about 1 month at all latitudes, and over sea ice by 1 and 3 months at low and high latitudes, respectively. This larger summer-length increase over sea ice at progressively higher latitudes is related to a retreat of summer sea ice during the 21st century, since open water freezes roughly 40 days later than ice-covered ocean. As a consequence, by the year 2100, the freeze onset is projected to be initiated within roughly 10 days across the whole Arctic Ocean, whereas this transition varies by about 80 days today.     

Assessment of Snow Depth over Arctic Sea Ice in CMIP6 Models Using Satellite Data

Snow depth over sea ice is an essential variable for understanding the Arctic energy budget.In this study,we evaluate snow depth over Arctic sea ice during 1993-2014 simulated by 31 models from phase 6 of the Coupled Model Intercomparison Project(CMIP6)against recent satellite retrievals.The CMIP6 models capture some aspects of the observed snow depth climatology and variability.The observed variability lies in the middle of the models’simulations.All the models show negative trends of snow depth during 1993-2014.However,substantial spatiotemporal discrepancies are identified.Compared to the observation,most models have late seasonal maximum snow depth(by two months),remarkably thinner snow for the seasonal minimum,an incorrect transition from the growth to decay period,and a greatly underestimated interannual variability and thinning trend of snow depth over areas with frequent occurrence of multi-year sea ice.Most models are unable to reproduce the observed snow depth gradient from the Canadian Arctic to the outer areas and the largest thinning rate in the central Arctic.Future projections suggest that snow depth in the Arctic will continue to decrease from 2015 to 2099.Under the SSP5-8.5 scenario,the Arctic will be almost snow-free during the summer and fall and the accumulation of snow starts from January.Further investigation into the possible causes of the issues for the simulated snow depth by some models based on the same family of models suggests that resolution,the inclusion of a hightop atmospheric model,and biogeochemistry processes are important factors for snow depth simulation.     

FGOALS-g3 Model Datasets for CMIP6 Flux-Anomaly-Forced Model Intercomparison Project

The Flux-Anomaly-Forced Model Intercomparison Project(FAFMIP) is an endorsed Model Intercomparison Project in phase 6 of the Coupled Model Intercomparison Project(CMIP6). The goal of FAFMIP is to investigate the spread in the atmosphere–ocean general circulation model projections of ocean climate change forced by increased CO_2, including the uncertainties in the simulations of ocean heat uptake, global mean sea level rise due to ocean thermal expansion and dynamic sea level change due to ocean circulation and density changes. The FAFMIP experiments have already been conducted with the Flexible Global Ocean–Atmosphere–Land System Model, gridpoint version 3.0(FGOALS-g3). The model datasets have been submitted to the Earth System Grid Federation(ESGF) node. Here, the details of the experiments,the output variables and some baseline results are presented. Compared with the preliminary results of other models, the evolutions of global mean variables can be reproduced well by FGOALS-g3. The simulations of spatial patterns are also consistent with those of other models in most regions except the North Atlantic and the Southern Ocean, indicating large uncertainties in the regional sea level projections of these two regions.     

Arctic climate change in 21st century CMIP5 simulations with EC-Earth

The Arctic climate change is analyzed in an ensemble of future projection simulations performed with the global coupled climate model EC-Earth2.3. EC-Earth simulates the twentieth century Arctic climate relatively well but the Arctic is about 2 K too cold and the sea ice thickness and extent are overestimated. In the twenty-first century, the results show a continuation and strengthening of the Arctic trends observed over the recent decades, which leads to a dramatically changed Arctic climate, especially in the high emission scenario RCP8.5. The annually averaged Arctic mean near-surface temperature increases by 12 K in RCP8.5, with largest warming in the Barents Sea region. The warming is most pronounced in winter and autumn and in the lower atmosphere. The Arctic winter temperature inversion is reduced in all scenarios and disappears in RCP8.5. The Arctic becomes ice free in September in all RCP8.5 simulations after a rapid reduction event without recovery around year 2060. Taking into account the overestimation of ice in the twentieth century, our model results indicate a likely ice-free Arctic in September around 2040. Sea ice reductions are most pronounced in the Barents Sea in all RCPs, which lead to the most dramatic changes in this region. Here, surface heat fluxes are strongly enhanced and the cloudiness is substantially decreased. The meridional heat flux into the Arctic is reduced in the atmosphere but increases in the ocean. This oceanic increase is dominated by an enhanced heat flux into the Barents Sea, which strongly contributes to the large sea ice reduction and surface-air warming in this region. Increased precipitation and river runoff lead to more freshwater input into the Arctic Ocean. However, most of the additional freshwater is stored in the Arctic Ocean while the total Arctic freshwater export only slightly increases.     

An Overview of BCC Climate System Model Development and Application for Climate Change Studies

This paper reviews recent progress in the development of the Beijing Climate Center Climate System Model(BCC-CSM) and its four component models(atmosphere,land surface,ocean,and sea ice).Two recent versions are described:BCC-CSM1.1 with coarse resolution(approximately 2.8125°×2.8125°) and BCC-CSM1.1(m) with moderate resolution(approximately 1.125°×1.125°).Both versions are fully coupled climate-carbon cycle models that simulate the global terrestrial and oceanic carbon cycles and include dynamic vegetation.Both models well simulate the concentration and temporal evolution of atmospheric CO_2 during the 20th century with anthropogenic CO2 emissions prescribed.Simulations using these two versions of the BCC-CSM model have been contributed to the Coupled Model Intercomparison Project phase five(CMIP5) in support of the Intergovernmental Panel on Climate Change(IPCC) Fifth Assessment Report(AR5).These simulations are available for use by both national and international communities for investigating global climate change and for future climate projections.Simulations of the 20th century climate using BCC-CSMl.l and BCC-CSMl.l(m) are presented and validated,with particular focus on the spatial pattern and seasonal evolution of precipitation and surface air temperature on global and continental scales.Simulations of climate during the last millennium and projections of climate change during the next century are also presented and discussed.Both BCC-CSMl.l and BCC-CSMl.l(m) perform well when compared with other CMIP5 models.Preliminary analyses indicate that the higher resolution in BCC-CSM1.1(m) improves the simulation of mean climate relative to BCC-CSMl.l,particularly on regional scales.     

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.     

Oceanic climatology in the coupled model FGOALS-g2: Improvements and biases

The present study examines simulated oceanic climatology in the Flexible Global Ocean-Atmosphere-Land System model, Grid-point Version 2 (FGOALS-g2) forced by historical external forcing data. The oceanic temperatures and circulations in FGOALS-g2 were found to be comparable to those observed, and substantially improved compared to those simulated by the previous version, FGOALS-g1.0. Compared with simulations by FGOALS-g1.0, the shallow mixed layer depths were better captured in the eastern Atlantic and Pacific Ocean in FGOALS-g2. In the high latitudes of the Northern Hemisphere, the cold biases of SST were about 1°C–5°C smaller in FGOALS-g2. The associated sea ice distributions and their seasonal cycles were more realistic in FGOALS-g2. The pattern of Atlantic Meridional Overturning Circulation (AMOC) was better simulated in FGOALS-g2, although its magnitude was larger than that found in observed data. The simulated Antarctic Circumpolar Current (ACC) transport was about 140 Sv through the Drake Passage, which is close to that observed. Moreover, Antarctic Intermediate Water (AAIW) was better captured in FGOALS-g2. However, large SST cold biases (>3°C) were still found to exist around major western boundary currents and in the Barents Sea, which can be explained by excessively strong oceanic cold advection and unresolved processes owing to the coarse resolution. In the Indo-Pacific warm pool, the cold biases were partly related to the excessive loss of heat from the ocean. Along the eastern coast in the Atlantic and Pacific Oceans, the warm biases were due to overestimation of shortwave radiation. In the Indian Ocean and Southern Ocean, the surface fresh biases were mainly due to the biases of precipitation. In the tropical Pacific Ocean, the surface fresh biases (>2 psu) were mainly caused by excessive precipitation and oceanic advection. In the Indo-Pacific Ocean, fresh biases were also found to dominate in the upper 1000 m, except in the northeastern Indian Ocean. There were warm and salty biases (3°C–4°C and 1–2 psu) from the surface to the bottom in the Labrador Sea, which might be due to large amounts of heat transport and excessive evaporation, respectively. For vertical structures, the maximal biases of temperature and salinity were found to be located at depths of >600 m in the Arctic Ocean, and their values exceeded 4°C and 2 psu, respectively.     

Arctic Climate Changes Based on Historical Simulations (1900‒2013) with the CAMS-CSM

The Chinese Academy of Meteorological Sciences Climate System Model (CAMS-CSM) is a newly developed global climate model that will participate in the Coupled Model Intercomparison Project phase 6. Based on historical simulations (1900?2013), we evaluate the model performance in simulating the observed characteristics of the Arctic climate system, which includes air temperature, precipitation, the Arctic Oscillation (AO), ocean temperature/salinity, the Atlantic meridional overturning circulation (AMOC), snow cover, and sea ice. The model?data comparisons indicate that the CAMS-CSM reproduces spatial patterns of climatological mean air temperature over the Arctic (60°?90°N) and a rapid warming trend from 1979 to 2013. However, the warming trend is overestimated south of the Arctic Circle, implying a subdued Arctic amplification. The distribution of climatological precipitation in the Arctic is broadly captured in the model, whereas it shows limited skills in depicting the overall increasing trend. The AO can be reproduced by the CAMS-CSM in terms of reasonable patterns and variability. Regarding the ocean simulation, the model underestimates the AMOC and zonally averaged ocean temperatures and salinity above a depth of 500 m, and it fails to reproduce the observed increasing trend in the upper ocean heat content in the Arctic. The large-scale distribution of the snow cover extent (SCE) in the Northern Hemisphere and the overall decreasing trend in the spring SCE are captured by the CAMS-CSM, while the biased magnitudes exist. Due to the underestimation of the AMOC and the poor quantification of air–sea interaction, the CAMS-CSM overestimates regional sea ice and underestimates the observed decreasing trend in Arctic sea–ice area in September. Overall, the CAMS-CSM reproduces a climatological distribution of the Arctic climate system and general trends from 1979 to 2013 compared with the observations, but it shows limited skills in modeling local trends and interannual variability.     

Preface

<正>The Flexible Global Ocean-Atmosphere-Land System model (FGOALS) is a coupled climate model that allows researchers to conduct fundamental research into the Earth's past, present, near-term and long-term future climate states. FGOALS couples the ocean, atmosphere, land, and sea ice through a coupler that coordinates the component models and passes the exchange of energy, momentum, and water among them. The     

近20年来中国极地大气科学研究进展

南极、北极和青藏高原是地球上的 3大气候敏感地区 ,是多个国际计划研究全球变化的关键地区。中国的南极和北极实地考察研究 ,分别始于 2 0世纪 80和 90年代 ,起步较晚 ,但近 2 0余年来有较大的进展。极地大气科学考察与研究是极地科学研究的重要组成部分。讫今为止 ,中国已组织了 2 0次南极考察和 3次北极考察 ,建立了中国南极长城站、中山站和北极黄河站等 3个常年科学考察站 ;进行了常规地面气象、Brewer大气臭氧、近地面物理、高层大气物理、冰雪和大气化学等观测 ,获得了较为系统的极地大气科学第一手资料 ;开展了有关极地与全球变化的研究 ,取得了新的进展。南极地区大气温度、臭氧和海冰的气候变化在时间和空间上都是多样的。南极地区的增暖主要发生在南极半岛地区 ,在南极大陆主体并不明显 ,近 10余年来还有降温趋势。中国南极长城站和中山站的观测资料也证实了这一点。此外 ,还揭示了南极半岛西侧和罗斯海外围的海冰变化具有“翘翘板”特征 ,由此定义的南极涛动指数可用来讨论南极海冰状况和海冰关键区的活动 ;用实地考察资料研究了极地不同下垫面的近地面物理和海 -冰 -气相互作用特征 ,给出了边界层特征参数 ;讨论了极地天气气候和大气环境特征及其对东亚大气环流和中国天气气候的影响 ;利用     

Interannual predictability of Arctic sea ice in a global climate model: regional contrasts and temporal evolution

The predictability of the Arctic sea ice is investigated at the interannual time scale using decadal experiments performed within the framework of the fifth phase of the Coupled Model Intercomparison Project with the CNRM-CM5.1 coupled atmosphere–ocean global climate model. The predictability of summer Arctic sea ice extent is found to be weak and not to exceed 2 years. In contrast, robust prognostic potential predictability (PPP) up to several years is found for winter sea ice extent and volume. This predictability is regionally contrasted. The marginal seas in the Atlantic sector and the central Arctic show the highest potential predictability, while the marginal seas in the Pacific sector are barely predictable. The PPP is shown to decrease drastically in the more recent period. Regarding sea ice extent, this decrease is explained by a strong reduction of its natural variability in the Greenland–Iceland–Norwegian Seas due to the quasi-disappearance of the marginal ice zone in the center of the Greenland Sea. In contrast, the decrease of predictability of sea ice volume arises from the combined effect of a reduction of its natural variability and an increase in its chaotic nature. The latter is attributed to a thinning of sea ice cover over the whole Arctic, making it more sensitive to atmospheric fluctuations. In contrast to the PPP assessment, the prediction skill as measured by the anomaly correlation coefficient is found to be mostly due to external forcing. Yet, in agreement with the PPP assessment, a weak added value of the initialization is found in the Atlantic sector. Nevertheless, the trend-independent component of this skill is not statistically significant beyond the forecast range of 3 months. These contrasted findings regarding potential predictability and prediction skill arising from the initialization suggest that substantial improvements can be made in order to enhance the prediction skill.