Advanced Search
Article Contents

The Arctic Sea Ice Thickness Change in CMIP6’s Historical Simulations


doi: 10.1007/s00376-022-1460-4

  • This study assesses sea ice thickness (SIT) from the historical run of the Coupled Model Inter-comparison Project Phase 6 (CMIP6). The SIT reanalysis from the Pan-Arctic Ice Ocean Modeling and Assimilation System (PIOMAS) product is chosen as the validation reference data. Results show that most models can adequately reproduce the climatological mean, seasonal cycle, and long-term trend of Arctic Ocean SIT during 1979–2014, but significant inter-model spread exists. Differences in simulated SIT patterns among the CMIP6 models may be related to model resolution and sea ice model components. By comparing the climatological mean and trend for SIT among all models, the Arctic SIT change in different seas during 1979–2014 is evaluated. Under the scenario of historical radiative forcing, the Arctic SIT will probably exponentially decay at –18% (10 yr)–1 and plausibly reach its minimum (equilibrium) of 0.47 m since the 2070s.
    摘要: 本研究将第六次国际耦合模式比较计划(CMIP6)历史试验结果与华盛顿大学泛北冰洋模式与同化系统(PIOMAS)的冰厚再分析数据进行了系统性的比较,评估了各耦合模式对1979-2014年期间北极海冰厚度变化的模拟能力。研究结果表明,大多数模式都能较好地重现再分析的北极海冰厚度气候态、年变化、年际变化和线性趋势,但依然存在明显的模式间差异。这种差异与模式分辨率以及各耦合模式对海冰模式的选择有关。随后,通过比较各模式间北极海冰厚度的气候态平均值和年际变化趋势,对1979-2014年间北极不同海域海冰厚度变化进行了进一步的分析。结果表明,在历史辐射强迫情景下,北极海冰年平均厚度可能以每十年-18%的速度指数衰减,并有可能在20世纪70年代达到0.47 m的平衡值。文章也对北极不同海域的海冰变化特征进行了讨论。
  • 加载中
  • Figure 1.  Spatial pattern of SIT from PIOMAS during 1979–2014. The uppercase letters represent the different seas defined by the NSIDC Arctic Ocean regional masks (https://nsidc.org/data/polar-stereo/tools_masks.html#region_masks). The black line is the boundary among the different seas. (A—Central Arctic; B—Beaufort Sea; C—Chukchi Sea; D—East Siberian Sea; E—Laptev Sea; F—Kara Sea; and G—Barents Sea).

    Figure 2.  The difference of climatological mean SIT between CMIP6 models and PIOMAS during 1979–2014.

    Figure 3.  Taylor Diagram of Arctic SIT for CMIP6 models and PIOMAS during 1979–2014.

    Figure 4.  The boxplot of regional averaged SIT from CMIP6s MMM and PIOMAS in different seas during 1979–2014. The five values in a box from top to bottom are the maximum, 75th percentile, median, 25th percentile, and minimum. The relative deviation of the averaged SIT between CMIP6s MMM and PIOMAS is shown as solid circles (right axis).

    Figure 5.  Taylor Score (TS) against (a) the ranking of model resolution and (b) the sea ice components in CMIP6 models during 1979–2014. The dashed lines indicate a TS of 0.8.

    Figure 6.  Seasonal cycles of (a) SIT and (b) TS from CMIP6 models and PIOMAS during 1979–2014.

    Figure 7.  Evolution of (a) annual mean, (b) May, and (c) September Arctic SIT for CMIP6 models and PIOMAS during 1979–2014.

    Figure 8.  The trend versus the climatological mean of the SIT during 1979–2014 from CMIP6 models and PIOMAS in different seas. x0 and y0 are the X- and Y-axis intercepts, respectively; s is the linear regression coefficient, and p is the confidence level.

    Table 1.  The basic information of CMIP6 models.

    ModelCountry (Institution)Sea Ice Model
    (resolution: longitude × latitude)
    SITANESM3China (NUIST)CICE4.1 (384 × 362)
    MIROC-ES2LJapan (MIROC)COCO4.9 (360 × 256)
    MIROC6
    EC-EARTH3Europe (EC)NEMO-LIM3 (362 × 292)
    EC-EARTH3-Veg
    EC-EARTH3-Veg-LR
    EC-EARTH3-AerChem
    E3SMLLNL (E3SM)MPAS-Seaice v6.0 (360 × 180)
    SITEACCESS-CM2Australia (CSIRO-ARCCSS)CICE5.1.2 (360 × 300)
    ACCESS-ESM1-5Australia (CSIRO)CICE4.1 (360 × 300)
    BCC-CSM2-MRChina (BCC)SIS2 (360 × 232)
    CAMS-CSM1-0China (CAMS)SIS1.0 (320 × 160)
    CESM2USA (NCAR)CICE5.1 (320 × 384)
    CESM2-WACCM
    CIESMChina (THU)CICE4 (320 × 384)
    CMCC-CM2-SR5Italy (CMCC)CICE4.0 (288 × 192)
    CNRM-CM6-1France (CNRM-CERFACS)Gelato6.1 (362 × 294)
    CNRM-CM6-1-HRGelato6.1 (1442 × 1050)
    CNRM-ESM2-1Gelato6.1 (362 × 294)
    EC-EARTH3Europe (EC)NEMO-LIM3 (362 × 292)
    EC-EARTH3-AerChem
    EC-EARTH3-Veg
    EC-EARTH3-Veg-LR
    FGOALS-f3-LChina (CAS)CICE4.0 (360 × 218)
    FIO-ESM-2-0China (FIO-QLNM)CICE4.0 (320 × 384)
    GFDL-ESM4USA (NOAA-GFDL)GFDL-SIM4p5 (720 × 576)
    GISS-E2-1-GUSA (NASA-GISS)GISS SI (144 × 90)
    HadGEM3-GC31-LLUK (MOHC)CICE-HadGEM3-GSI8 (360 × 330)
    HadGEM3-GC31-MMCICE-HadGEM3-GSI8 (1440 × 1205)
    IPSL-CM6A-LRFrance (IPSL)NEMO-LIM3 (362 × 332)
    MPI-ESM-1-2-HAMEurope (HAMMOZ-Consortium)unnamed (256 × 220)
    MPI-ESM1-2-HRGermany (MPI-M)unnamed (384 × 192)
    MPI-ESM1-2-LRunnamed (256 × 220)
    MRI-ESM2-0Japan (MRI)MRI.COM4.4 (360 × 363)
    NorESM2-LMNorway (NCC)CICE (360 × 384)
    NorESM2-MM
    TaiESM1China (AS-RCEC)CICE4 (320 × 384)
    UKESM1-0-LLUK (MOHC)CICE-HadGEM3-GSI8 (360 × 330)
    DownLoad: CSV
  • Arzel, O., T. Fichefet, and H. Goosse, 2006: Sea ice evolution over the 20th and 21st centuries as simulated by current AOGCMs. Ocean Modelling, 12(3-4), 401−415, https://doi.org/10.1016/j.ocemod.2005.08.002.
    Bitz, C. M., and G. H. Roe, 2004: A mechanism for the high rate of sea ice thinning in the Arctic Ocean. J. Climate, 17(18), 3623−3632, https://doi.org/10.1175/1520-0442(2004)017<3623:AMFTHR>2.0.CO;2.
    Bracegirdle, T. J., 2022: Early-to-late winter 20th century North Atlantic multidecadal atmospheric variability in observations, CMIP5 and CMIP6. Geophys. Res. Lett., 49(11), e2022GL098212, https://doi.org/10.1029/2022GL098212.
    Bryan, K., 1969: A numerical method for the study of the circulation of the world ocean. J. Comput. Phys., 4(3), 347−376, https://doi.org/10.1016/0021-9991(69)90004-7.
    Day, J. J., J. C. Hargreaves, J. D. Annan, and A. Abe-Ouchi, 2012: Sources of multi-decadal variability in Arctic sea ice extent. Environmental Research Letters, 7(3), 034011, https://doi.org/10.1088/1748-9326/7/3/034011.
    Divoky, G. J., P. M. Lukacs, and M. L. Druckenmiller, 2015: Effects of recent decreases in arctic sea ice on an ice-associated marine bird. Progress in Oceanography, 136, 151−161, https://doi.org/10.1016/j.pocean.2015.05.010.
    Dörr, J., M. Årthun, T. Eldevik, and E. Madonna, 2021: Mechanisms of regional winter sea-ice variability in a warming Arctic. J. Climate, 34(21), 8635−8653, https://doi.org/10.1175/JCLI-D-21-0149.1.
    Goosse, H., and V. Zunz, 2014: Decadal trends in the Antarctic sea ice extent ultimately controlled by ice-ocean feedback. The Cryosphere, 8(2), 453−470, https://doi.org/10.5194/tc-8-453-2014.
    Hezel, P. J., X. Zhang, C. M. Bitz, B. P. Kelly, and F. Massonnet, 2012: Projected decline in spring snow depth on Arctic sea ice caused by progressively later autumn open ocean freeze-up this century. Geophys. Res. Lett., 39(17), L17505, https://doi.org/10.1029/2012GL052794.
    Holland, M. M., C. M. Bitz, and B. Tremblay, 2006: Future abrupt reductions in the summer Arctic sea ice. Geophys. Res. Lett., 33(23), L23503, https://doi.org/10.1029/2006GL028024.
    IPCC, 2013: Climate Change 2013: The Physical Science Basis. Cambridge University Press.
    Kashiwase, H., K. I. Ohshima, S. Nihashi, and H. Eicken, 2017: Evidence for ice-ocean albedo feedback in the Arctic Ocean shifting to a seasonal ice zone. Scientific Reports, 7(1), 8170, https://doi.org/10.1038/s41598-017-08467-z.
    Kraus, E. B., and J. S. Turner, 1967: A one-dimensional model of the seasonal thermocline II. The general theory and its consequences. Tellus, 19, 98−106.
    Kumar, A., J. Yadav, and R. Mohan, 2021: Spatio-temporal change and variability of Barents-Kara sea ice, in the Arctic: Ocean and atmospheric implications. Science of the Total Environment, 753, 142046, https://doi.org/10.1016/j.scitotenv.2020.142046.
    Kwok, R., 2011: Observational assessment of Arctic Ocean sea ice motion, export, and thickness in CMIP3 climate simulations. J. Geophys. Res.: Oceans, 116, C00D05, https://doi.org/10.1029/2011JC007004.
    Kwok, R., and G. F. Cunningham, 2015: Variability of Arctic sea ice thickness and volume from CryoSat-2. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 373(2045), 20140157, https://doi.org/10.1098/rsta.2014.0157.
    Kwok, R., G. F. Cunningham, M. Wensnahan, I. Rigor, H. J. Zwally, and D. Yi, 2009: Thinning and volume loss of the Arctic Ocean sea ice cover: 2003−2008. J. Geophys. Res.: Oceans, 114(C7), C07005, https://doi.org/10.1029/2009JC005312.
    Laxon, S. W., and Coauthors, 2013: CryoSat-2 estimates of Arctic sea ice thickness and volume. Geophys. Res. Lett., 40(4), 732−737, https://doi.org/10.1002/grl.50193.
    Lindsay, R. W., and J. Zhang, 2006: Assimilation of ice concentration in an ice–ocean model. J. Atmos. Oceanic Technol., 23(5), 742−749, https://doi.org/10.1175/JTECH1871.1.
    Long, M. Y., L. J. Zhang, S. Y. Hu, and S. M. Qian, 2021: Multi-aspect assessment of CMIP6 models for Arctic sea ice simulation. J. Climate, 34(4), 1515−1529, https://doi.org/10.1175/JCLI-D-20-0522.1.
    Mallett, R. D. C., J. C. Stroeve, M. Tsamados, J. C. Landy, R. Willatt, V. Nandan, and G. E. Liston, 2021: Faster decline and higher variability in the sea ice thickness of the marginal Arctic seas when accounting for dynamic snow cover. The Cryosphere, 15(5), 2429−2450, https://doi.org/10.5194/tc-15-2429-2021.
    Massonnet, F., M. Vancoppenolle, H. Goosse, D. Docquier, T. Fichefet, and E. Blanchard-Wrigglesworth, 2018: Arctic sea-ice change tied to its mean state through thermodynamic processes. Nature Climate Change, 8(7), 599−603, https://doi.org/10.1038/s41558-018-0204-z.
    Meier, W. N., and Coauthors, 2014: Arctic sea ice in transformation: A review of recent observed changes and impacts on biology and human activity. Rev. Geophys., 52(3), 185−217, https://doi.org/10.1002/2013RG000431.
    Melia, N., K. Haines, and E. Hawkins, 2016: Sea ice decline and 21st century trans-Arctic shipping routes. Geophys. Res. Lett., 43(18), 9720−9728, https://doi.org/10.1002/2016GL069315.
    Min, C., Q. H. Yang, D. K. Chen, Y. J. Yang, X. Y. Zhou, Q. Shu, and J. P. Liu, 2022a: The emerging Arctic shipping corridors. Geophys. Res. Lett., 49, e2022GL099157, https://doi.org/10.1029/2022GL099157.
    Min, C., X. Y. Zhou, H. Luo, Y. J. Yang, Y. G. Wang, J. L. Zhang, and Q. H. Yang., 2022b: Toward quantifying the increasing accessibility of the Arctic Northeast Passage in the past four decades. Adv. Atmos. Sci.,
    Min, C., L. J. Mu, Q. H. Yang, R. Ricker, Q. Shi, B. Han, R. H. Wu, and J. P. Liu, 2019: Sea ice export through the Fram Strait derived from a combined model and satellite data set. The Cryosphere, 13(12), 3209−3224, https://doi.org/10.5194/tc-13-3209-2019.
    Notz, D., and SIMIP Community, 2020: Arctic sea ice in CMIP6. Geophys. Res. Lett., 47(10), e2019GL086749, https://doi.org/10.1029/2019GL086749.
    Notz, D., A. Jahn, M. Holland, E. Hunke, F. Massonnet, J. Stroeve, B. Tremblay, and M. Vancoppenolle, 2016: The CMIP6 Sea-Ice Model Intercomparison Project (SIMIP): Understanding sea ice through climate-model simulations. Geoscientific Model Development, 9(9), 3427−3446, https://doi.org/10.5194/gmd-9-3427-2016.
    Park, H., E. Watanabe, Y. Kim, I. Polyakov, K. Oshima, X. D. Zhang, J. S. Kimball, and D. Q. Yang, 2020: Increasing riverine heat influx triggers Arctic sea ice decline and oceanic and atmospheric warming. Science Advances, 6(45), eabc4699, https://doi.org/10.1126/sciadv.abc4699.
    Parkinson, C. L., and D. J. Cavalieri, 2008: Arctic sea ice variability and trends, 1979−2006. J. Geophys. Res.: Oceans, 113(C7), C07003, https://doi.org/10.1029/2007JC004564.
    Petrick, S., K. Riemann-Campe, S. Hoog, C. Growitsch, H. Schwind, R. Gerdes, and K. Rehdanz, 2017: Climate change, future Arctic Sea ice, and the competitiveness of European Arctic offshore oil and gas production on world markets. Ambio, 46(S3), 410−422, https://doi.org/10.1007/s13280-017-0957-z.
    Post, E., and Coauthors, 2013: Ecological consequences of sea-ice decline. Science, 341(6145), 519−524, https://doi.org/10.1126/science.1235225.
    Ricker, R., F. Girard-Ardhuin, T. Krumpen, and C. Lique, 2018: Satellite-derived sea ice export and its impact on Arctic ice mass balance. The Cryosphere, 12(9), 3017−3032, https://doi.org/10.5194/tc-12-3017-2018.
    Ricker, R., S. Hendricks, D. K. Perovich, V. Helm, and R. Gerdes, 2015: Impact of snow accumulation on CryoSat-2 range retrievals over Arctic sea ice: An observational approach with buoy data. Geophys. Res. Lett., 42(11), 4447−4455, https://doi.org/10.1002/2015GL064081.
    Roach, L. A., and Coauthors, 2020: Antarctic sea ice area in CMIP6. Geophys. Res. Lett., 47(9), e2019GL086729, https://doi.org/10.1029/2019GL086729.
    Schweiger, A., R. Lindsay, J. L. Zhang, M. Steele, H. Stern, and R. Kwok, 2011: Uncertainty in modeled Arctic sea ice volume. J. Geophys. Res.: Oceans, 116, C00D06, https://doi.org/10.1029/2011JC007084.
    Screen, J. A., and I. Simmonds, 2010: The central role of diminishing sea ice in recent Arctic temperature amplification. Nature, 464(7293), 1334−1337, https://doi.org/10.1038/nature09051.
    Semenov, V. A., T. Martin, L. K. Behrens, and M. Latif, 2015: Arctic sea ice area in CMIP3 and CMIP5 climate model ensembles–variability and change. The Cryosphere Discussions, 9(1), 1077−1131, https://doi.org/10.5194/tcd-9-1077-2015.
    Serreze, M. C., and W. N. Meier, 2019: The Arctic’s sea ice cover: Trends, variability, predictability, and comparisons to the Antarctic. Annals of the New York Academy of Sciences, 1436(1), 36−53, https://doi.org/10.1111/nyas.13856.
    Sévellec, F., A. V. Fedorov, and W. Liu, 2017: Arctic sea-ice decline weakens the Atlantic Meridional Overturning Circulation. Nature Climate Change, 7(8), 604−610, https://doi.org/10.1038/nclimate3353.
    Shen, Z. L., A. M. Duan, D. L. Li, and J. X. Li, 2021: Assessment and ranking of climate models in Arctic sea ice cover simulation: From CMIP5 to CMIP6. J. Climate, 34(9), 3609−3627, https://doi.org/10.1175/JCLI-D-20-0294.1.
    Shu, Q., Q. Wang, J. Su, X. Li, and F. L. Qiao, 2019: Assessment of the Atlantic water layer in the Arctic Ocean in CMIP5 climate models. Climate Dyn., 53(9−10), 5279−5291, https://doi.org/10.1007/s00382-019-04870-6.
    Shu, Q., Q. Wang, Z. Y. Song, F. L. Qiao, J. C. Zhao, M. Chu, and X. F. Li, 2020: Assessment of sea ice extent in CMIP6 with comparison to observations and CMIP5. Geophys. Res. Lett., 47(9), e2020GL087965, https://doi.org/10.1029/2020GL087965.
    Smith, L. C., and S. R. Stephenson, 2013: New Trans-Arctic shipping routes navigable by midcentury. Proceedings of the National Academy of Sciences of the United States of America, 110(13), E1191−E1195, https://doi.org/10.1073/pnas.1214212110.
    Spall, M. A., 2019: Dynamics and thermodynamics of the mean transpolar drift and ice thickness in the Arctic Ocean. J. Climate, 32(24), 8449−8463, https://doi.org/10.1175/JCLI-D-19-0252.1.
    Stroeve, J., A. Barrett, M. Serreze, and A. Schweiger, 2014: Using records from submarine, aircraft and satellites to evaluate climate model simulations of Arctic sea ice thickness. The Cryosphere, 8(5), 1839−1854, https://doi.org/10.5194/tc-8-1839-2014.
    Stroeve, J. C., M. C. Serreze, M. M. Holland, J. E. Kay, J. Malanik, and A. P. Barrett, 2012a: The Arctic’s rapidly shrinking sea ice cover: A research synthesis. Climatic Change, 110(3−4), 1005−1027, https://doi.org/10.1007/s10584-011-0101-1.
    Stroeve, J. C., V. Kattsov, A. Barrett, M. Serreze, T. Pavlova, M. Holland, and W. N. Meier, 2012b: Trends in Arctic sea ice extent from CMIP5, CMIP3 and observations. Geophys. Res. Lett., 39(16), L16502, https://doi.org/10.1029/2012GL052676.
    Taylor, K. E., 2001: Summarizing multiple aspects of model performance in a single diagram. J. Geophys. Res.: Atmos., 106(D7), 7183−7192, https://doi.org/10.1029/2000JD900719.
    Thackeray, C. W., and A. Hall, 2019: An emergent constraint on future Arctic sea-ice albedo feedback. Nature Climate Change, 9(12), 972−978, https://doi.org/10.1038/s41558-019-0619-1.
    Tian-Kunze, X., L. Kaleschke, N. Maaß, M. Mäkynen, N. Serra, M. Drusch, and T. Krumpen, 2014: SMOS-derived thin sea ice thickness: Algorithm baseline, product specifications and initial verification. The Cryosphere, 8(3), 997−1018, https://doi.org/10.5194/tc-8-997-2014.
    Turner, J., T. J. Bracegirdle, T. Phillips, G. J. Marshall, and J. S. Hosking, 2013: An initial assessment of Antarctic sea ice extent in the CMIP5 models. J. Climate, 26(5), 1473−1484, https://doi.org/10.1175/JCLI-D-12-00068.1.
    Watts, M., W. Maslowski, Y. J. Lee, J. C. Kinney, and R. Osinski, 2021: A spatial evaluation of Arctic sea ice and regional limitations in CMIP6 historical simulations. J. Climate, 34, 6399−6420, https://doi.org/10.1175/JCLI-D-20-0491.1.
    Yu, L. J., S. Y. Zhong, T. Vihma, and B. Sun, 2021: Attribution of late summer early autumn Arctic sea ice decline in recent decades. NPJ Climate and Atmospheric Science, 4(1), 3, https://doi.org/10.1038/S41612-020-00157-4.
    Zhang, J., S. T. Stegall, and X. D. Zhang, 2018: Wind-sea surface temperature-sea ice relationship in the Chukchi-Beaufort Seas during autumn. Environmental Research Letters, 13(3), 034008, https://doi.org/10.1088/1748-9326/aa9adb.
    Zhang, J., W. D. Hibler, M. Steele, and D. A. Rothrock, 1998: Arctic ice-ocean modeling with and without climate restoring. J. Phys. Oceanogr., 28(2), 191−217, https://doi.org/10.1175/1520-0485(1998)028<0191:AIOMWA>2.0.CO;2.
    Zhang, J. L., and D. A. Rothrock, 2003: Modeling global sea ice with a thickness and enthalpy distribution model in generalized curvilinear coordinates. Mon. Wea. Rev., 131(5), 845−861, https://doi.org/10.1175/1520-0493(2003)131<0845:MGSIWA>2.0.CO;2.
    Zhang, R., 2015: Mechanisms for low-frequency variability of summer Arctic sea ice extent. Proceedings of the National Academy of Sciences of the United States of America, 112(15), 4570−4575, https://doi.org/10.1073/pnas.1422296112.
    Zhou, T. J., L. W. Zou, and X. L. Chen, 2019: Commentary on the Coupled Model Intercomparison Project Phase 6 (CMIP6). Climate Change Research, 15(5), 445−456, https://doi.org/10.12006/j.issn.1673-1719.2019.193. (in Chinese with English abstract
    Zhu, H. H., Z. H. Jiang, J. Li, W. Li, C. X. Sun, and L. Li, 2020: Does CMIP6 inspire more confidence in simulating climate extremes over China? Adv. Atmos. Sci., 37(10), 1119−1132, https://doi.org/10.1007/s00376-020-9289-1.
  • [1] Xinyao RONG, Jian LI, Haoming CHEN, Jingzhi SU, Lijuan HUA, Zhengqiu ZHANG, Yufei XIN, 2021: The CMIP6 Historical Simulation Datasets Produced by the Climate System Model CAMS-CSM, ADVANCES IN ATMOSPHERIC SCIENCES, 38, 285-295.  doi: 10.1007/s00376-020-0171-y
    [2] Yuyang GUO, Yongqiang YU, Pengfei LIN, Hailong LIU, Bian HE, Qing BAO, Shuwen ZHAO, Xiaowei WANG, 2020: Overview of the CMIP6 Historical Experiment Datasets with the Climate System Model CAS FGOALS-f3-L, ADVANCES IN ATMOSPHERIC SCIENCES, 37, 1057-1066.  doi: 10.1007/s00376-020-2004-4
    [3] Yinghan SANG, Hong-Li REN, Xueli SHI, Xiaofeng XU, Haishan CHEN, 2021: Improvement of Soil Moisture Simulation in Eurasia by the Beijing Climate Center Climate System Model from CMIP5 to CMIP6, ADVANCES IN ATMOSPHERIC SCIENCES, 38, 237-252.  doi: 10.1007/s00376-020-0167-7
    [4] Shengzhe CHEN, Jiping LIU, Yifan DING, Yuanyuan ZHANG, Xiao CHENG, Yongyun HU, 2021: Assessment of Snow Depth over Arctic Sea Ice in CMIP6 Models Using Satellite Data, ADVANCES IN ATMOSPHERIC SCIENCES, 38, 168-186.  doi: 10.1007/s00376-020-0213-5
    [5] Xiaoyong YU, Chengyan LIU, Xiaocun WANG, Jian CAO, Jihai DONG, Yu LIU, 2022: Evaluation of Arctic Sea Ice Drift and its Relationship with Near-surface Wind and Ocean Current in Nine CMIP6 Models from China, ADVANCES IN ATMOSPHERIC SCIENCES, 39, 903-926.  doi: 10.1007/s00376-021-1153-4
    [6] Yuyang GUO, Yongqiang YU, Pengfei LIN, Hailong LIU, Bian HE, Qing BAO, Bo AN, Shuwen ZHAO, Lijuan HUA, 2020: Simulation and Improvements of Oceanic Circulation and Sea Ice by the Coupled Climate System Model FGOALS-f3-L, ADVANCES IN ATMOSPHERIC SCIENCES, 37, 1133-1148.  doi: 10.1007/s00376-020-0006-x
    [7] Yu Rucong, Jin Xiangze, Zhang Xuehong, 1995: Design and Numerical Simulation of an Arctic Ocean Circulation and Thermodynamic Sea-Ice Model, ADVANCES IN ATMOSPHERIC SCIENCES, 12, 289-310.  doi: 10.1007/BF02656978
    [8] Yuanxin LIU, Lijing CHENG, Yuying PAN, Zhetao TAN, John ABRAHAM, Bin ZHANG, Jiang ZHU, Junqiang SONG, 2022: How Well Do CMIP6 and CMIP5 Models Simulate the Climatological Seasonal Variations in Ocean Salinity?, ADVANCES IN ATMOSPHERIC SCIENCES, 39, 1650-1672.  doi: 10.1007/s00376-022-1381-2
    [9] Bian HE, Qing BAO*, Xiaocong WANG, Linjiong ZHOU, Xiaofei WU, Yimin LIU, Guoxiong WU, Kangjun CHEN, Sicheng HE, Wenting HU, Jiandong LI, Jinxiao LI, Guokui NIAN, Lei WANG, Jing YANG, Minghua ZHANG, Xiaoqi ZHANG, 2019: CAS FGOALS-f3-L Model Datasets for CMIP6 Historical Atmospheric Model Intercomparison Project Simulation, ADVANCES IN ATMOSPHERIC SCIENCES, , 771-778.  doi: 10.1007/s00376-019-9027-8
    [10] Liang ZHAO, Yunwen LIU, Yihui DING, Qingquan LI, Wei DONG, Xinyong SHEN, Wei CHENG, Haoxin YAO, Ziniu XIAO, 2023: The Warm Arctic–Cold Eurasia Pattern and Its Key Region in Winter in CMIP6 Model Simulations, ADVANCES IN ATMOSPHERIC SCIENCES, 40, 2138-2153.  doi: 10.1007/s00376-022-2201-4
    [11] Yin ZHAO, Tianjun ZHOU, Wenxia ZHANG, Jian LI, 2022: Change in Precipitation over the Tibetan Plateau Projected by Weighted CMIP6 Models, ADVANCES IN ATMOSPHERIC SCIENCES, 39, 1133-1150.  doi: 10.1007/s00376-022-1401-2
    [12] Huanhuan ZHU, Zhihong JIANG, Juan LI, Wei LI, Cenxiao SUN, Laurent LI, 2020: Does CMIP6 Inspire More Confidence in Simulating Climate Extremes over China?, ADVANCES IN ATMOSPHERIC SCIENCES, 37, 1119-1132.  doi: 10.1007/s00376-020-9289-1
    [13] Bruno FERRERO, Marcos TONELLI, Fernanda MARCELLO, Ilana WAINER, 2021: Long-term Regional Dynamic Sea Level Changes from CMIP6 Projections, ADVANCES IN ATMOSPHERIC SCIENCES, 38, 157-167.  doi: 10.1007/s00376-020-0178-4
    [14] Changyu LI, Jianping HUANG, Lei DING, Yu REN, Linli AN, Xiaoyue LIU, Jiping HUANG, 2022: The Variability of Air-sea O2 Flux in CMIP6: Implications for Estimating Terrestrial and Oceanic Carbon Sinks, ADVANCES IN ATMOSPHERIC SCIENCES, 39, 1271-1284.  doi: 10.1007/s00376-021-1273-x
    [15] Rongrong PAN, Qi SHU, Zhenya SONG, Shizhu WANG, Yan HE, Fangli QIAO, 2023: Simulations and Projections of Winter Sea Ice in the Barents Sea by CMIP6 Climate Models, ADVANCES IN ATMOSPHERIC SCIENCES, 40, 2318-2330.  doi: 10.1007/s00376-023-2235-2
    [16] Odd Helge OTTER, Helge DRANGE, 2004: A Possible Feedback Mechanism Involving the Arctic Freshwater,the Arctic Sea Ice, and the North Atlantic Drift, ADVANCES IN ATMOSPHERIC SCIENCES, 21, 784-801.  doi: 10.1007/BF02916375
    [17] Jinping WANG, Xianyao CHEN, 2023: Arctic Sea Level Variability from Oceanic Reanalysis and Observations, ADVANCES IN ATMOSPHERIC SCIENCES, 40, 2362-2377.  doi: 10.1007/s00376-023-3004-y
    [18] ZHANG Lixia* and ZHOU Tianjun, , 2014: An Assessment of Improvements in Global Monsoon Precipitation Simulation in FGOALS-s2, ADVANCES IN ATMOSPHERIC SCIENCES, 31, 165-178.  doi: 10.1007/s00376-013-2164-6
    [19] Pengfei LIN, Zhipeng YU, Hailong LIU, Yongqiang YU, Yiwen LI, Jirong JIANG, Wei XUE, Kangjun CHEN, Qian YANG, Bowen ZHAO, Jilin WEI, Mengrong DING, Zhikuo SUN, Yaqi WANG, Yao MENG, Weipeng ZHENG, Jinfeng MA, 2020: LICOM Model Datasets for the CMIP6 Ocean Model Intercomparison Project, ADVANCES IN ATMOSPHERIC SCIENCES, 37, 239-249.  doi: 10.1007/s00376-019-9208-5
    [20] Jie ZHANG, Tongwen WU, Fang ZHANG, Kalli FURTADO, Xiaoge XIN, Xueli SHI, Jianglong LI, Min CHU, Li ZHANG, Qianxia LIU, Jinghui Yan, Min WEI, Qiang MA, 2021: BCC-ESM1 Model Datasets for the CMIP6 Aerosol Chemistry Model Intercomparison Project (AerChemMIP), ADVANCES IN ATMOSPHERIC SCIENCES, 38, 317-328.  doi: 10.1007/s00376-020-0151-2

Get Citation+

Export:  

Share Article

Manuscript History

Manuscript received: 18 December 2021
Manuscript revised: 10 September 2022
Manuscript accepted: 14 September 2022
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

The Arctic Sea Ice Thickness Change in CMIP6’s Historical Simulations

    Corresponding author: Bo HAN, hanbo5@mail.sysu.edu.cn
  • 1. School of Atmospheric Sciences, Sun Yat-sen University, and Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), Zhuhai 519082, China
  • 2. First Institute of Oceanography, Ministry of Natural Resources, Qingdao 266061, China

Abstract: This study assesses sea ice thickness (SIT) from the historical run of the Coupled Model Inter-comparison Project Phase 6 (CMIP6). The SIT reanalysis from the Pan-Arctic Ice Ocean Modeling and Assimilation System (PIOMAS) product is chosen as the validation reference data. Results show that most models can adequately reproduce the climatological mean, seasonal cycle, and long-term trend of Arctic Ocean SIT during 1979–2014, but significant inter-model spread exists. Differences in simulated SIT patterns among the CMIP6 models may be related to model resolution and sea ice model components. By comparing the climatological mean and trend for SIT among all models, the Arctic SIT change in different seas during 1979–2014 is evaluated. Under the scenario of historical radiative forcing, the Arctic SIT will probably exponentially decay at –18% (10 yr)–1 and plausibly reach its minimum (equilibrium) of 0.47 m since the 2070s.

摘要: 本研究将第六次国际耦合模式比较计划(CMIP6)历史试验结果与华盛顿大学泛北冰洋模式与同化系统(PIOMAS)的冰厚再分析数据进行了系统性的比较,评估了各耦合模式对1979-2014年期间北极海冰厚度变化的模拟能力。研究结果表明,大多数模式都能较好地重现再分析的北极海冰厚度气候态、年变化、年际变化和线性趋势,但依然存在明显的模式间差异。这种差异与模式分辨率以及各耦合模式对海冰模式的选择有关。随后,通过比较各模式间北极海冰厚度的气候态平均值和年际变化趋势,对1979-2014年间北极不同海域海冰厚度变化进行了进一步的分析。结果表明,在历史辐射强迫情景下,北极海冰年平均厚度可能以每十年-18%的速度指数衰减,并有可能在20世纪70年代达到0.47 m的平衡值。文章也对北极不同海域的海冰变化特征进行了讨论。

    • Sea ice is a significant component of the Arctic environment (Parkinson and Cavalieri, 2008). It can modulate the global atmosphere and ocean circulation mainly by changing heat, momentum, and moisture exchanges between the atmosphere and ocean (Sévellec et al., 2017; Mallett et al., 2021). Arctic sea ice is one of the crucial indicators of global climate change (Notz et al., 2016; Serreze and Meier, 2019). With ongoing global warming, sea ice has retreated dramatically in the Arctic Ocean (Stroeve et al., 2012a, b; Kwok and Cunningham, 2015), which has a significant impact on local ecosystem evolution (Post et al., 2013; Divoky et al., 2015), navigation (Smith and Stephenson, 2013; Melia et al., 2016; Min et al., 2022a, b), and resource extraction (Meier et al., 2014; Petrick et al., 2017).

      Sea ice thickness (SIT) is one of the most important physical properties of the cryosphere as it directly modulates the energy exchange efficiency between the adjacent ocean and atmosphere. Among the existing observations, field observations, including sea ice draft measurements from submarine upward-looking sonar (ULS), moored ULS, Ice Mass Balance (IMB) buoys, and airborne electromagnetic measurements, are a common way to derive Arctic SIT (Schweiger et al., 2011). Additionally, draft observations from submarines and moorings are converted to ice thicknesses using an ice density of 928 kg m–3, and the snow water equivalent estimated by the model and uncertainty in the draft-to-thickness conversion is relatively small (<10%) (Schweiger et al., 2011). The maximum uncertainties of the laser altimetry data from satellites, like ICESat and ICESat-2, were estimated to be 0.5 m for individual 25-km ICESat grid cells (Kwok et al., 2009). And the sea ice freeboard detected by the satellite altimeter mission CryoSat-2 can be converted to sea ice thickness by considering the floe’s hydrostatic equilibrium, given the sea ice density and weight of overlying snow (Ricker et al., 2015 Mallett et al., 2021). Moreover, the L-band radiometer on the Soil Moisture and Ocean Salinity (SMOS) mission can also be used to obtain sea ice thickness (Tian-Kunze et al., 2014). The uncertainties of SMOS-derived ice thickness are small over thin ice (< 1 m), while the sea ice estimates retrieved from CryoSat-2 have high accuracy over thick ice (> 1 m) (Ricker et al., 2018).

      However, observations of Arctic SIT, especially long-term and wide-coverage observations, are still scarce, which yields a significant uncertainty in revisiting the sea ice volume (SIV) and SIT changes in the Arctic Ocean. Therefore, SIT reanalysis data have tremendous scientific significance. Among various sea ice reanalysis products, the Pan-Arctic Ice Ocean Modeling and Assimilation System (PIOMAS) (Zhang and Rothrock, 2003), developed by the University of Washington, covers the period from 1978 to the present. Several studies have already verified the validity and quantitatively discussed the uncertainty of PIOMAS data (Schweiger et al., 2011; Laxon et al., 2013; Stroeve et al., 2014). Compared with satellite data and in situ observations, PIOMAS overestimates SIT in thin ice areas and underestimates SIT in thick ice areas (Stroeve et al., 2014). Nevertheless, PIOMAS data helps express the spatial distribution and trend of Arctic SIT (Schweiger et al., 2011) and has advantages in climate assessment compared with remote sensing products like ICESat and Cryosat-2 (Laxon et al., 2013).

      In addition, a coupled model is a powerful tool for studying the mechanism of climate change within the Arctic region (Arzel et al., 2006; Stroeve et al., 2012b; Semenov et al., 2015). Under the auspices of the Working Group on Coupled Modelling (WGCM) of the World Climate Research Program (WCRP), Coupled Modeling Inter-comparison Project (CMIP) has carried out a series of fruitful works on formulating formats and mechanisms for sharing simulation data with the global scientific community (Zhou et al., 2019). And the role of sea ice in climate change is one of the key topics in CMIP6, especially in the Sea-Ice Model Inter-comparison Project (SIMIP) (Notz et al., 2016). As a representation of natural climate, the historical run is a fundamental experiment in all CMIP projects, and it can represent the ability of a coupled model to simulate the real earth system.

      Sea ice simulations, like those from CMIP, are often used in the discussions of climate change (Turner et al., 2013; Notz and SIMIP Community, 2020; Roach et al., 2020; Shu et al., 2020; Shen et al., 2021). However, the representation of model results for the earth system remains verified. In recent years, with the data from PIOMAS, several papers have evaluated the SIT data from CMIP, especially on its spatial distribution (Stroeve et al., 2014; Watts et al., 2021). In this study, we systematically evaluate the Arctic SIT from the historical run of CMIP6 against PIOMAS SIT data for the climatological mean, seasonal cycle, and long-term trend of Arctic Ocean SIT during 1979–2014.

      This paper is organized as follows. The datasets used in this study are described in section 2. In section 3, we discuss the performance of CMIP6 models in simulating the Arctic SIT for the mean state (section 3.1), seasonal cycle (section 3.2), and long-term trend (section 3.3), and discuss the Arctic SIT diversities from the models’ ensemble (section 3.4). The last section (section 4) concludes this study and provides a further discussion. This study has two main aims: (1) Assessing CMIP6’s SIT simulation ability in different regions of the Arctic Ocean; (2) Summarizing the general features of Arctic sea ice response to historical radiative forcing.

    2.   Data
    • This study uses SIT data from the CMIP6 historical run and PIOMAS. The SIT reanalysis has been provided by PIOMAS since 1979, whereas the historical run spans from 1850 to 2014. Therefore, the study focuses on the period from 1979 to 2014. For convenience, all data have been gridded into 1º × 1º before comparison and analysis. The analysis domain is shown in Fig. 1.

      Figure 1.  Spatial pattern of SIT from PIOMAS during 1979–2014. The uppercase letters represent the different seas defined by the NSIDC Arctic Ocean regional masks (https://nsidc.org/data/polar-stereo/tools_masks.html#region_masks). The black line is the boundary among the different seas. (A—Central Arctic; B—Beaufort Sea; C—Chukchi Sea; D—East Siberian Sea; E—Laptev Sea; F—Kara Sea; and G—Barents Sea).

    • PIOMAS is developed by the University of Washington (Zhang and Rothrock, 2003). The ocean model in PIOMAS is based on the Bryan-Cox-Semtner-type model (Bryan, 1969; Zhang and Rothrock, 2003), with an embedded mixed layer described by Kraus and Turner (Kraus and Turner, 1967), and the ice model is a multi-category ice thickness and enthalpy distribution model (Lindsay and Zhang, 2006). Detailed information about the ice and ocean model can be found in Lindsay and Zhang (2006) and Zhang et al. (1998). The near-real-time satellite sea ice concentration (SIC) observations from the National Snow and Ice Data Center (NSIDC) and sea surface temperature (SST) reanalysis in ice-free areas from the NCEP/NCAR are assimilated into the ice–ocean model. From 1978 to the present, PIOMAS provides a multitude of essential information about the ocean and ice. In this study, the PIOMAS SIT is used as a reference to evaluate the SIT simulated by CMIP6 models.

    • Since the start of CMIP, CMIP6 has the most models participating and the most consistent experiment design (Zhou et al., 2019). The SIT simulations from 34 CMIP6 models are evaluated in this assessment, and the basic information of these models is listed in Table 1. However, it should be noted that there are two types of SIT (i.e., sivol and sithick) provided by CMIP6 (Notz et al., 2016):

      ModelCountry (Institution)Sea Ice Model
      (resolution: longitude × latitude)
      SITANESM3China (NUIST)CICE4.1 (384 × 362)
      MIROC-ES2LJapan (MIROC)COCO4.9 (360 × 256)
      MIROC6
      EC-EARTH3Europe (EC)NEMO-LIM3 (362 × 292)
      EC-EARTH3-Veg
      EC-EARTH3-Veg-LR
      EC-EARTH3-AerChem
      E3SMLLNL (E3SM)MPAS-Seaice v6.0 (360 × 180)
      SITEACCESS-CM2Australia (CSIRO-ARCCSS)CICE5.1.2 (360 × 300)
      ACCESS-ESM1-5Australia (CSIRO)CICE4.1 (360 × 300)
      BCC-CSM2-MRChina (BCC)SIS2 (360 × 232)
      CAMS-CSM1-0China (CAMS)SIS1.0 (320 × 160)
      CESM2USA (NCAR)CICE5.1 (320 × 384)
      CESM2-WACCM
      CIESMChina (THU)CICE4 (320 × 384)
      CMCC-CM2-SR5Italy (CMCC)CICE4.0 (288 × 192)
      CNRM-CM6-1France (CNRM-CERFACS)Gelato6.1 (362 × 294)
      CNRM-CM6-1-HRGelato6.1 (1442 × 1050)
      CNRM-ESM2-1Gelato6.1 (362 × 294)
      EC-EARTH3Europe (EC)NEMO-LIM3 (362 × 292)
      EC-EARTH3-AerChem
      EC-EARTH3-Veg
      EC-EARTH3-Veg-LR
      FGOALS-f3-LChina (CAS)CICE4.0 (360 × 218)
      FIO-ESM-2-0China (FIO-QLNM)CICE4.0 (320 × 384)
      GFDL-ESM4USA (NOAA-GFDL)GFDL-SIM4p5 (720 × 576)
      GISS-E2-1-GUSA (NASA-GISS)GISS SI (144 × 90)
      HadGEM3-GC31-LLUK (MOHC)CICE-HadGEM3-GSI8 (360 × 330)
      HadGEM3-GC31-MMCICE-HadGEM3-GSI8 (1440 × 1205)
      IPSL-CM6A-LRFrance (IPSL)NEMO-LIM3 (362 × 332)
      MPI-ESM-1-2-HAMEurope (HAMMOZ-Consortium)unnamed (256 × 220)
      MPI-ESM1-2-HRGermany (MPI-M)unnamed (384 × 192)
      MPI-ESM1-2-LRunnamed (256 × 220)
      MRI-ESM2-0Japan (MRI)MRI.COM4.4 (360 × 363)
      NorESM2-LMNorway (NCC)CICE (360 × 384)
      NorESM2-MM
      TaiESM1China (AS-RCEC)CICE4 (320 × 384)
      UKESM1-0-LLUK (MOHC)CICE-HadGEM3-GSI8 (360 × 330)

      Table 1.  The basic information of CMIP6 models.

      (1) Equivalent sea ice thickness (SITE) is grid cell-averaged ice thickness. It used to be referred to as simply “ice thickness” in earlier CMIPs (such as CMIP5), which caused some misunderstanding among users who thought this variable would refer to actual thickness. In CMIP6, SITE is represented by sea-ice volume per area (sivol).

      (2) The thickness (sithick in CMIP6) of sea ice averaged over an ice-covered portion of a grid cell is known as actual sea ice thickness (SITA). It corresponds well to real (or floe) sea ice thickness.

      Among the 34 models, eight CMIP6 models provide both SITE and SITA and 26 models only produce SITE. Because PIOMAS produces SITE, all SITA is converted into SITE for further analysis. The conversion between these two SITs is:

      SIC is the sea ice concentration in a grid. SIT stands for SITE (hereafter) without any further definition in this investigation.

    3.   Results
    • Figure 1 gives the spatial pattern of climatological mean Arctic SIT from PIOMAS. Compared with PIOMAS, the results from 34 models show significant differences (Fig. 2). Some models (e.g., CESM2, CIESM, CMCC-CM2-SR5, and three CNRMs) underestimate SIT across the Arctic Ocean region. In contrast, some other models overestimate SIT (e.g., E3SM, EC-Earth3s, and FGOALS-f3-L), especially in the thin-ice areas in the marginal seas. As PIOMAS overestimates the SIT in the thin-ice area and underestimates it in the thick-ice area, the deviation between the CMIP6 multi-model ensemble mean (CMIP6s MMM) and the actual condition may be greater than what is shown in Fig. 2. In addition, several models (e.g., EC-Earth3s and FGOALS-f3-L) show a thick ice pack extending from the northern Greenland area along the Lomonosov Ridge to the East Siberian Continental Shelf, significantly overestimating the ice thickness in the Siberian Sea and the east part of the Central Arctic. The underestimation of SIT is most significant in the thick-ice areas near the Canadian Arctic Archipelago (CAA) and the northern coast of Greenland.

      Figure 2.  The difference of climatological mean SIT between CMIP6 models and PIOMAS during 1979–2014.

      Most models capture the spatial pattern of climatological mean SIT well, with spatial correlation coefficients ranging from 0.77 to 0.95 (Fig. 3). The normalized standard deviation ranges from 0.2 to 1.9. Interestingly, the models that most significantly underestimate the mean SIT have the smallest normalized standard deviation (e.g., BCC-CSM2-R, CIESM, CMCC, CNRMs). In contrast, the model with the greatest normalized standard deviation most significantly overestimates the mean SIT (UKESM1-0-LL).

      Figure 3.  Taylor Diagram of Arctic SIT for CMIP6 models and PIOMAS during 1979–2014.

      As shown in Fig. 4, over the Arctic Ocean, the regional averaged SIT in CMIP6s MMM is about 0.19 m smaller than PIOMAS. CMIP6s MMM underestimates the climatological mean SIT in the Central Arctic, the Beaufort Sea, and the Kara Sea but overestimates it in the other seas. The most considerable overestimation and underestimation of SIT exist in the East Siberian Sea and the Central Arctic, with values of 0.2 m and –0.3 m, respectively. However, the most significant negative (~ –18%) relative deviation appears in the Kara Sea due to the small mean SIT. The relative deviation shown in the parentheses above is defined as:

      Figure 4.  The boxplot of regional averaged SIT from CMIP6s MMM and PIOMAS in different seas during 1979–2014. The five values in a box from top to bottom are the maximum, 75th percentile, median, 25th percentile, and minimum. The relative deviation of the averaged SIT between CMIP6s MMM and PIOMAS is shown as solid circles (right axis).

      Kwok (2011) previously found that the deficiency in simulating SIT fields in CMIP3 models could be ascribed to their inability to mimic the natural sea level pressure pattern. Stroeve et al. (2014) then figured out that thick ice in the Siberian Sea in CMIP5 models can be partly explained by the biases in the surface wind fields and the mean flow of the Beaufort Gyre and the Transpolar Drift Stream. This problem may still exist in CMIP6 models and lead to biases in the spatial pattern of simulated SIT, but the specific situation needs to be further analyzed.

      The performance of a climate model for SIT seems to be related to its spatial resolution and sea ice physics components (Massonnet et al., 2018; Long et al., 2021). Here, we use the Taylor Score (Taylor, 2001) to represent the simulation ability of CMIP6 models. The Taylor Score (TS) is calculated as follows:

      where R represents the spatial correlation coefficient between the simulations and PIOMAS, and R0 is the maximum of R. σsm and σso represent the standard deviation of the variables in the simulations and PIOMAS, respectively. TS ranges between 0 and 1; the higher the TS is, the closer the simulation is to PIOMAS (Zhu et al., 2020).

      According to Fig. 5, models with high resolution are slightly more capable of reproducing the spatial pattern of Arctic SIT, consistent with previous results (Long et al., 2021) on the relationship between resolution and SIC. However, the dependence of the simulation results on the resolution decreases when the resolution reaches a certain level. In addition, the TS values corresponding to some sea ice components (e.g., Gelato and SIS) are all less than 0.8, while other components (like COCO and NEMO-LIM) have much better TS values close to 0.9. Because CICE involves a more reasonable consideration of dynamic and thermodynamic processes of sea ice than SIS (Long et al., 2021), TS values from most models with CICE are > 0.8, except for CMCC-CM2-SR5 and CIESM. The SIT from these two models is the thinnest among all models (Fig. 2), and the overestimated open water area might significantly lower the TS. Therefore, enhancing model resolution and choosing a proper sea ice component can improve the simulation of the climatological SIT spatial pattern.

      Figure 5.  Taylor Score (TS) against (a) the ranking of model resolution and (b) the sea ice components in CMIP6 models during 1979–2014. The dashed lines indicate a TS of 0.8.

    • The climatological mean seasonal cycles of SIT during 1979–2014 are given in Fig. 6a. From PIOMAS, the Arctic SIT reaches its maximum in May and reaches its minimum in September. Most CMIP6 models reproduce this seasonal cycle well, except for E3SM, which gives its minimum in November. CIESM and CMCC-CM2-SR5 even give a nearly ice-free period from August to October, which is undoubtedly not in agreement with PIOMAS.

      Figure 6.  Seasonal cycles of (a) SIT and (b) TS from CMIP6 models and PIOMAS during 1979–2014.

      The spatial patterns of climatological SIT in each month are quite different between all models. As seen in Fig. 6b, 22 out of 34 models give a TS>0.6 each month, and of those 22 models, the TaiESM1 has the best performance. The TS from almost all the other models is greater than 0.2 each month, except for CMCC-CM2-SR5 and CIESM, which have a TS less than 0.2 from August to December. The TS of CMIP6s MMM is about 0.97 throughout the year, higher than any individual model.

      For most models, the maximum and minimum of TS appear in September and May, respectively, and TS increases from September to May and then decreases. Therefore, the spatial pattern of simulated SIT is better in the sea ice growing season than in the melting season. The different performances of CMIP6 models in the two periods may be due to the complexity of the feedback process.

      SIT change is stimulated by external forcing and modulated by internal feedback. The sea ice albedo feedback, which is due to the albedo contrast between water and ice, is a crucial driver of Arctic climate change and is also a significant factor in seasonal sea ice retreat (Screen and Simmonds, 2010; Kashiwase et al., 2017; Thackeray and Hall, 2019). In the polar region, shrinking snow and ice would reduce the surface albedo, intensify the surface solar absorption, and promote further sea ice reduction, which is a typical positive (or amplifying) feedback in the polar region. However, the sea ice loss also results in a negative (or damping) feedback. During the melting season, the enhanced heat loss of seawater within the newly formed open water area, mainly through thermal radiation, will accelerate the sea ice growth rate afterward (Bitz and Roe, 2004; Hezel et al., 2012). In addition, the complexity of feedback processes makes the correlation between wind speed and SST negative over open water but positive in sea-ice-dominated areas (Zhang et al., 2018). Moreover, the ice–ocean feedback related to salinity also complicates the SIT response to a warming climate (Goosse and Zunz, 2014). The difference in each model’s ability to reproduce these feedback processes should be responsible for the SIT’s spread in simulations.

    • The linear trend of the Arctic SIT during 1979–2014 is captured well by CMIP6s MMM, referring to PIOMAS (Fig. 7). PIOMAS gives a declining rate of about –0.26 m (10 yr)–1, with an intensified declining rate in September [–0.28 m (10 yr)–1] and a weakened one in May [–0.25 m (10 yr)–1]. Furthermore, the declining rate of annual mean SIT from CMIP6s MMM is about –0.23 m (10 yr)–1, similar to the result from PIOMAS data. For individual models, the declining rate is between –0.012 m (10 yr)–1 (CAMS-CSM1-0) and –0.60 m (10 yr)–1 (EC-Earth3) in May, and between –0.0001 m (10 yr)–1 (CIESM) and –0.59 m (10 yr)–1 (EC-Earth3) in September.

      Figure 7.  Evolution of (a) annual mean, (b) May, and (c) September Arctic SIT for CMIP6 models and PIOMAS during 1979–2014.

      The decline of SIT accelerated after 2000 in PIOMAS. The declining rate was about –0.14 m (10 yr)–1 in May and –0.15 m (10 yr)–1 in September before 2000. After that, it grew to –0.41 m (10 yr)–1 and –0.48 m (10 yr)–1. Most CMIP6 models, including CMIP6s MMM, fail to reproduce the acceleration in SIT decreasing. As a result, the difference in the declining rate of Arctic SIT between simulations and the observation is significant after 2000.

      The global warming slowed down at the end of the 20th century (IPCC, 2013), but the sea ice reduction accelerated in the 2010s, which means that in addition to global warming, the acceleration can be partly ascribed to the internal variability in the climate system, whose phase in a fully coupled climate model can be pretty different from reality. The North Atlantic variability can impact the Arctic sea ice by changing the heat transported to the Arctic. Day et al. (2012) found a significant correlation between the annual Atlantic multidecadal oscillation (AMO) index and Arctic sea ice extent (SIE), and ~0.3–3.1% (10 yr)–1 out of 10.1% (10 yr)–1 in the SIE decline is attributed to the AMO-driven variability during the period 1979–2010. Meanwhile, the historical realization of the CMIP5 and CMIP6 models generally failed to reproduce the low-frequency variability of the North Atlantic and associated links to SSTs, which could bring biases to the simulated Arctic SIT declining trend (Bracegirdle, 2022).

      The SIT trends in different seas of the Arctic Ocean vary among all models (Fig. 8). Most CMIP6 models underestimate the SIT declination in marginal seas; some (e.g., CAMS-CSM2-MR, CIESM, and CNRM-CM6-1) even show a positive SIT trend in the Kara Sea and the Barents Sea. The unrealistic simulation of Atlantic Water (AW), the primary input of warm water for the Arctic Ocean, may contribute to considerable SIT bias (Zhang, 2015; Shu et al., 2019). Some studies have pointed out that the CMIP5 models tend to show a thicker but slow-warming Atlantic Water in the historical simulation (Zhang, 2015). This problem is still significant in CMIP6 models (Shu et al., 2019), resulting in the weak heat flux from the North Atlantic into the Arctic and possibly causing the underestimation of the Arctic Ocean SIT decrease.

      Figure 8.  The trend versus the climatological mean of the SIT during 1979–2014 from CMIP6 models and PIOMAS in different seas. x0 and y0 are the X- and Y-axis intercepts, respectively; s is the linear regression coefficient, and p is the confidence level.

    • Putting all models’ results together, we can find that the higher the climatological mean SIT is, the faster the decreasing of SIT is during the study period (Fig. 8). The correlation between the mean SIT and SIT trend is beyond the confidence level of 99.9% in most seas of the Arctic Ocean, except for the Kara Sea (P>90%). The Arctic sea ice change is tied to its mean state through thermodynamic processes, as reported by Massonnet et al. (2018). The large spread of the climatological seasonal mean SIT of the CMIP6 models (as discussed in sections 3.1 and 3.2) makes reproducing the acceleration in SIT decreasing even more challenging. Regardless of the complicated mechanism modulating the Arctic sea ice change process, the relationship between the mean SIT and SIT trend might help us summarize some features of Arctic sea ice change.

      If we use X to represent the SIT, then the declining trend of SIT (the ordinate of Fig. 8) can be expressed as $ \mathrm{d}X/\mathrm{d}t $, where t represents the time. We can obtain the following formula by linear regression using the independent variables X and the dependent variables $ \mathrm{d}X/\mathrm{d}t $:

      where $ {\beta }_{0} $ is a constant term and $ \beta $ is the regression coefficient, which is also referred to as the SIT decrease ratio (DR) hereafter and represented by s in Fig. 8. Therefore, the decay of SIT with time would have the following formal solution:

      where C is a constant calculated using the SIT of CMIP6s MMM in 2014 as the initial value (i.e., t=0) in this study. We changed the period to 30 years during 1979–2014 and found that the choice of the period will not fundamentally alter the statistical results from Eq. (4). Thus, we give the results of 1979–2014 hereafter.

      Unlike the linear trend, DR ($ \beta $) reflects the exponential decay rate, which we believe is closer to the change in natural SIT. According to Fig. 8h, the DR for the Arctic Ocean is –18% (10 yr)–1. The minimum of DR appears in the Central Arctic, about –18% (10 yr)–1; then, in the Laptev Sea and the East Siberian Sea, the DR is about –17% (10 yr)–1; and the maximum DR is about –9% (10 yr)–1 in the Kara Sea. Since CMIP6s MMM gives an average of Arctic SIT of about 1.96 m in 1979 (Fig. 7) and an SIT trend of –0.23 m (10 yr)–1, the Arctic SIT will reduce to 0.2 m (10%) in 2056 and 0.02 m (1%) in 2063 under the same radiative forcing as the historical simulation. In contrast, if sea ice change follows the exponential decaying trend, the Arctic SIT will not linearly reduce to 0.22 m but eventually reach an equilibrium SIT of about 0.47 m.

      The x-intercept, $ -{\beta }_{0}/\beta $, represents the equilibrium SIT (ESIT) when the SIT trend equals zero under the historical forcing, i.e., the minimum SIT under the same historical forcing. The maximum ESIT appears in the Central Arctic Sea and the Laptev Sea, being more than 0.5 m; the minimums, both negative, appear in the Kara Sea and the Barents Sea, which means that these two areas will eventually become ice-free. ESIT for the Arctic Ocean is 0.48 m.

      In contrast, the y-intercept, $ {\beta }_{0} $, reflects the sea ice recovery potential (RP) when all the Arctic SIT has melted (an ice-free scenario). We suggest that RP reflects the sea ice’s recoverability, which means some processes could enforce the growth of local sea ice and resist the SIT decline caused by global warming. The RP for the Arctic Ocean is about 0.08 m (10 yr)–1. The greatest RP appears in the Central Arctic , being about 0.1 m (10 yr)–1, while the RPs of the Kara and the Barents Seas are both negative. Therefore, the Kara Sea and the Barents Sea most likely meet the ice-free scenario from their ESIT and RP.

      Being derived from ensemble models’ results, DR, ESIT, and RP reflect humans’ general understanding of the inherent features of the Arctic sea ice change under the specific radiative forcing. They are insensitive to both the systematic bias of numerical models and the internal climate variability, thus helping assess the features of climate change in different scenarios, especially in future projections.

    4.   Conclusion and discussion
    • The simulated Arctic sea ice thickness from 34 CMIP6 models is assessed against PIOMAS data for the period 1979–2014. Most models reasonably produce the spatial pattern of the climatological mean sea ice thickness, although significant inter-model spread exists. Increasing model resolution and properly choosing sea ice model physics can lead to improved SIT simulation. The averaged Arctic SIT of the selected CMIP6 models is about 0.19 m thinner than the PIOMAS data. The most considerable overestimation and underestimation for regional SIT appear in the East Siberian and the Central Arctic, respectively. Most models reproduce a much better sea ice pattern in the growing season than in the melting season. Regardless of the systematic bias, most models can produce the historical SIT declination in each sea and the entire Arctic Ocean.

      For the climatological mean and the trend of the Arctic SIT, three indicators, i.e., the decrease ratio (DR), the equilibrium SIT (ESIT), and the recovery potential (RP), are studied. Under the historical radiative forcing, DR is –18% decade–1, ESIT is 0.47 m, and RP is 0.08 m decade–1 for the Arctic. Different seas in the Arctic Ocean have similar DRs, ranging from –18% decade–1 to –9% decade–1, but their ESITs and RPs vary significantly among models. The Kara Sea and the Barents Sea have the two smallest ESITs and RPs, implying that they are most likely to be ice-free all year round in the future. In contrast, the Laptev Sea and Central Arctic are unlikely to be ice-free. Therefore, DR, ESIT, and RP can reflect the general features of sea ice change from the viewpoint of numerical simulations without knowing the detailed mechanism.

      Dynamic and thermodynamic processes modulate sea ice change (Spall, 2019; Yu et al., 2021). Some literature suggests that the thermodynamic processes play a more critical role in the sea ice loss in the whole Arctic Ocean (Bitz and Roe, 2004; Holland et al., 2006; Massonnet et al., 2018), especially in the Chukchi Sea, the East Siberian Sea, and the Kara Sea (Yu et al., 2021). The transpolar drift transports ice dynamically from the Pacific Arctic to the Atlantic Arctic and delivers sea ice out of the Arctic Ocean through the Fram Strait (Ricker et al., 2018; Min et al., 2019). SIT change on a regional scale owns spatial-dependent mechanisms. For example, the SIT change in the central basins is linked to the Atlantic water entering through the Fram Strait (Dörr et al., 2021; Kumar et al., 2021); the SIT in the Laptev Sea is also affected by river runoff (Holland et al., 2006; Park et al., 2020). Moreover, a quantitative analysis to distinguish sea ice’s thermodynamic and dynamic processes is even more complicated and has enormous uncertainties (Spall, 2019; Yu et al., 2021). And the three indicators’ regional differences and temporal variations can reflect the competitive relationship among various thermodynamic and dynamic processes in sea ice under a changing external forcing. But this might help us understand the future Arctic climate projections.

      Many factors contribute to the biases of model simulation results. Previous studies have shown that the model resolution, sea ice component, and the uncertainty of reference data could have an impact on the biases. Therefore, to improve the results’ reliability, it is necessary to use the data from satellite and in situ observation platforms for further comparison. Moreover, it also should be realized that the evaluation of model simulations should not be limited to the comparison of model simulation results and observation results for a single parameter but should focus on the differences between models and highlight the potential reasons for these differences, which is the next step of this work.

      Acknowledgements. This is a contribution to the Year of Polar Prediction (YOPP), a flagship activity of the Polar Prediction Project (PPP), initiated by the World Weather Research Programme (WWRP) of the World Meteorological Organization (WMO). This study was funded by the National Natural Science Foundation of China (Grant Nos. 41922044 and 41941009), the National Key R&D Program of China (Grant No. 2019YFA0607004 and 2022YFE0106300), the Guangdong Basic and Applied Basic Research Foundation (Grant Nos. 2020B1515020025 and 2019A1515110295), and the Norges Forskningsråd (Grant No. 328886).

Reference

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return