Advanced Search
Article Contents

Changes of Air-sea Coupling in the North Atlantic over the 20th Century


doi: 10.1007/s00376-014-4090-7

  • Changes of air-sea coupling in the North Atlantic Ocean over the 20th century are investigated using reanalysis data, climate model simulations, and observational data. It is found that the ocean-to-atmosphere feedback over the North Atlantic is significantly intensified in the second half of the 20th century. This coupled feedback is characterized by the association between the summer North Atlantic Horseshoe (NAH) SST anomalies and the following winter North Atlantic Oscillation (NAO). The intensification is likely associated with the enhancement of the North Atlantic storm tracks as well as the NAH SST anomalies. Our study also reveals that most IPCC AR4 climate models fail to capture the observed NAO/NAH coupled feedback.
  • 加载中
  • Bjerknes J., 1964: Atlantic air-sea interaction. Vol. 10, Advances in Geophysics, Academic Press, 1- 82.
    Carton J. A., B. S. Giese, 2008: A reanalysis of ocean climate using Simple Ocean Data Assimilation (SODA). Mon. Wea. Rev.,136, 2999-3017, doi: 10.1175/2007MWR1978.1.
    Cassou C., C. Deser, L. Terray, J. W. Hurrell, and M. Dr\'evillon, 2004: Summer sea surface temperature conditions in the North Atlantic and their impact upon the atmospheric circulation in early winter. J. Climate,17, 3349-3363, doi: 10.1175/ 1520-0442(2004)017<3349:SSSTCI>2.0.CO;2.
    Chang E. K. M., S. Lee, and K. Swanson, 2002: Storm track dynamics. J. Climate,15, 2163-2183, doi: 10.1175/1520-0442.
    Chen G., J. Lu, and D. M. W. Frierson, 2008: Phase speed spectra and the latitude of surface westerlies: Interannual variability and global warming trend. J. Climate, 21, 5942-5959, doi: 10.1175/2008JCLI2306.1.
    Compo G.P., Coauthors, 2011: The twentieth century reanalysis project. Quart. J. Roy. Meteor. Soc.,137, 1-28, doi: 10.1002/qj.776.
    Czaja A., C. Frankignoul, 2002: Observed impact of Atlantic SST anomalies on the North Atlantic Oscillation. J. Climate,15, 606-623, doi: 10.1175/1520-0442(2002)015 <0606:OIOASA>2.0.CO;2.
    Czaja A., A. W. Robertson, and T. Huck, 2003: The role of Atlantic ocean-atmosphere coupling in affecting North Atlantic Oscillation variability. Vol. 134, The North Atlantic Oscillation, Climatic Significance and Environmental Impact, J. W. Hurrell, Y. Kushnir, G. Ottersen, M. Visbeck, Eds., AGU Geophys. Monogr., 147- 172.
    Delworth T., 1996: North Atlantic interannual variability in a coupled ocean-atmosphere model. J. Climate,9, 2356-2375, doi: 10.1175/1520-0442(1996)009<2356:NAVIA>2.0.CO;2.
    Frankignoul C., E. Kestenare, 2005: Observed Atlantic SST anomaly impact on the NAO: An update. J.Climate, 18, 4089- 4094.
    Gan B., L. Wu, 2012: Modulation of atmospheric response to North Pacific SST anomalies under global warming: A statistical assessment. J. Climate,25, 6554-6566, doi: 10.1175/JCLI-D-11-00493.1.
    Gastineau G., F. D'Andrea, and C. Frankignoul, 2013: Atmospheric response to the North Atlantic Ocean variability on seasonal to decadal time scales . Climate Dyn.,40, 2311-2330, doi: 10.1007/s00382-012-1333-0.
    Hu Z.-Z., B. Huang, 2006: On the significance of the relationship between the North Atlantic Oscillation in early winter and Atlantic SST anomalies. J. Geophys. Res., 111,D12103, doi: 10.1029/2005JD006339.
    Hurrell J. W., 1995: Decadal trends in the North Atlantic Oscillation: Regional temperatures and precipitation. Science,269, 676-679, doi: 10.1126/science.269.5224.676.
    Hurrell J. W., C. Deser, 2009: North Atlantic climate variability: The role of the North Atlantic Oscillation. J. Mar. Syst.,78, 28-41, doi: 10.1016/j.jmarsys.2008.11.026.
    Hurrell J. W., Y. Kushnir, G. Ottersen, and M. Visbeck, 2003: An overview of the North Atlantic Oscillation. The North Atlantic Oscillation: Climatic Significance and Environmental Impact, Vol. 134, Geophys. Monogr., Amer. Geophys. Union, 1- 35.
    Hu Z.-Z., Z. H. Wu, 2004: The intensification and shift of the annual North Atlantic Oscillation in a global warming scenario simulation. Tellus (A),56, 112-124, doi: 10.1111/ j.1600-0870.2004.00050.
    Kaplan A., M. A. Cane, Y. Kushnir, A. C. Clement, M. B. Blumenthal, and B. Rajagopalan, 1998: Analyses of global sea surface temperature 1856-1991. J. Geophys. Res.,103, 18 567-18 589, doi: 10.1029/97JC01736.
    Kushnir Y., W. A. Robinson, I. Blade, N. M. J. Hall, S. Peng, and R. Sutton, 2002: Atmospheric GCM response to extratropical SST anomalies: Synthesis and evaluation. J.Climate, 15, 2233- 2256, doi: 10.1175/1520-0442(2002)015<2233: AGRTES>2.0.CO;2.
    Marshall J., H. Johnson, and J. Goodman, 2001a: A study of the interaction of the North Atlantic Oscillation with the ocean circulation. J. Climate,14, 1399-1421, doi: 10.1175/1520-0442(2001)014<1399:ASOTIO>2.0.CO;2.
    Marshall J., Coauthors, 2001b: North Atlantic climate variability: Phenomena,impacts and mechanisms. Int. J. Climatol., 21, 1863-1898, doi: 10.1002/joc.693.
    Mehta V. M., M. J. Suarez, J. V. Manganello, and T. Delworth, 2000: Oceanic influence on the North Atlantic Oscillation and associated Northern Hemisphere climate variations,1959-1993. Geophys. Res. Lett., 27, 121-124, doi: 10.1029/ 1999GL002381.
    Peng S., J. S. Whitaker, 1999: Mechanisms determining the atmospheric response to midlatitude SST anomalies. J. Climate,12, 1393-1408, doi: 10.1175/1520-0442(1999)012 <1393:MDTART>2.0.CO;2.
    Peng S., W. A. Robinson, S. Li, and M. P. Hoerling, 2005: Tropical Atlantic SST forcing of coupled North Atlantic seasonal responses. J. Climate,18, 480-496, doi: 10.1175/JCLI-3270.1.
    Rayner N. A., D. E. Parker, E. B. Horton, C. K. Folland , L. V. Alexand er, D. P. Rowell, E. C. Kent, and A. Kaplan, 2003: Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res., 108,4407, doi: 10.1029/2002JD002670.
    Rodwell M. J., D. P. Rowell, and C. K. Folland, 1999: Oceanic forcing of the wintertime North Atlantic Oscillation and European climate. Nature,398, 320-323, doi: 10.1038/18648.
    Smith T. M., R. W. Reynolds, T. C. Peterson, and J. Lawrimore, 2008: Improvements to NOAA's historical merged land ocean surface temperature analysis (1880-2006). J. Climate,21, 2283-2296, doi: 10.1175/2007JCLI2100.1.
    Timlin M. S., M. A. Alexand er, and C. Deser, 2002: On the reemergence of North Atlantic SST anomalies. J.Climate, 15, 2707- 2712.
    Thompson D. W. J., J. M. Wallace, 1998: The Arctic oscillation signature in the wintertime geopotential height and temperature fields. Geophys. Res. Lett.,25, 1297-1300, doi: 10.1029/98GL00950.
    Thompson D. W. J., J. M. Wallace, and G. C. Hegerl, 2000: Annular modes in the extratropical circulation. Part II: Trends. J. Climate,13, 1018-1036, doi: 10.1175/1520-0442(2000)013 <1018:AMITEC>2.0.CO;2.
    Visbeck M., E. Chassignet, R. Curry, T. Delworth, B. Dickson, and G. Rahmann, 2003: The ocean's response to NAO variability. The North Atlantic Oscillation: Climatic Significance and Environment Impact. Vol. 134, Geophys. Monogr., J. W. Hurrell et al., Eds, 113-146, doi: 10.1016/j.jmarsys.2008.11. 026.
    Wu L., Z. Liu, 2005: North Atlantic decadal variability: Air-sea coupling,oceanic memory, and potential Northern Hemisphere resonance. J. Climate, 18, 331-349, doi: 10.1175/ 1520-0442(2003)16<1101:PDVTTP>2.0.CO;2.
    Yang Y., L. Wu, and C. Fang, 2012: Will global warming suppress North Atlantic Tripole decadal variability? J. Climate,25, 2040-2055, doi: 10.1175/JCLI-D-11-00164.1.
    Yin J. H., 2005: A consistent poleward shift of the storm tracks in simulations of 21st century climate. Geophys. Res. Lett., 32,L18701, doi: 10.1029/2005GL023684.
    Zheng X.-T., S.-P. Xie, G. A. Vecchi, Q. Liu, and J. Hafner, 2010: Indian Ocean Dipole response to global warming: Analysis of ocean-atmospheric feedbacks in a coupled model. J. Climate,23, 1240-1253, doi: 10.1175/2009JCLI3326.1.
  • [1] Charlie C. F. LOK, Johnny C. L. CHAN, Ralf TOUMI, 2022: Importance of Air-Sea Coupling in Simulating Tropical Cyclone Intensity at Landfall, ADVANCES IN ATMOSPHERIC SCIENCES, 39, 1777-1786.  doi: 10.1007/s00376-022-1326-9
    [2] ZhouTianjun, Zhang Xuehong, Yu Yongqiang, Yu Rucong, Liu Xiying, Jin Xiangze, 2000: Response of IAP/ LASG GOALS Model to the Coupling of Air-Sea Fresh Water Exchange, ADVANCES IN ATMOSPHERIC SCIENCES, 17, 473-486.  doi: 10.1007/s00376-000-0037-2
    [3] REN Guoyu, DING Yihui, ZHAO Zongci, ZHENG Jingyun, WU Tongwen, TANG Guoli, XU Ying, 2012: Recent Progress in Studies of Climate Change in China, ADVANCES IN ATMOSPHERIC SCIENCES, 29, 958-977.  doi: 10.1007/s00376-012-1200-2
    [4] Gao Ge, Huang Chaoying, 2001: Climate Change and Its Impact on Water Resources in North China, ADVANCES IN ATMOSPHERIC SCIENCES, 18, 718-732.  doi: 10.1007/BF03403497
    [5] Zhang Tao, Guo Yufu, Wu Guoxiong, 2002: Analysis of the Zonal Mean Atmospheric Climate State in IAP/ LASG GOALS Model Simulations, ADVANCES IN ATMOSPHERIC SCIENCES, 19, 1091-1102.  doi: 10.1007/s00376-002-0067-z
    [6] Paul D. WILLIAMS, 2017: Increased Light, Moderate, and Severe Clear-Air Turbulence in Response to Climate Change, ADVANCES IN ATMOSPHERIC SCIENCES, 34, 576-586.  doi: 10.1007/s00376-017-6268-2
    [7] XU Yongfu, HUANG Yao, LI Yangchun, 2012: Summary of Recent Climate Change Studies on the Carbon and Nitrogen Cycles in the Terrestrial Ecosystem and Ocean in China, ADVANCES IN ATMOSPHERIC SCIENCES, 29, 1027-1047.  doi: 10.1007/s00376-012-1206-9
    [8] BUHE Cholaw, Ulrich CUBASCH, LIN Yonghui, JI Liren, 2003: The Change of North China Climate in Transient Simulations Using the IPCC SRES A2 and B2 Scenarios with a Coupled Atmosphere-Ocean General Circulation Model, ADVANCES IN ATMOSPHERIC SCIENCES, 20, 755-766.  doi: 10.1007/BF02915400
    [9] HAN Zuoqiang, YAN Zhongwei*, LI Zhen, LIU Weidong, and WANG Yingchun, 2014: Impact of Urbanization on Low-Temperature Precipitation in Beijing during 19602008, ADVANCES IN ATMOSPHERIC SCIENCES, 31, 48-56.  doi: 10.1007/s00376-013-2211-3
    [10] 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
    [11] Emily BLACK, 2024: Global Change in Agricultural Flash Drought over the 21st Century, ADVANCES IN ATMOSPHERIC SCIENCES, 41, 209-220.  doi: 10.1007/s00376-023-2366-5
    [12] FU Yuanhai, LU Riyu, 2010: Simulated Change in the Interannual Variability of South Asian Summer Monsoon in the 21st Century, ADVANCES IN ATMOSPHERIC SCIENCES, 27, 992-1002.  doi: 10.1007/s00376-009-9124-1
    [13] Lanying CHEN, Renhao WU, Qi SHU, Chao MIN, Qinghua YANG, Bo HAN, 2023: The Arctic Sea Ice Thickness Change in CMIP6’s Historical Simulations, ADVANCES IN ATMOSPHERIC SCIENCES, 40, 2331-2343.  doi: 10.1007/s00376-022-1460-4
    [14] Tianjun ZHOU, Wenxia ZHANG, Lixia ZHANG, Robin CLARK, Cheng QIAN, Qinghong ZHANG, Hui QIU, Jie JIANG, Xing ZHANG, 2022: 2021: A Year of Unprecedented Climate Extremes in Eastern Asia, North America, and Europe, ADVANCES IN ATMOSPHERIC SCIENCES, 39, 1598-1607.  doi: 10.1007/s00376-022-2063-9
    [15] LI Chongyin, ZHOU Wen, JIA Xiaolong, WANG Xin, 2006: Decadal/Interdecadal Variations of the Ocean Temperature and its Impacts on Climate, ADVANCES IN ATMOSPHERIC SCIENCES, 23, 964-981.  doi: 10.1007/s00376-006-0964-7
    [16] Aiguo Dai, G.A. Meehl, W.M. Washington, T.M.L. Wigley, 2001: Climate Changes in the 21st Century over the Asia-Pacific Region Simulated by the NCAR CSM and PCM, ADVANCES IN ATMOSPHERIC SCIENCES, 18, 639-658.
    [17] Yan XIA, Yongyun HU, Jiankai ZHANG, Fei XIE, Wenshou TIAN, 2021: Record Arctic Ozone Loss in Spring 2020 is Likely Caused by North Pacific Warm Sea Surface Temperature Anomalies, ADVANCES IN ATMOSPHERIC SCIENCES, 38, 1723-1736.  doi: 10.1007/s00376-021-0359-9
    [18] Johnny C. L. CHAN, Kin Sik LIU, 2022: Recent Decrease in the Difference in Tropical Cyclone Occurrence between the Atlantic and the Western North Pacific, ADVANCES IN ATMOSPHERIC SCIENCES, 39, 1387-1397.  doi: 10.1007/s00376-022-1309-x
    [19] SUN Guodong, MU Mu, 2011: Response of a Grassland Ecosystem to Climate Change in a Theoretical Model, ADVANCES IN ATMOSPHERIC SCIENCES, 28, 1266-1278.  doi: 10.1007/s00376-011-0169-6
    [20] Ge Ling, Liang Jiaxing, Chen Yiliang, 1996: Spatial / Temporal Features of Antarctic Climate Change, ADVANCES IN ATMOSPHERIC SCIENCES, 13, 375-382.  doi: 10.1007/BF02656854

Get Citation+

Export:  

Share Article

Manuscript History

Manuscript received: 29 April 2014
Manuscript revised: 26 June 2014
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

Changes of Air-sea Coupling in the North Atlantic over the 20th Century

  • 1. Physical Oceanography Laboratory, Ocean University of China, Qingdao 266100

Abstract: Changes of air-sea coupling in the North Atlantic Ocean over the 20th century are investigated using reanalysis data, climate model simulations, and observational data. It is found that the ocean-to-atmosphere feedback over the North Atlantic is significantly intensified in the second half of the 20th century. This coupled feedback is characterized by the association between the summer North Atlantic Horseshoe (NAH) SST anomalies and the following winter North Atlantic Oscillation (NAO). The intensification is likely associated with the enhancement of the North Atlantic storm tracks as well as the NAH SST anomalies. Our study also reveals that most IPCC AR4 climate models fail to capture the observed NAO/NAH coupled feedback.

1. Introduction
  • The dominant atmospheric variability over the North Atlantic Ocean, known as the North Atlantic Oscillation (NAO), is demonstrated as a redistribution of atmospheric mass between the Arctic and the subtropical North Atlantic, with strengthened (weakened) Icelandic Low and Subtropical High (Hurrell, 1995; Hurrell and Deser, 2009). The ocean, in the meantime, displays a tripole pattern, the North Atlantic Tripole (NAT), with cooling (warming) over the subpolar and eastern subtropical North Atlantic and warming (cooling) over the western subtropical North Atlantic (see reviews by Marshall et al., 2001b; Czaja et al., 2003; Hurrell et al., 2003). This NAO/NAT system is strongest in wintertime (Thompson and Wallace, 1998). In summer, the NAO still exists but features a weaker pattern, with the Icelandic Low and Subtropical High both weakened (strengthened). The NAT, however, evolves into a horseshoe pattern, the North Atlantic Horseshoe (NAH), with warming (cooling) over the western subtropical North Atlantic and cooling (warming) over part of the eastern ocean (Czaja and Frankignoul, 2002; Cassou et al., 2004).

    Figure 1.  Seasonal variations of squared covariance of the first MCA mode of midlatitude (20°-70°N) SST and Z500 anomalies for (a, c) 1950-99 and (b, d) 1900-49 using (a, b) Z500 and (c, d) SLP from 56 ensemble members of 20CRv2 and SST from HadISST. The shaded area indicates where the covariance is statistically significant at the 80%, 85% and 95% confidence level (light to dark shading, respectively). The x-axis denotes the months of Z500, and the y-axis indicates SST lags in months (positive when Z500 leads SST).

    Figure 2.  Spatial patterns of the first MCA mode during 1950-99 of SST (shading in K) and Z500 (contoured at 10 m intervals; solid, positive; dashed, negative) using Z500 from 56 ensemble members of 20CRv2 and SST from HadISST. The Z500 is fixed on NDJ, with SST lagged by months (positive as Z500 leads). The correlation coefficient r between the SST and Z500 MCA time series as well as the explained variance of MCA modes are given for each lag.

    Previous studies suggest that the NAO/NAT and NAO/ NAH systems are mainly sustained by atmospheric internal stochastic forcing (Bjerknes, 1964; Delworth, 1996; Marshall et al., 2001a); however, air-sea coupling also plays an important role in the reddening of the seasonal to decadal time scale fluctuations in this system (Rodwell et al., 1999; Mehta et al., 2000; Wu and Liu, 2005). Observational studies show that the air-sea coupling is strongest when the atmosphere and the ocean are in phase or when the former precedes by several months, which supports the theory that the NAO variability intrinsic to the atmosphere is responsible for the development of the NAT (Delworth, 1996; Visbeck et al., 2003; Wu and Liu, 2005). Strong coupling, however, also exists when the ocean leads, indicating that the ocean in turn has an influence on the NAO, instead of passively responding to the atmosphere (Czaja and Frankignoul, 2002). This coupling happens between the winter atmosphere and summer ocean, when the atmosphere displays a NAO pattern and the ocean a NAH pattern. Therefore, (Czaja and Frankignoul, 2002) explain this atmospheric wintertime air-sea coupling system as the NAO feedback, in response to the NAH anomalies in the previous summer, which can further generate the NAT SST anomalies in winter. The whole process illustrates the importance of summer SST in predicting winter atmosphere and ocean scenarios over the North Atlantic Ocean. This predictability may also be attributable to the re-emergence process (Timlin et al., 2002; Hu and Huang, 2006), but we focus on the summer SST contribution in this study. The observed NAO/NAT and NAO/NAH associations are further assessed by (Cassou et al., 2004) using a model simulation. However, (Peng et al., 2005) argue that the summer tropical Atlantic anomalies, instead of the NAH, are partly responsible for the winter NAO variability. In the 20th century, the circulations of both the atmosphere and the ocean over the North Atlantic were modulated significantly. This modulation included not only the mean state, such as the strengthening and poleward shift of the westerlies (e.g. Thompson et al., 2000; Chen et al., 2008), but also the regime shifts of climate variability, for instance the shift of the NAO action center (Hu and Wu, 2004) and the suppression of the NAT decadal variability (Yang et al., 2012). To understand these changes, it is important to comprehend the modulation of ocean-atmosphere coupling over the North Atlantic. This is the major focus of the current study. Studies suggest that the ocean-to-atmosphere feedback in the tropics is weakened due to an increase of static stability of the troposphere in response to global warming (e.g. Zheng et al., 2010). The coupling in the midlatitudes, however, is more complicated than in the tropics, due to the nonlinear interactions between eddies, jets and planetary waves (Kushnir et al., 2002). (Gan and Wu, 2012) find that a warm-ridge response over the North Pacific, usually identified in early winter, is intensified substantially in a warm climate, which suggests a strengthening of midlatitude air-sea coupling by global warming. In this paper, we investigate the modulation of air-sea coupling over the North Atlantic based on observations, reanalysis data, and multiple climate model data.

    The remainder of the paper is organized as follows. The data and method used in this study are described in section 2. The air-sea coupling and its change in global warming are investigated in sections 3 and 4. The mechanisms of this change are illustrated in section 5. The paper concludes with a summary and some further discussion in section 6.

2. Data and method
  • In this study, we use monthly mean observational SST data from the Hadley Centre Global Sea Ice and Sea Surface Temperature (HadISST) dataset (Rayner et al., 2003). It consists of data from 1870 to 2010, with a resolution of 1.0°× 1.0°. The monthly mean atmospheric data for geopotential height and heat flux are both from the Twentieth-Century Reanalysis version 2 (20CRv2) dataset (Compo et al., 2011) from 1871 to 2010, with a resolution of 2.0°× 2.0° on a global grid. This atmospheric reanalysis data consists of 56 runs and we use both the ensemble mean and individual runs. To verify the results, we also use another three SST datasets: Simple Ocean Data Assimilation (SODA) (Carton and Giese, 2008); National Oceanic and Atmospheric Administration Extended Reconstructed SST version 3 (ERSSTv3b) (Smith et al., 2008); and Kaplan Extended SST version 2 (Kaplanv2) (Kaplan et al., 1998). In addition, we use model outputs of the 20th century simulation for the Fourth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC AR4 20c3m).

    To quantify the strength of the air-sea coupling, we use the maximum covariance analysis (MCA) method, the same method used in the study of (Czaja and Frankignoul, 2002). This technique is based on the singular value decomposition (SVD) of the temporal covariance matrix between SST and an atmospheric variable, say 500 hPa geopotential height (Z500). Both variables, with their seasonal cycle and trend removed (using third-order polynomial), are weighted by the square root of the cosine of latitude, and normalized by the corresponding mean (domain averaged) seasonal cycle of standard deviation. Each MCA is repeated 100 times, with a randomized order of SST and Z500 sequence of years to ensure these two variables do not correlate at the same time. The significance level is calculated as the percentage of randomized squared covariance below that of the value being tested. The MCA is a useful method in statistically distinguishing causal relationships. If the ocean responds passively to atmospheric forcing, there should be no covariance when SST leads by more than the atmospheric persistence time; and if SST can influence the atmosphere, their cross covariance does not vanish when SST leads. In this paper, we focus on the extratropical North Atlantic over the domain (20°-70°N, 100°W-20°E).

3.Air-sea coupling over the North Atlantic Ocean
  • Figure 3.  Regression of SST (shading in K) and net surface heat flux (contoured at 5 W m-2 intervals; solid, positive; dashed, negative) against the standardized MCA PC1 index during 1950-99. The index is set with SST lagged from -6 to 5 and Z500 fixed at NDJ. SST and net surface heat flux are regressed simultaneously and at one-month lead against the PC index, respectively. Heat flux is defined positive (negative) when downward (upward).

    To study the air-sea coupling, we examine the squared covariance (SC) between monthly SST and Z500 from HaDISST and 20CRv2, respectively. The air-sea coupling is measured by SC passing the significance test, and the ocean-to-atmosphere feedback represents the coupling when SST leads Z500. The SC in the period 1950-99 is similar to that in (Czaja and Frankignoul, 2002) in spite of the different atmospheric and SST dataset used (Fig. 1a). Strong SC can be found in all seasons when Z500 leads SST anomalies, indicating a dominant forcing of the atmosphere in all seasons. The spatial patterns of the first MCA modes display a negative NAO pattern with weakening of both the Icelandic Low and Subtropical High, accompanying a NAT pattern with cooling in the western subtropics and warming in the subpolar region and eastern subtropics (Figs. 2a-c). The correlation between the first MCA time series of SST and Z500 reaches 0.73 at lag 1 month. This tight association between atmosphere and ocean suggests the NAO variability intrinsic to the atmosphere is responsible for the development of the NAT.

    Significant SC can also be found when SST leads early winter Z500 by 3 to 6 months (Fig. 1a). The corresponding pattern between the atmosphere and the ocean, however, evolves from the NAO/NAT to the NAO/NAH (Figs. 2d-i), with the cold SST anomalies retreating and warm anomalies growing, and together forming a horseshoe pattern. This relationship persists even when SST precedes Z500 by 6 months, but the corresponding NAO anomalies are much weaker than those in winter (Fig. 2i vs. c). The correlation between the MCA time series decreases with time (Figs. 2d-i) but still remains significant, with a magnitude of 0.39 at lag -6 months. This NAO/NAH corresponding relationship indicates that the atmosphere anomalies generated by the NAH can partially contribute to the NAO anomalies in winter.

    To further explore the mechanism of the NAO/NAT and NAO/NAH systems, we study the evolution of SST anomalies from the NAH to the NAT. We calculate the lead-lag regression of the net surface heat flux onto standardized MCA PC1s, with Z500 fixed at November-January (NDJ) (Fig. 3). It can be seen that the heat flux is highly responsible for the following month's SST anomalies, with latent and sensible heat flux explaining about 2/3 and 1/3 of heat flux anomalies, respectively (not shown). Specifically, the summer heat flux anomalies feature a horseshoe pattern with negative anomalies over the west and positive anomalies in part of the eastern basin, in response to which the SST anomalies show a similar pattern in the following month (Figs. 3a-e). In October-December (OND), the surface heat flux anomalies transform from a horseshoe to a tripole pattern, which forces the SST anomalies to evolve from the NAH to the NAT pattern (Fig. 3f). After OND, the surface heat flux displays a tripole pattern and further forces the tripole SST anomalies (Figs. 3g-l). Therefore, the evolution of SST from the NAH to the NAT is attributed to the surface heat flux forcing.

    Figure 4.  As in Fig. 1, except for IPCC models during 1950-99.

    The surface heat flux associated with the NAH (summer ocean) also helps to form the NAT in winter. By regressing OND heat flux onto the MCA PC1 index with SST and Z500 fixed on June-August (JJA) and NDJ (not shown), it is found that the heat flux anomaly can still help to form a NAT pattern in NDJ, although this anomaly is weaker than that in Fig. 3g. Moreover, the atmospheric response displays a NAO-like pattern (Figs. 2f-i), which can also sustain the NAT anomalies.

    Figure 5.  Time series of (a) averaged SC and (b) normalized SC at SST lagging from -6 to -3 months with Z500 fixed on OND to DJF. The thick black line demonstrates the 56 ensemble mean SC and shading illustrates the maximum and minimum SC of the 56 individual runs.

    In summary, the air-sea coupling is robust in the second half of the 20th century. Although it is dominated by the forcing of the atmosphere on the ocean, the ocean-to-atmosphere feedback also exists, featuring the summer NAH influencing the winter NAO. The results are consistent with (Czaja and Frankignoul, 2002), although different reanalysis datasets are used.

    Figure 6.  (a-d) As in Fig. 2, except for (a, b) bccr bcm2_0 and (c, d) ingv_echam4. The left panel is for winter SST and atmosphere, and the right for summer SST and winter atmosphere. (e-f) As in Fig. 1, except for (e) bccr bcm2_0 and (f) ingv_echam4 during 1900-49.

4. Changes of air-sea coupling in the 20th century
  • In the first half of the 20th century (1900-49), the SC remains significant when Z500 leads SST (Fig. 1b), with a coupled NAO/NAT pattern (not shown) consistent with that in the second half of the 20th century (1950-99). However, there is no strong SC signal when SST leads, forming a sharp contrast to that in the second half of the 20th century. Moreover, the association between the NAO and the NAH does not exist (not shown). When calculating the averaged SC time series with SST lagged from -3 to -6 months and Z500 fixed on OND to DJF (thick black line in Fig. 4a), the SC displays an increasing trend superimposed with some decadal variability. This indicates an enhancement of the ocean-to-atmosphere feedback over the North Atlantic in the second half of the 20th century.

    The result above is based on the ensemble mean of 56 runs of the 20CRv2; however, the changes of air-sea coupling may be potentially biased by the ensemble mean (Compo et al., 2011). To further verify this enhancing of the ocean-to-atmosphere feedback, we calculate the averaged SC time series for each run. The SC in ensemble members displays a relatively larger dispersion in the early period than in the late period, likely due to sparse observations of the former. Nevertheless, it is also clear that the SC increases with time, suggesting an enhancement of the ocean-to-atmosphere feedback in the 20th century.

    To further substantiate the results drawn from reanalysis data, we use the IPCC AR4 20c3m experiments (Table 1). First, we assess whether these models can reproduce the observed SC over the North Atlantic during 1950-99. This serves as a criterion to select models to be used in our study. Although almost all models can reproduce the observed SC with Z500 preceding SST, most of them fail to capture the significant SC when SST leads the winter Z500 by 3 to 6 months (Fig. 5), except the bccr_bcm2_0 and ingv_echam4 models (Figs. 5a and k). Both models capture the coupled NAO/NAT pattern in early winter (Figs. 6a and c), although the subpolar lobe of the NAT in the ingv_echam4 model is too strong (Fig. 6c). The spatial patterns of summer SST and following winter Z500 in both models display a coupled NAO/NAH pattern (Figs. 6b and d), resembling that in the observations (Fig. 2). In the first half of the 20th century (1900-49), the significant SC disappears in ingv_echam4 and becomes less robust in bccr_bcm2_0 when SST leads early winter Z500 (Fig. 5a vs. 6e, 5k vs. 6f). Moreover, the associated NAO/NAH spatial pattern no longer exists in both models (not shown). The weakening of the coupling between the winter Z500 and the summer SST during 1900-49 in both models is consistent with the reanalysis, further indicating an enhancement of ocean-to-atmosphere feedback in the late 20th century.

5. Potential mechanisms
  • Although air-sea coupling over the North Atlantic can be influenced by the tropics, (Frankignoul and Kestenare, 2005) argue that the bulk of the NAO signals comes from the midlatitudes, based on a separate analysis of midlatitude (20°-70°N) and tropical (20°S-20°N) SST anomalies. Therefore, it is possible that the enhanced ocean-to-atmosphere feedback over the North Atlantic in the second half of the 20th century may be associated with changes of atmosphere and/or ocean over the North Atlantic.

  • The ocean-to-atmosphere feedback in the midlatitudes has been suggested to be associated with strong nonlinear interaction between synoptic eddies, stationary waves and the jet stream (e.g., Kushnir et al., 2002). Considering the intimate relationship between the midlatitude ocean-to-atmosphere feedback and transient eddy feedback tied to storm tracks (Peng and Whitaker, 1999; Chang et al., 2002), we examine the changes in storm tracks in the 20th century using daily data of the 56 ensemble runs of the 20CRv2 reanalysis product. The storm track is defined as bandpass-filtered (2-8 days) 200 hPa standard deviation of geopotential height in DJF. The climatology of storm tracks over 1950-99 using 20CRv2 (Fig. 7b) resembles the observed (Chang et al., 2002), with a maximum across the midlatitudes. It is found that the storm track index [storm tracks averaged over (40°-60°N, 20°-80°W)] demonstrates a steady increasing trend in the 20th century (Fig. 7a). The individual runs in the early period display relatively larger fluctuation than in the late period, due to sparse observations in the early period. Nevertheless, the increasing trend is clear for each individual run (not shown) and ensemble mean. When comparing the early (1900-49) and late (1950-99) periods of the 20th century, the storm tracks are enhanced over the entire North Atlantic Ocean, with an increase up to 12% over the subpolar ocean (Fig. 7c). The intensification of storm tracks in a warm climate has also been identified in 21st century climate projections from IPCC AR4 (Yin, 2005). Such intensifi-cation may amplify the midlatitude transient eddy feedback, and thus the ocean-to-atmosphere dynamic feedback.

    Figure 7.  (a) Temporal evolution of storm tracks (gpm) averaged with a 50-yr sliding window over the green box in Fig. 7b (thick black line for the mean, shading for the standard variation of 56 ensemble members). (b) Climatology of storm tracks (gpm) during 1950-99. (c) Ratio (percentage) of storm track changes in two different periods: 1900-49 and 1950-99; shading indicates significance passed the 90% confidence level.

    Figure 8.  (a, b) The second EOF modes of JJA SST and (c, d) the regression of NDJ 500 hPa geopotential height onto corresponding EOF PC index during 1950-99 (top row) and 1900-49 (bottom row). Units for SST EOFs and Z500 regression are °C and gpm, respectively.

    Figure 9.  As in Fig. 8, except (a-d) are for bccr bcm2_0 and (e-h) for ingv echam4 during 1950-99 (first and third rows) and 1900-49 (second and last rows).

    Figure 10.  EOF modes of JJA SST for (a, b) ERSSTv3b, (c, d) Kaplanv2 and (e, f) SODA during 1950-99 (left panel) and 1900-49 (right panel). Units for SST EOFs are °C.

  • Section 3 suggests that the summer NAH is closely related to the winter NAO variability. Therefore, we examine whether the summer NAH changes in the 20th century.

    The NAH is enhanced over the 20th century. In the second half of the 20th century, the second mode of summer SST features a typical NAH mode, with cold anomalies in the western subtropics and warm anomalies in part of the eastern basin (Fig. 8a). In response to this NAH forcing, the winter (NDJ) atmosphere features a NAO pattern with the Icelandic Low and Subtropical High both weakened (Fig. 8b). Here, we call this SST mode an effective NAH mode, which not only shows a horseshoe spatial pattern but also contributes to winter NAO variability. In the first half of the 20th century, the second EOF of summer SST shows a weak and not well-formed horseshoe pattern (Fig. 8c); moreover, the corresponding winter atmospheric regression does not display a dipole pattern (Fig. 8d).

    The strengthening of the NAH in the second half of the 20th century can also be seen in the two IPCC models that capture coupling between summer SST and winter Z500 during 1950-99. In the second half of the 20th century, the two models display a NAH pattern as the second and first EOF mode of summer SST (Figs. 9a and e), respectively, although the eastern lobe is too strong in the ingv_echam4 model. In response to these SST forcings, the winter atmosphere in both models displays a NAO pattern (Figs. 9b and f). During 1900-49, however, neither of these two models captures an effective NAH signal. The first EOF modes of both models resemble a horseshoe pattern. However, the corresponding winter Z500 regressions show a NAO-like pattern but with opposite sign and a non-NAO-related pattern for the bccr_bcm2_0 and ingv_echam4 models, respectively (Figs. 9c and d). For those models that fail to capture the coupling between the summer SST and winter atmosphere, most of them do not display an effective NAH signal. Some of these models cannot capture EOFs with a horseshoe pattern for summer SST, while other models either yield too weak or opposite NAO anomalies in response to the horseshoe EOF. Only the ukmo_hadcm3 model can produce an effective NAH, and this is the model that captures a weak ocean-to-atmosphere feedback (Fig. 5t).

    We also investigate the NAH changes in the 20th century using other SST reanalysis data (ERSSTv3b, Kaplanv2 and SODA). During the second half of the 20th century, all these three datasets display a well-formed NAH signal as the second EOF for JJA SST (Figs. 10a, c and e). In the first half of the 20th century, however, the second summer SST modes in these datasets display a tripole pattern, instead of a horseshoe pattern (Figs. 10b, d and f). Moreover, the amplitude of the western pole is much weaker than that in the second half of the 20th century. In addition, all of these EOF modes are associated with the following winter NAO in the second half of the 20th century, but none in the first half of the 20th century, based on the 20CRv2 data.

6. Summary and discussion
  • In this paper, the changes of air-sea coupling in the North Atlantic Ocean over the 20th century are investigated using 20CRv2 reanalysis data, climate model simulations, and observational data. The 20CRv2 reanalysis data capture the observed ocean-to-atmosphere feedback during the second half of the 20th century (Czaja and Frankignoul, 2002; Cassou et al., 2004; Gastineau et al., 2013). In the early 20th century, the ocean-to-atmosphere feedback weakens in both the reanalysis data and the climate model simulations, suggesting an intensification trend of air-sea coupling over the North Atlantic in the 20th century.

    The enhancement of this feedback may be related with changes of both atmosphere and ocean in the 20th century. Over the 20th century, the North Atlantic storm tracks strengthen, which may amplify the midlatitude transient eddy feedback, and thus the atmospheric response to SST. In addition, the summer NAH appears to be stronger in the late 20th century, which may also lead to a stronger NAO response in winter. The mechanism for the enhancement of the NAH is beyond the scope of the current study. It is conceivable that the shallowing of the mixed layer in warm climate appears to favor a stronger SST in response to atmospheric forcing. However, we admit that understanding the dynamics of ocean-atmosphere coupling in the midlatitudes remains a challenge.

    The poor ability of climate models in simulating the coupling between the winter atmosphere and preceding summer SST remains an obstacle in predicting the climate variability over the North Atlantic. Only 2 out of 24 models can capture this coupling found in the observation during 1950-99. Moreover, over half of the models fail to simulate the summer NAH pattern. Therefore, it remains a great challenge to improve model ability in simulating and predicting the North Atlantic climate variability.

    The 20CRv2 data might be questioned to be responsible for the enhancement of air-sea coupling over the 20th century, due to lack of observations in the upper-level atmosphere over the early part of the 20th century. Therefore, we repeat the MCA analysis using SLP data (which has rich observations in the first half of the 20th century, relative to the upper troposphere) and attain similar results as Z500 (Figs. 1c and d). Moreover, the trend of variability in both the atmosphere (e.g. strengthening of storm tracks) and ocean (e.g. warming of SST) might not be entirely removed, although we did perform trend removal before the MCA analysis. To better assess the strength of ocean-atmosphere feedback, we re-plot the averaged SC time series, which is scaled by the variance of SST and Z500, to remove the positive variance trend in both ocean and atmosphere (Fig. 4b). The SC still shows a positive increasing trend, which further proves the enhancement of air-sea coupling over the 20th century.

    One should also keep in mind that the MCA method cannot remove noise thoroughly, such as the strong SC showed in the summer atmosphere with SST leading in the period 1900-49 (Fig. 1b), and in the IPCC results (Fig. 5). To distinguish between noise and air-sea coupling, we need to check the spatial patterns of ocean and atmosphere to confirm whether this strong SC has physical meaning.

Reference

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return