• Aagaard, K., , J. H. Swift, , and E. C. Carmack, 1985: Thermohaline circulation in the Arctic Mediterranean Seas. J. Geophys. Res., 90 (C3), 48334846, doi:10.1029/JC090iC03p04833.

    • Search Google Scholar
    • Export Citation
  • Ådlandsvik, B., , and H. Loeng, 1991: A study of the climatic system in the Barents Sea. Polar Res., 10, 4550, doi:10.1111/j.1751-8369.1991.tb00633.x.

    • Search Google Scholar
    • Export Citation
  • Alexander, M. A., , U. S. Bhatt, , J. E. Walsh, , M. S. Timlin, , J. S. Miller, , and J. D. Scott, 2004: The atmospheric response to realistic Arctic sea ice anomalies in an AGCM during winter. J. Climate, 17, 890905, doi:10.1175/1520-0442(2004)017<0890:TARTRA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Årthun, M., , T. Eldevik, , L. H. Smedsrud, , Ø. Skagseth, , and R. B. Ingvaldsen, 2012: Quantifying the influence of Atlantic heat on Barents Sea ice variability and retreat. J. Climate, 25, 47364743, doi:10.1175/JCLI-D-11-00466.1.

    • Search Google Scholar
    • Export Citation
  • Bengtsson, L., , V. A. Semenov, , and O. M. Johannessen, 2004: The early twentieth-century warming in the Arctic—A possible mechanism. J. Climate, 17, 40454057, doi:10.1175/1520-0442(2004)017<4045:TETWIT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Bitz, C. M., , M. M. Holland, , E. C. Hunke, , and R. E. Moritz, 2005: On the maintenance of sea-ice edge. J. Climate, 18, 29032921, doi:10.1175/JCLI3428.1.

    • Search Google Scholar
    • Export Citation
  • Blanchard-Wrigglesworth, E., , K. C. Armour, , C. M. Bitz, , and E. DeWeaver, 2011: Persistence and inherent predictability of Arctic sea ice in a GCM ensemble and observations. J. Climate, 24, 231250, doi:10.1175/2010JCLI3775.1.

    • Search Google Scholar
    • Export Citation
  • Bretherton, C. S., , M. Widmann, , V. P. Dymnikov, , J. M. Wallace, , and I. Bladé, 1999: The effective number of spatial degrees of freedom of a time-varying field. J. Climate, 12, 19902009, doi:10.1175/1520-0442(1999)012<1990:TENOSD>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Dee, D. P., and et al. , 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553597, doi:10.1002/qj.828.

    • Search Google Scholar
    • Export Citation
  • Deser, C., , G. Magnusdottir, , R. Saravanan, , and A. Philips, 2004: The effects of North Atlantic SST and sea ice anomalies on the winter circulation in CCM3. Part II: Direct and indirect components of the response. J. Climate, 17, 877889, doi:10.1175/1520-0442(2004)017<0877:TEONAS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Deser, C., , R. A. Tomas, , and S. Peng, 2007: The transient atmospheric circulation response to North Atlantic SST and sea ice anomalies. J. Climate, 20, 47514767, doi:10.1175/JCLI4278.1.

    • Search Google Scholar
    • Export Citation
  • Ferreira, D., , and C. Frankignoul, 2005: The transient atmospheric response to midlatitude SST anomalies. J. Climate, 18, 10491067, doi:10.1175/JCLI-3313.1.

    • Search Google Scholar
    • Export Citation
  • Francis, J. A., , and E. Hunter, 2007: Drivers of declining sea ice in the Arctic winter: A tale of two seas. Geophys. Res. Lett., 34, L17503, doi:10.1029/2007GL030995.

    • Search Google Scholar
    • Export Citation
  • Francis, J. A., , and S. J. Vavrus, 2012: Evidence linking Arctic amplification to extreme weather in mid-latitudes. Geophys. Res. Lett., 39, L06801, doi:10.1029/2012GL051000.

    • Search Google Scholar
    • Export Citation
  • Furevik, T., 2001: Annual and interannual variability of Atlantic Water temperatures in the Norwegian and Barents Seas. Deep-Sea Res. I, 48, 383404, doi:10.1016/S0967-0637(00)00050-9.

    • Search Google Scholar
    • Export Citation
  • Hendon, H. H., , and D. L. Hartmann, 1982: Stationary waves on a sphere: Sensitivity to thermal feedback. J. Atmos. Sci., 39, 19061920, doi:10.1175/1520-0469(1982)039<1906:SWOASS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Holopainen, E. O., 1978: On the dynamic forcing of the long-term mean flow by the large-scale Reynolds’ stresses in the atmosphere. J. Atmos. Sci., 35, 15961604, doi:10.1175/1520-0469(1978)035<1596:OTDFOT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Honda, M., , K. Yamazaki, , H. Nakamura, , and K. Takeuchi, 1999: Dynamic and thermodynamic characteristics of atmospheric response to anomalous sea-ice extent in the Sea of Okhotsk. J. Climate, 12, 33473358, doi:10.1175/1520-0442(1999)012<3347:DATCOA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hoskins, B. J., , and D. J. Karoly, 1981: The steady linear response of a spherical atmosphere to thermal and orographic forcing. J. Atmos. Sci., 38, 11791196, doi:10.1175/1520-0469(1981)038<1179:TSLROA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Ikeda, M., 1990: Decadal oscillations of the air–ice–ocean system in the Northern Hemisphere. Atmos.–Ocean, 28, 106139, doi:10.1080/07055900.1990.9649369.

    • Search Google Scholar
    • Export Citation
  • Inoue, J., , M. E. Hori, , and K. Takaya, 2012: The role of Barents Sea ice in the wintertime cyclone track and emergence of a warm-Arctic cold-Siberian anomaly. J. Climate, 25, 25612568, doi:10.1175/JCLI-D-11-00449.1.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and et al. , 1996: The NCEP–NCAR 40-yr Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437471, doi:10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Koenigk, T., , and L. Brodeau, 2014: Ocean heat transport into the Arctic in the twentieth and twenty-first century in EC-Earth. Climate Dyn., 42, 31013120, doi:10.1007/s00382-013-1821-x.

    • Search Google Scholar
    • Export Citation
  • Koenigk, T., , U. Mikolajewicz, , J. H. Jungclaus, , and A. Kroll, 2009: Sea ice in the Barents Sea: Seasonal to interannual variability and climate feedbacks in a global coupled model. Climate Dyn., 32, 11191138, doi:10.1007/s00382-008-0450-2.

    • Search Google Scholar
    • Export Citation
  • Koenigk, T., , L. Brodeau, , R. Graversen, , J. Karlsson, , G. Svensson, , M. Tjernström, , U. Willén, , and K. Wyser, 2013: Arctic climate change in 21st century CMIP5 simulations with EC-Earth. Climate Dyn., 40, 27192743, doi:10.1007/s00382-012-1505-y.

    • Search Google Scholar
    • Export Citation
  • Kushnir, Y., , W. A. Robinson, , I. Bladé, , N. M. Hall, , S. Peng, , and R. Sutton, 2002: Atmospheric GCM response to extratropical SST anomalies: Synthesis and evaluation. J. Climate, 15, 22332256, doi:10.1175/1520-0442(2002)015<2233:AGRTES>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Liu, Y., , J. R. Key, , Z. Liu, , X. Wang, , and S. J. Vavrus, 2012: A cloudier Arctic expected with diminishing sea ice. Geophys. Res. Lett., 39, L05705, doi:10.1029/2012GL051251.

    • Search Google Scholar
    • Export Citation
  • Overland, J. E., , and M. Wang, 2010: Large-scale atmospheric circulation changes are associated with the recent loss of Arctic sea ice. Tellus, 62A, 19, doi:10.1111/j.1600-0870.2009.00421.x.

    • Search Google Scholar
    • Export Citation
  • Palmer, T. N., , and S. Zhaobo, 1985: A modeling and observational study of the relationship between sea surface temperature in the northwest Atlantic and the atmospheric general circulation. Quart. J. Roy. Meteor. Soc., 111, 947975, doi:10.1002/qj.49711147003.

    • Search Google Scholar
    • Export Citation
  • Pedlosky, J., 1987: Geophysical Fluid Dynamics. 2nd ed. Springer-Verlag, 710 pp.

  • Petoukhov, V., , and V. A. Semenov, 2010: A link between reduced Barents-Kara sea ice and cold winter extremes over northern continents. J. Geophys. Res., 115, D21111, doi:10.1029/2009JD013568.

    • Search Google Scholar
    • Export Citation
  • Polyakov, I. V., and et al. , 2011: Fate of early 2000s Arctic warm water pulse. Bull. Amer. Meteor. Soc., 92, 561566, doi:10.1175/2010BAMS2921.1.

    • Search Google Scholar
    • Export Citation
  • Schlichtholz, P., 2011: Influence of oceanic heat variability on sea ice anomalies in the Nordic Seas. Geophys. Res. Lett., 38, L05705, doi:10.1029/2010GL045894.

    • Search Google Scholar
    • Export Citation
  • Schlichtholz, P., 2013: Observational evidence for oceanic forcing of atmospheric variability in the Nordic seas area. J. Climate, 26, 29572975, doi:10.1175/JCLI-D-11-00594.1.

    • Search Google Scholar
    • Export Citation
  • Schlichtholz, P., , and M.-N. Houssais, 2011: Forcing of oceanic heat anomalies by air–sea interactions in the Nordic Seas area. J. Geophys. Res., 116, C01006, doi:10.1029/2009JC005944.

    • Search Google Scholar
    • Export Citation
  • Seager, R., , D. S. Battisti, , J. Yin, , N. Gordon, , N. Naik, , A. C. Clement, , and M. A. Cane, 2002: Is the Gulf Stream responsible for Europe’s mild winters? Quart. J. Roy. Meteor. Soc., 128, 25632586, doi:10.1256/qj.01.128.

    • Search Google Scholar
    • Export Citation
  • Skagseth, Ø., 2008: Recirculation of Atlantic Water in the western Barents Sea. Geophys. Res. Lett., 35, L11606, doi:10.1029/2008GL033785.

    • Search Google Scholar
    • Export Citation
  • White, W. B., , and S.-C. Chen, 2002: Thermodynamic mechanisms responsible for the tropospheric response to SST anomalies in the Antarctic circumpolar wave. J. Climate, 15, 25772596, doi:10.1175/1520-0442(2002)015<2577:TMRFTT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • White, W. B., , P. Glorsen, , and I. Simmonds, 2004: Tropospheric response in the Antarctic circumpolar wave along the sea ice edge. J. Climate, 17, 27652779, doi:10.1175/1520-0442(2004)017<2765:TRITAC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Wu, B., , J. Wang, , and J. Walsh, 2004: Possible feedback of winter sea ice in the Greenland and Barents Seas on the local atmosphere. Mon. Wea. Rev., 132, 18681876, doi:10.1175/1520-0493(2004)132<1868:PFOWSI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Yang, S., , and J. H. Christensen, 2012: Arctic sea ice reduction and European cold winters in CMIP5 climate change experiments. Geophys. Res. Lett., 39, L20707, doi:10.1029/2012GL053338.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Climatological mean of the winter (December–March) air temperature (°C, contours) and geostrophic wind (m s−1, arrows scaled as in the bottom right corner) at 1000 hPa and sea ice extent (sea ice concentration above 15%, shading) in the Nordic seas area based on data in the period 1982–2006 from the NCEP–NCAR reanalysis. The box delineates the Barents Sea opening (BSO) area.

  • View in gallery

    Anomalies of the air temperature (contours) and geostrophic wind (arrows) at (a) 1000, (b) 850, and (c) 500 hPa in the extended Nordic seas area in winters 1982/83–2005/06 regressed onto the previous summer AWT index. The contour interval (CI) for the temperature anomalies is 0.2°C per unit AWT index. Red and blue contours represent positive and negative anomalies, respectively. The zero contour is omitted, and pink and aquamarine shading denote positive and negative anomalies statistically significant at the 90% confidence level, respectively. The wind anomalies are in meters per second per unit AWT index (scaled as in the bottom right corner and masked if both components are nonsignificant at the 90% confidence level). The thick yellow line shows the climatological mean position of the winter ice edge (15% sea ice concentration contour). In (a) and (c), the magenta box indicates the area used for calculating the average zonal wind at the entrance to the Barents Sea and average air temperature over the Barents Sea, respectively (see Figs. 4a,b).

  • View in gallery

    Anomalies of (a) the sea ice concentration, (b) skin temperature, (c) turbulent surface heat flux (positive upward), and (d) surface wind stress (arrows) and Ekman pumping velocity (positive downward) at the top of the atmospheric planetary boundary layer (contours) in the extended Nordic seas area in winters 1982/83–2005/06 regressed onto the previous summer AWT index. In (a)–(d), the CI is 5%, 0.5°C, 10 W m−2, and 2 × 10−3 Pa s−1 per unit AWT index, respectively. In (d), the wind stress anomalies are in newtons per square meter per unit AWT index (scaled as in the bottom-right corner and masked if both components are nonsignificant at the 90% confidence level). The contour and shading colors and the thick yellow line are explained in the caption to Fig. 2. In (a) and (c), the black box indicates the area over which the sea ice concentration and surface heat flux anomalies are integrated (see Figs. 17 and 5, respectively).

  • View in gallery

    Vertical section of the time-lagged regression coefficients of the anomalies of (a) air temperature averaged over the magenta box in Fig. 2c, (b) westerly geostrophic velocity averaged over the magenta box in Fig. 2a, and (c) vertical velocity (positive downward) averaged over the black box in Fig. 6f regressed onto the summer AWT index. In (a)–(c), the CI is 0.2°C, 0.2 m s−1, and 1 × 10−3 Pa s−1 per unit AWT index, respectively. The contour and shading colors are explained in the caption to Fig. 2. The thick black contours are the correlation coefficients (only contours of |r| ≥ 0.4 are plotted). Positive lags correspond to the AWT index leading the atmospheric variables calculated as 4-month averages with the interval of 1 month. Lag 6 months corresponds to winter (December–March).

  • View in gallery

    Magnitude of the time-lagged regression coefficients of the anomalies of the turbulent surface heat flux regressed onto the summer AWT index and then integrated over the positive (circles) and negative (triangles) lobes of the regression pattern over the Nordic seas area (black box in Fig. 3c). The anomalies are in 1012 W per unit AWT index. Positive lags correspond to the AWT index leading the surface heat flux anomalies calculated as 4-month averages with the interval of 1 month. Lag 6 months corresponds to winter (December–March).

  • View in gallery

    Anomalies of the geopotential height (contours) and geostrophic wind (arrows) at (a) 1000 and (b) 300 hPa, total wind (arrows) and its curl (contours) at (c) 1000 and (d) 300 hPa, and total wind divergence (contours) and ageostrophic wind (arrows) at (e) 1000 and (f) 300 hPa in the extended Nordic seas area in winters 1982/83–2005/06 regressed onto the previous summer AWT index. In (a)–(f), the CI is 2.5 gpm, 2.5 gpm, 5 × 10−7 s−1, 5 × 10−7 s−1, 2 × 10−7 s−1, and 0.5 × 10−7 s−1 per unit AWT index, respectively. In (a)–(f), the wind anomalies are in meters per second per unit AWT index (scaled as in the bottom right corner and masked if both components are nonsignificant at the 90% confidence level), respectively. The contour and shading colors and the thick yellow line are explained in the caption to Fig. 2. The thick magenta line indicates the zonal section at which some variables are shown in Fig. 8. In (f), the black box indicates the area for calculating the average vertical velocity in the Barents Sea MIZ (see Figs. 4c and 17).

  • View in gallery

    Anomalies of the meridional wind (positive northward) averaged between 5° and 25°W along a meridional section from 47.5° to 90°N in winters 1982/83–2005/06 regressed onto the previous summer AWT index. The CI is 0.2 m s−1 per unit AWT index. The contour and shading colors are explained in the caption to Fig. 2. The thick contours are the climatological mean winter meridional wind (CI = 2 m s−1; southerlies red and northerlies blue; zero contour omitted).

  • View in gallery

    Anomalies of (a) the geostrophic vorticity, (b) vertical velocity (positive downward), and (c) wind divergence along a vertical section at 80°N (thick magenta line in Fig. 6) in winters 1982/83–2005/06 regressed onto the previous summer AWT index. In (a)–(c), the CI is 5 × 10−7 s−1, 1 × 10−3 Pa s−1, and 1 × 10−7 s−1 per unit AWT index, respectively. The contour and shading colors are explained in the caption to Fig. 2. In (a)–(c), the thick contours are the climatological mean winter zonal winds (m s−1), meridional winds (m s−1) and air temperature (°C), respectively (magenta and black contours are for positive and negative values, respectively, and the zero contour is omitted).

  • View in gallery

    Anomalies of vertical velocity (positive downward) averaged between 5°W and 60°E along a meridional section from 47.5° to 90°N in winters 1982/83–2005/06 regressed onto the previous summer AWT index. The CI is 1 × 10−3 Pa s−1 per unit AWT index. The contour and shading colors are explained in the caption to Fig. 2. The thick contours are the climatological mean winter vertical velocity (CI = 5 × 10−3 Pa s−1; downward red and upward blue; zero contour omitted).

  • View in gallery

    Precipitation anomalies in the extended Nordic seas area in winters 1982/83–2005/06 regressed onto the previous summer AWT index. The CI is 5 × 10−2 mm day−1 per unit AWT index. The contour and shading colors and the thick yellow line are explained in the caption to Fig. 2.

  • View in gallery

    The friction term F′ from Eq. (3) at 1000 hPa in the extended Nordic seas area in winters 1982/83–2005/06 regressed onto the previous summer AWT index. The CI is 2 × 10−11 s−2 per unit AWT index. The contour and shading colors and the thick yellow line are explained in the caption to Fig. 2. The thick magenta line indicates the zonal section at which the terms from Eq. (3) and some related variables are shown in Figs. 12 and 13.

  • View in gallery

    (a) The mean zonal advection of anomalous vorticity and (b) the sum of the total advection and vortex-tube stretching terms from Eq. (3) along a vertical section at 80°N (thick magenta line in Fig. 11) in winters 1982/83–2005/06 regressed onto the previous summer AWT index. The CI is 1 × 10−11 s−2 per unit AWT index. The contour and shading colors are explained in the caption to Fig. 2.

  • View in gallery

    (a) The vorticity anomaly, (b) vorticity tendency anomaly, (c) mean advection of anomalous vorticity, (d) anomalous advection of mean absolute vorticity, (e) vortex-tube stretching anomaly, (f) anomalous eddy vorticity flux convergence, (g) friction term from Eq. (3), and (h) the sum of (c) and (f) along a vertical section at 80°N (thick magenta line in Fig. 11) in early winters (November–February) of the period 1982/83–2005/06 regressed onto the previous summer AWT index. In (a), (b), and (c)–(h), the CI is 5 × 10−7 s−1, 1 × 10−13 s−2, and 1 × 10−11 s−2 per unit AWT index, respectively. The contour and shading colors are explained in the caption to Fig. 2. In (b), the thick black contours are the correlation coefficients (only contours of |r| ≥ 0.4 are plotted).

  • View in gallery

    (a) The mean horizontal advection of anomalous temperature, (b) anomalous horizontal advection of mean temperature, (c) anomalous vertical advection of mean temperature, (d) anomalous eddy heat flux convergence, (e) anomalous diabatic heating, and (f) the sum of (d) and (e) at the 925-hPa level in the extended Nordic seas area in winters 1982/83–2005/06 regressed onto the previous summer AWT index. The CI is 0.2°C day−1 per unit AWT index. The contour and shading colors and the thick yellow line are explained in the caption to Fig. 2. The thick magenta line indicates the zonal section at which the terms from Eq. (9) are shown in Fig. 16. In (e), the black box indicates the area used for calculating averages of the terms from Eq. (9) (see Fig. 15).

  • View in gallery

    Time-lagged regression coefficients of the terms from Eq. (9) at 925 hPa averaged over the black box in Fig. 14e regressed onto the summer AWT index in the 1982–2005 period. The anomalies are in degrees Celsius per day per unit AWT index. Filled symbols denote anomalies statistically significant at the 90% confidence level. Positive lags correspond to the AWT index leading the atmospheric variables calculated as 4-month averages with the interval of 1 month. Lag 6 months corresponds to winter (December–March).

  • View in gallery

    (a) The mean horizontal advection of anomalous temperature, (b) anomalous horizontal advection of mean temperature, (c) anomalous vertical advection of mean temperature, (d) anomalous eddy heat flux convergence, (e) anomalous diabatic heating, and (f) the sum of (d) and (e) along a vertical section at 77.5°N (thick magenta line in Fig. 14) in winters 1982/83–2005/06 regressed onto the previous summer AWT index. The CI is 0.2°C day−1 per unit AWT index. The contour and shading colors are explained in the caption to Fig. 2.

  • View in gallery

    Time-lagged correlation of the anomalies of vertical velocity (positive downward) at 400 hPa in the Barents Sea MIZ (averaged over the black box in Fig. 6f) in winters (December–March) 1982/83–2005/06 with the anomalies of the total sea ice area in the Nordic seas. Filled symbols denote correlations statistically significant at the 90% confidence level. SIA is calculated as 4-month averages of the sea ice concentration integrated over the black box in Fig. 3a with the interval of 1 month. Negative lags correspond to SIA leading the vertical velocity.

  • View in gallery

    Schematic of atmospheric response to warm anomalies of Atlantic water temperature (AWT+). AWT+ leads to negative sea ice cover anomalies (c), warm surface temperature anomalies , upward surface heat flux anomalies , and warm air temperature anomalies (T+) in the marginal ice zone (MIZ) of the Barents and Greenland Seas. A cyclonic wind disturbance (ζ+) develops around the hot spots in the planetary boundary layer (PBL). Strong anomalies of geostrophic flow occur on its southern rim along the Iceland–Barents Sea (IBS) corridor. In this area, warming by wind anomalies is compensated by anomalous diabatic cooling (J) and results in downward surface heat flux anomalies . In the Greenland Sea MIZ, cooling by wind anomalies balances the anomalous diabatic warming (J+) in the PBL. In the Barents Sea MIZ, the sum of J+ and in the PBL is balanced with anomalous cooling by eddies . At the tropopause, J+ plus anomalous warming by eddies are balanced with adiabatic cooling by anomalous ascending motion . The anomalous ascending motion (ω) is forced by upward anomalies of the Ekman pumping resulting from frictional convergence in the cyclonic surface wind disturbance. A quasi-meridional overturning circulation anomaly is closed by southward anomalies of ageostrophic wind () at the tropopause, anomalous descending motion (ω+) south of the IBS corridor, and northward in the PBL. Increases (P+) and decreases (P) in precipitation occur in the area of ascending and descending motion, respectively. Downward anomalies of the Ekman pumping south of the IBS corridor are associated with an anticyclonic wind disturbance (ζ) in the PBL. In the upper troposphere, a broad ζ and a narrow ζ+ are mainly driven by anomalous wind divergence and eddy vorticity flux convergence , respectively.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 36 36 8
PDF Downloads 11 11 2

Local Wintertime Tropospheric Response to Oceanic Heat Anomalies in the Nordic Seas Area

View More View Less
  • 1 Institute of Oceanology, Polish Academy of Sciences, Sopot, Poland
© Get Permissions
Full access

Abstract

A regression analysis between observed summertime Atlantic water temperature anomalies at the entrance to the Barents Sea and atmospheric fields in the following winter from the NCEP–NCAR reanalysis in the period 1982–2006 is carried out. It shows that the ocean plays a key role in shaping wintertime tropospheric variability in the Nordic seas (Greenland–Iceland–Norwegian and Barents Seas) region. The oceanically driven atmospheric circulation anomaly around the Nordic seas marginal ice zone is intensified at the surface as a result of a thermally direct baroclinic adjustment. Frictional convergence in the cyclonic disturbance corresponding to warm ocean temperature anomalies forces ascending motion at the top of the planetary boundary layer and a compensating divergence aloft, which over the Barents Sea is extreme at the tropopause. A quasi-meridional overturning circulation anomaly is closed by descending motion south of the cyclonic disturbance. In addition, an equivalent barotropic flow anomaly appears in the upper troposphere. It is partly driven by eddy–mean flow interactions, as revealed by anomalies in the vorticity budget. The atmospheric response to oceanic forcing in the Nordic seas area is unique because of the prominent role of surface friction and because of specific profiles of diabatic heating and eddy heat flux convergence over the Barents Sea. As revealed by anomalies in the heat budget, the combined effect of diabatic heating and thermal eddy forcing acts, when the diabatic heating is positive, as a heat sink at the surface and a heat source aloft. The strongest anomalous heating occurs in the upper troposphere/lower stratosphere where it counteracts the dynamic cooling owing to anomalous ascending motion.

Corresponding author address: Pawel Schlichtholz, Institute of Oceanology, Polish Academy of Sciences, Powstancow Warszawy 55, 81-712 Sopot, Poland. E-mail: schlicht@iopan.gda.pl

Abstract

A regression analysis between observed summertime Atlantic water temperature anomalies at the entrance to the Barents Sea and atmospheric fields in the following winter from the NCEP–NCAR reanalysis in the period 1982–2006 is carried out. It shows that the ocean plays a key role in shaping wintertime tropospheric variability in the Nordic seas (Greenland–Iceland–Norwegian and Barents Seas) region. The oceanically driven atmospheric circulation anomaly around the Nordic seas marginal ice zone is intensified at the surface as a result of a thermally direct baroclinic adjustment. Frictional convergence in the cyclonic disturbance corresponding to warm ocean temperature anomalies forces ascending motion at the top of the planetary boundary layer and a compensating divergence aloft, which over the Barents Sea is extreme at the tropopause. A quasi-meridional overturning circulation anomaly is closed by descending motion south of the cyclonic disturbance. In addition, an equivalent barotropic flow anomaly appears in the upper troposphere. It is partly driven by eddy–mean flow interactions, as revealed by anomalies in the vorticity budget. The atmospheric response to oceanic forcing in the Nordic seas area is unique because of the prominent role of surface friction and because of specific profiles of diabatic heating and eddy heat flux convergence over the Barents Sea. As revealed by anomalies in the heat budget, the combined effect of diabatic heating and thermal eddy forcing acts, when the diabatic heating is positive, as a heat sink at the surface and a heat source aloft. The strongest anomalous heating occurs in the upper troposphere/lower stratosphere where it counteracts the dynamic cooling owing to anomalous ascending motion.

Corresponding author address: Pawel Schlichtholz, Institute of Oceanology, Polish Academy of Sciences, Powstancow Warszawy 55, 81-712 Sopot, Poland. E-mail: schlicht@iopan.gda.pl

1. Introduction

The warm and salty Atlantic water of subtropical origin flows toward the Arctic Ocean through the Nordic seas, which comprise the Greenland, Iceland, and Norwegian Seas on their western side and the Barents Sea on their eastern side (Fig. 1). On its way northward, the Atlantic water interacts with the cold polar atmosphere. It also interacts with the fresh and cold polar water and sea ice exported from the Arctic Ocean throughout the year and formed locally in winter (e.g., Aagaard et al. 1985). Convergence of oceanic heat transport maintains the climatological mean winter ice edge (border of the shaded area in Fig. 1) exceptionally far north in the Nordic seas relative to other subarctic seas (e.g., Bitz et al. 2005). It also contributes to formation of a wedge of warm air over the open water adjacent to the ice edge (Fig. 1, contours). In winter, this wedge is warmer by more than 20°C compared to the zonal mean temperature at the corresponding latitudes (Seager et al. 2002). The contrast between warm air over the ocean and cold air over the sea ice and land leads to a reverse thermal wind shear in the planetary boundary layer (PBL) that maintains a cyclonic surface geostrophic circulation in the Nordic seas area (Fig. 1, arrows).

Presently, enhanced northward oceanic heat fluxes contribute to the Arctic amplification of the global warming and may further increase in the future. This is indicated, for instance, by the global coupled climate model of the EC-Earth Consortium (EC-Earth) (Koenigk and Brodeau 2014). In particular, increased ocean heat transport to the Barents Sea leads to enhanced bottom ice melt and a northward sea ice retreat. The corresponding changes in the surface heat flux enhance atmospheric warming.

The Barents Sea is also known to be a flywheel of regional climate feedbacks on shorter time scales. In particular, a positive feedback mechanism exists in which anomalous wind-driven inflow of Atlantic water to the Barents Sea sustains the anomalous winds by affecting the surface heat flux in the marginal ice zone (MIZ). This feedback was deemed as responsible for the Arctic warming of the 1930s based on experiments with a climate model and its atmospheric component (Bengtsson et al. 2004). This feedback may also drive quasidecadal oscillations in the Arctic region, as suggested by some conceptual models (e.g., Ikeda 1990; Ådlandsvik and Loeng 1991). Regional coupled sea ice–ocean models show that sea ice variability in the Barents Sea is significantly linked to anomalous volume transport of Atlantic water through the Barents Sea opening (BSO) also on the interannual time scale (e.g., Årthun et al. 2012). On the other hand, a global climate model indicates that seasonal-to-interannual atmospheric variability in the Nordic seas area does significantly depend on the Barents Sea ice cover anomalies, but the latter are not strongly linked to oceanic heat transport to the Barents Sea on these time scales (Koenigk et al. 2009).

A recent study based on oceanic observations and atmospheric reanalysis data suggests that climate feedbacks in the Nordic seas area involve autumn-to-winter reemergence of sea surface temperature (SST) anomalies formed by local air–sea interactions in the preceding winter-to-spring season and then sequestered below the summer mixed layer (Schlichtholz 2013). The reemergence mechanism explains strong links of the Atlantic water temperature (AWT) anomalies observed in the BSO area in summer to both anomalous sea ice extent in the Nordic seas in the following winter (Schlichtholz 2011) and anomalous air–sea interactions over the open water part of the Nordic seas during the previous winter-to-spring season (Schlichtholz and Houssais 2011). Some AWT anomalies may also be advected from the south (e.g., Furevik 2001; Francis and Hunter 2007). In any case, they strongly influence not only the sea ice extent but also the surface air temperature and winds over the Nordic seas (Schlichtholz 2013).

Here we address the problem of how deep is the local atmospheric response to oceanic forcing in the Nordic seas area and what mechanisms govern this response. The study is organized as follows. Data and methods are described in section 2. Then, relations of wintertime anomalies in air temperature, horizontal wind, and vertical motion over the Nordic seas and adjacent areas to AWT anomalies observed half a year earlier in the BSO area are analyzed in section 3. The corresponding anomalous atmospheric vorticity and heat budgets are presented in sections 4 and 5, respectively. A discussion follows in section 6. It includes comparison with studies of atmospheric response to surface forcing in other regions and some remarks on climatic implications of the results. Finally, concluding remarks are given in section 7.

2. Data and methods

Oceanic forcing is characterized by the summer (June–September) AWT index constructed by Schlichtholz and Houssais (2011) for the period 1982–2005. This index is based on temperature data from the International Council for the Exploration of the Sea (ICES) oceanographic database (http://www.ices.dk/) and from the National Oceanic and Atmospheric Administration National Oceanographic Data Center (NODC; http://www.nodc.noaa.gov/). The temperature data were averaged over the Atlantic water core (100–300 m) in the BSO area (70°–76°N, 13°–17°E; box in Fig. 1). Then, they were linearly detrended and divided by the standard deviation (σAWT ≈ 0.4°C) of the anomalies obtained.

Fig. 1.
Fig. 1.

Climatological mean of the winter (December–March) air temperature (°C, contours) and geostrophic wind (m s−1, arrows scaled as in the bottom right corner) at 1000 hPa and sea ice extent (sea ice concentration above 15%, shading) in the Nordic seas area based on data in the period 1982–2006 from the NCEP–NCAR reanalysis. The box delineates the Barents Sea opening (BSO) area.

Citation: Journal of Climate 27, 23; 10.1175/JCLI-D-13-00763.1

Air–sea interactions are described by seasonal (4 month)-mean fields of sea ice concentration c, skin temperature (surface temperature of sea ice or seawater, Ts), turbulent (sensible plus latent) surface heat flux Qt (positive upward), and zonal and meridional components of the surface wind stress τs. These fields are constructed using monthly mean data from the National Centers for Environmental prediction (NCEP)–National Center for Atmospheric Research (NCAR) reanalysis on the spectral T62 (~2° latitude × ~2° longitude) grid (Kalnay et al. 1996), downloaded from http://www.esrl.noaa.gov/psd/. Seasonal mean atmospheric fields are constructed using monthly or daily mean data from the NCEP–NCAR reanalysis at constant pressure (p) levels on a 2.5° latitude × 2.5° longitude grid and monthly mean precipitation P on the T62 grid.

The monthly atmospheric dataset at constant pressure levels includes the temperature T, geopotential height Z, pressure vertical velocity ω (positive downward), and zonal (positive eastward) and meridional (positive northward) components of the wind velocity u = (u, υ). The geopotential height and surface wind stress data are used to calculate the geostrophic wind ug and the Ekman pumping velocity ωE in units of the pressure velocity (positive downward) at the top of the PBL, respectively,
e1
where g, f, k, and are acceleration due to gravity, the Coriolis parameter, the vertical unit vector, and the horizontal gradient operator, respectively. The horizontal wind velocity and the geostrophic wind data are used to calculate the relative vorticity, ζ = k · × u, and the geostrophic relative vorticity, ζg = k · × ug, respectively.
The daily atmospheric dataset includes the temperature (corrected for the lapse rate to obtain the potential temperature θ) and the horizontal wind velocity components. It is used to calculate the horizontal convergence of eddy vorticity and heat fluxes, Eζ and Eθ respectively, from the following formulas:
e2
where the double overbar and double prime indicate seasonal averaging and daily anomaly with respect to the seasonal average, respectively.

All seasonal-mean surface and atmospheric variables are decomposed into the seasonally varying long-term mean (i.e., a climatological annual cycle in the period 1982–2006) denoted by an overbar and the anomaly (denoted by a prime) from this mean (e.g., ). The anomalies are then linearly detrended and regressed onto the AWT index. The statistical significance of the correlations r is assessed using a two-tailed t test carried out with an effective number of degrees of freedom (Bretherton et al. 1999). Horizontal and vertical distributions of several regression coefficients (referred to as AWT-associated anomalies) are shown in the figures (e.g., Fig. 2) in which positive (negative) anomalies of the scalar fields are plotted as red (blue) contours, and those anomalies that are statistically significant at the 90% confidence level are marked through pink and aquamarine shading, respectively. Anomalies of the vector fields are plotted only at locations where either the zonal or meridional components are significant at the 90% confidence level. As a linear method is applied, the plotted distributions correspond to warm (positive) AWT anomalies equal to σAWT and the same distributions but with opposite sign correspond to cold (negative) AWT anomalies equal to −σAWT.

Fig. 2.
Fig. 2.

Anomalies of the air temperature (contours) and geostrophic wind (arrows) at (a) 1000, (b) 850, and (c) 500 hPa in the extended Nordic seas area in winters 1982/83–2005/06 regressed onto the previous summer AWT index. The contour interval (CI) for the temperature anomalies is 0.2°C per unit AWT index. Red and blue contours represent positive and negative anomalies, respectively. The zero contour is omitted, and pink and aquamarine shading denote positive and negative anomalies statistically significant at the 90% confidence level, respectively. The wind anomalies are in meters per second per unit AWT index (scaled as in the bottom right corner and masked if both components are nonsignificant at the 90% confidence level). The thick yellow line shows the climatological mean position of the winter ice edge (15% sea ice concentration contour). In (a) and (c), the magenta box indicates the area used for calculating the average zonal wind at the entrance to the Barents Sea and average air temperature over the Barents Sea, respectively (see Figs. 4a,b).

Citation: Journal of Climate 27, 23; 10.1175/JCLI-D-13-00763.1

3. Covariations between the summer AWT index and winter atmospheric anomalies

a. Air temperature anomalies

It has been previously shown that more than 80% of the variance in the regional average of the surface (2-m height) air temperature over the Nordic seas (within the geographical limits of Fig. 1) in winter (December–March) is linked to the concurrent anomalies of the sea ice extent in this area (Schlichtholz 2013). More than 70% of the variance of the winter sea ice extent anomalies is, in turn, accounted for by the previous summer AWT anomalies in the BSO area (Schlichtholz 2011). Consequently, 60% of the variance of the wintertime air temperature anomalies over the Nordic seas is driven by the ocean in the marginal ice zone (Schlichtholz 2013). Consistent with these findings, the map of winter surface air temperature regressed on AWT has significant local maxima in the MIZ (Fig. 2a) in both the Barents Sea (~2.5°C) and Greenland Sea (~1.5°C). These “hot spots” are associated with minima in the corresponding map of sea ice concentration (Fig. 3a).

Fig. 3.
Fig. 3.

Anomalies of (a) the sea ice concentration, (b) skin temperature, (c) turbulent surface heat flux (positive upward), and (d) surface wind stress (arrows) and Ekman pumping velocity (positive downward) at the top of the atmospheric planetary boundary layer (contours) in the extended Nordic seas area in winters 1982/83–2005/06 regressed onto the previous summer AWT index. In (a)–(d), the CI is 5%, 0.5°C, 10 W m−2, and 2 × 10−3 Pa s−1 per unit AWT index, respectively. In (d), the wind stress anomalies are in newtons per square meter per unit AWT index (scaled as in the bottom-right corner and masked if both components are nonsignificant at the 90% confidence level). The contour and shading colors and the thick yellow line are explained in the caption to Fig. 2. In (a) and (c), the black box indicates the area over which the sea ice concentration and surface heat flux anomalies are integrated (see Figs. 17 and 5, respectively).

Citation: Journal of Climate 27, 23; 10.1175/JCLI-D-13-00763.1

The atmospheric hot spots (or cold spots in the negative phase of the AWT index) appear mainly in response to anomalous sensible surface heat flux maintained by large surface temperature anomalies. The latter should result from oceanically driven anomalies of sea ice cover as the ice surface temperature in the cold season is several degrees lower than the freezing point. Indeed, on both sides of the Nordic seas, the AWT-associated winter skin temperature anomalies are much larger in the MIZ (~5°C in the Barents Sea MIZ and ~3°C in the Greenland Sea MIZ) than the corresponding surface temperature anomalies in the open water (Fig. 3b). They are also larger than the air temperature anomalies in the MIZ, which they force (note a different contour interval in Figs. 3b and 2a). As sea ice influences also the moisture flux from the surface, the total AWT-associated turbulent surface heat flux anomalies in the MIZ have a substantial contribution (~35%) from the latent heat. The upward surface heat flux anomalies in the MIZ corresponding to warm oceanic heat anomalies and the subsequent anomalous sea ice retreat coexist with two distinct lobes of downward surface heat flux anomalies: one north of Iceland and one in the northeastern Norwegian Sea/southern Barents Sea (Fig. 3c). Both occur directly southeast and south, respectively, of the climatological ice edge.

Consistent with the oceanic forcing scenario, the AWT-associated winter anomalies of air temperature decrease upward on both sides of the Nordic seas. Over the Greenland Sea, they are strongly attenuated already within the PBL. At 850 hPa, significant temperature anomalies in this area appear only on the open water side of the ice edge (Fig. 2b). At 500 hPa, significant temperature anomalies (~0.5°C) are found only over the Barents Sea (Fig. 2c).

Figure 4a displays the postsummer evolution (from lag 0 to lag 12 months) of the vertical structure of the AWT-associated anomalies of air temperature in the Barents Sea area (averaged over the magenta box in Fig. 2c). It shows that expansion of significant temperature anomalies far beyond the PBL occurs only in the developed stage of the atmospheric response, that is, in early winter (lag 5 months) and winter (lag 6 months) when the AWT-associated air temperature anomalies are extreme. During this stage, significant temperature anomalies extend nearly to the tropopause (~300 hPa). Reasons for such a deep response will be discussed in section 5b.

Fig. 4.
Fig. 4.

Vertical section of the time-lagged regression coefficients of the anomalies of (a) air temperature averaged over the magenta box in Fig. 2c, (b) westerly geostrophic velocity averaged over the magenta box in Fig. 2a, and (c) vertical velocity (positive downward) averaged over the black box in Fig. 6f regressed onto the summer AWT index. In (a)–(c), the CI is 0.2°C, 0.2 m s−1, and 1 × 10−3 Pa s−1 per unit AWT index, respectively. The contour and shading colors are explained in the caption to Fig. 2. The thick black contours are the correlation coefficients (only contours of |r| ≥ 0.4 are plotted). Positive lags correspond to the AWT index leading the atmospheric variables calculated as 4-month averages with the interval of 1 month. Lag 6 months corresponds to winter (December–March).

Citation: Journal of Climate 27, 23; 10.1175/JCLI-D-13-00763.1

In the PBL, the AWT-associated air temperature anomalies remain significant from late autumn to early spring (i.e., for lags 4–8 months) (Fig. 4a). They are maintained by the large surface heat flux anomalies in the MIZ that occur during this season. This is illustrated in Fig. 5, which shows the postsummer evolution of the surface heat flux anomalies integrated over the upward (circles) and downward (triangles) lobes of their AWT-associated pattern in the Nordic seas area (black box in Fig. 3c). The upward anomalies in the MIZ increase from ~10 to ~20 TW between autumn and late autumn, and reach a maximum of ~25 TW in late winter (lag 7 months). The downward anomalies in the open water are as large as the upward anomalies in the MIZ only in early winter/winter. As discussed later (see sections 3b and 5a), they should mainly result from a feedback from oceanically driven wind anomalies.

Fig. 5.
Fig. 5.

Magnitude of the time-lagged regression coefficients of the anomalies of the turbulent surface heat flux regressed onto the summer AWT index and then integrated over the positive (circles) and negative (triangles) lobes of the regression pattern over the Nordic seas area (black box in Fig. 3c). The anomalies are in 1012 W per unit AWT index. Positive lags correspond to the AWT index leading the surface heat flux anomalies calculated as 4-month averages with the interval of 1 month. Lag 6 months corresponds to winter (December–March).

Citation: Journal of Climate 27, 23; 10.1175/JCLI-D-13-00763.1

The observations in the Nordic seas area are consistent with the following sequence, which we illustrate with a warm AWT anomaly. The anomalously warm Atlantic inflow inhibits ice formation, which allows greater heat flux from ocean to atmosphere, which warms the atmosphere. The ice and atmospheric anomalies may persist in winter because of a feedback between oceanically driven wind anomalies and wind-driven AWT anomalies (Schlichtholz 2013). Indeed, significant AWT-associated winter anomalies of the surface wind stress that appear in the Nordic seas area (Fig. 3d) should drive anomalous geostrophic currents through the Ekman suction at the base of the ocean surface frictional layer and/or a coastal water pile-up mechanism.

b. Surface wind anomalies

The winter atmospheric response to oceanic heat anomalies in the Nordic seas is characterized by a strong anomaly of the surface geostrophic airflow on the open water side of the ice edge, in the area extending approximately from south of Denmark Strait to the ice edge in the southeastern Barents Sea (Fig. 2a). Hereafter, this area will be referred to as the Iceland–Barents Sea (IBS) corridor. The anomalous geostrophic winds reach a magnitude of ~1.5 m s−1 per unit AWT index in this corridor. In the positive phase of the AWT index, they blow eastward on the southern rim of the Greenland Sea and Barents Sea hot spots (Fig. 2a). Then, they turn cyclonically northward over the Kara Sea. The cyclonic disturbance is antiparallel to the corresponding anomaly of the thermal wind shear, which is anticyclonic around the hot spots (, where R ≡ gas constant of air). In other words, the oceanically driven surface geostrophic wind anomalies result from a baroclinic adjustment of the atmospheric circulation to anomalous heating from below in the MIZ. This heating builds up anomalous horizontal temperature gradients that through the geostrophic and hydrostatic constraints induce a vertically sheared flow around the MIZ. Since the air temperature anomalies decrease upward (Figs. 2 and 4a), the anomalous geostrophic winds are intensified at the surface. Therefore, forcing by warm oceanic anomalies moves the surface air temperature front associated with the ice edge northward. This consequently leads to a northward shift of the generally cyclonic surface circulation in the Nordic seas area (Fig. 1).

The surface intensification of the anomalous airflow on the rim of the MIZ hot spots is illustrated in Fig. 4b. It displays the postsummer evolution of the vertical structure of the AWT-associated anomalies of the zonal geostrophic flow at the entrance to the Barents Sea (averaged over the magenta box in Fig. 2a). These anomalies are extreme in winter (lag 6 months) when they clearly weaken with height from a surface maximum. Significant anomalous winds start in early winter (lag 5 months). This explains a sudden jump in the magnitude of the integrated downward AWT-associated surface heat flux anomalies in the open water from ~10 TW in late autumn to nearly 25 TW in early winter (Fig. 5, triangles). These anomalies should mainly result from advection of the mean temperature (Fig. 1) by the anomalous surface winds along the IBS corridor (see section 5a for more details).

The anomalous surface winds in the IBS corridor blow along a common rim of an Arctic trough in the geopotential height anomalies centered at the northern tip of Spitsbergen and a subarctic ridge centered at the southern tip of Scandinavia (Fig. 6a). The height anomalies are not significant in the Arctic trough. However, the corresponding positive (cyclonic) vorticity anomalies in the MIZ are significant on both sides of the Nordic seas (Fig. 6c). The negative (anticyclonic) vorticity anomalies south of the IBS corridor are significant as well. This reflects the fact that the low-level wind anomalies are directly driven by the thermal surface forcing in the MIZ while the low-level pressure anomalies are also influenced by barotropic forcing in the upper troposphere. This forcing may partly be steered by the oceanic anomalies themselves (see section 4c), but should also result from intrinsic atmospheric variability. Confounded baroclinic and barotropic influences on recent sea level pressure anomalies in the Arctic region were also reported by Overland and Wang (2010).

Fig. 6.
Fig. 6.

Anomalies of the geopotential height (contours) and geostrophic wind (arrows) at (a) 1000 and (b) 300 hPa, total wind (arrows) and its curl (contours) at (c) 1000 and (d) 300 hPa, and total wind divergence (contours) and ageostrophic wind (arrows) at (e) 1000 and (f) 300 hPa in the extended Nordic seas area in winters 1982/83–2005/06 regressed onto the previous summer AWT index. In (a)–(f), the CI is 2.5 gpm, 2.5 gpm, 5 × 10−7 s−1, 5 × 10−7 s−1, 2 × 10−7 s−1, and 0.5 × 10−7 s−1 per unit AWT index, respectively. In (a)–(f), the wind anomalies are in meters per second per unit AWT index (scaled as in the bottom right corner and masked if both components are nonsignificant at the 90% confidence level), respectively. The contour and shading colors and the thick yellow line are explained in the caption to Fig. 2. The thick magenta line indicates the zonal section at which some variables are shown in Fig. 8. In (f), the black box indicates the area for calculating the average vertical velocity in the Barents Sea MIZ (see Figs. 4c and 17).

Citation: Journal of Climate 27, 23; 10.1175/JCLI-D-13-00763.1

c. Upper-tropospheric wind anomalies

Not only the magnitude but also the pattern of oceanically driven wind anomalies change with height. While the northward turning of the significant flow at 1000 hPa on the exit from the IBS corridor occurs on the eastern rim of the Barents Sea hot spot (Fig. 2a), at 850 hPa it occurs across the center of this spot (Fig. 2b). At 500 hPa, the anomalous flow in the IBS corridor is significant only west of the Barents Sea (Fig. 2c). At this level, significant anomalous westerlies (easterlies) appear northeast (southeast) of the Barents Sea hot spot. These anomalous winds are a manifestation of an anticyclonic disturbance in the upper-tropospheric circulation. This disturbance is intensified at the tropopause where the anomalous geostrophic winds blow around an elongated ridge in the geopotential height anomalies centered over northern Scandinavia (Fig. 6b).

The anomalous upper-tropospheric winds reach the same magnitude (~1.5 m s−1 per unit AWT index) as the anomalous surface winds in the IBS corridor. The latter blow under the upper-tropospheric ridge (Figs. 6a,b). This, on the one hand, reflects the baroclinic nature of the atmospheric response in the lower troposphere to oceanic forcing, and, on the other hand, shows that this forcing affects also the upper-tropospheric circulation by generating equivalent barotropic wind anomalies (i.e., wind anomalies having a magnitude that increases upward). This increase is illustrated in Fig. 7, which shows a latitudinal section of the AWT-associated winter meridional wind anomalies west of the Greenwich meridian (averaged between 5° and 25°W) on the background of the corresponding climatological mean winds (thick contours). The latter are characterized by a southward flow at the surface and a northward flow at the tropopause. Apparently, forcing by warm oceanic anomalies results in a northward shift of this antiparallel flow and moves the upper-level core of the southerly winds upward.

Fig. 7.
Fig. 7.

Anomalies of the meridional wind (positive northward) averaged between 5° and 25°W along a meridional section from 47.5° to 90°N in winters 1982/83–2005/06 regressed onto the previous summer AWT index. The CI is 0.2 m s−1 per unit AWT index. The contour and shading colors are explained in the caption to Fig. 2. The thick contours are the climatological mean winter meridional wind (CI = 2 m s−1; southerlies red and northerlies blue; zero contour omitted).

Citation: Journal of Climate 27, 23; 10.1175/JCLI-D-13-00763.1

The vertical structure of anomalous airflow is further illustrated in Fig. 8a, which displays the AWT-associated winter anomalies of the geostrophic vorticity along a zonal section at 80°N. The section runs through the center of the low-level trough near Spitsbergen (Fig. 6a). It also runs through a local maximum in the upper-tropospheric band of cyclonic vorticity anomalies squeezed along Greenland (Fig. 6d). Over the Barents Sea, the extreme in the vertical profile of is attached to the surface and increases westward to an absolute maximum near Spitsbergen (Fig. 8a). Then, after detaching from the surface in Fram Strait, the extreme tilts westward with height up to the tropopause. It loses its significance at the Greenwich meridian under a stratospheric core in the mean westerlies, shown as thick contours in Fig. 8a. This core results from meandering of the mean polar vortex (see the thick contours in Fig. 8b for across the 80°N section). The maximum in the zonal profile of extends from the stratosphere to the troposphere where it tilts eastward with pressure along the axis of the anomalous vortex (Fig. 8a). This indicates that the atmospheric response to oceanic forcing in the Nordic seas area may be influenced by the mean circulation. The mean winds do indeed affect the AWT-associated vorticity budget (see section 4b).

Fig. 8.
Fig. 8.

Anomalies of (a) the geostrophic vorticity, (b) vertical velocity (positive downward), and (c) wind divergence along a vertical section at 80°N (thick magenta line in Fig. 6) in winters 1982/83–2005/06 regressed onto the previous summer AWT index. In (a)–(c), the CI is 5 × 10−7 s−1, 1 × 10−3 Pa s−1, and 1 × 10−7 s−1 per unit AWT index, respectively. The contour and shading colors are explained in the caption to Fig. 2. In (a)–(c), the thick contours are the climatological mean winter zonal winds (m s−1), meridional winds (m s−1) and air temperature (°C), respectively (magenta and black contours are for positive and negative values, respectively, and the zero contour is omitted).

Citation: Journal of Climate 27, 23; 10.1175/JCLI-D-13-00763.1

d. Anomalous vertical motion

The section at 80°N runs through an upper-tropospheric center of anomalous wind divergence ( · u′ > 0) that develops over the Barents Sea hot spot (Fig. 6f). This divergence is connected by anomalous ascending motion (ω′ < 0) in the middle troposphere (Fig. 8b) to a strong anomalous wind convergence ( · u′ < 0) at the surface (Fig. 8c). The upper-tropospheric divergence reaches a maximum magnitude at the tropopause (see the black contours in Fig. 8c for the vertical structure of along the 80°N section). A less significant upper-tropospheric wind divergence coexisting with a significant surface wind convergence appears in the Greenland/Iceland Sea marginal ice zone (Figs. 6f,e).

The anomalous vertical motion in the Barents Sea MIZ follows the lag structure of the surface wind anomalies; that is, it is significant from early winter to early spring (lags 5–8 months) and extreme in winter. This is shown in Fig. 4c for ω′ averaged over the black box in Fig. 6f. The winter extreme is highly significant and linked to the AWT index as strongly (r < −0.7) as the corresponding surface westerlies in the BSO area (r > 0.7). This is illustrated by superimposing the correlations (thick black contours) on the anomalies in Fig. 4. The correlations exceeding 0.45–0.55 and 0.55–0.65 are significant at the 95% and 99% confidence level, respectively. The correlation for the vertical velocity in the Barents Sea MIZ is extreme at 400 hPa (r = −0.73) and for the corresponding wind divergence (averaged over the black box in Fig. 6f) at 300 hPa (r = 0.76).

The anomalous surface wind convergences in the MIZ are accompanied by significant anomalous surface wind divergences south of the IBS corridor (Fig. 6e). These divergences coexist with upper-tropospheric convergences (Fig. 6f). This implies that an anomalous quasimeridional overturning cell appears in response to oceanic forcing in the MIZ that joins the ascending motion in the north to a descending motion in the south. This cell is illustrated in Fig. 9, which shows a latitudinal section of the AWT-associated winter anomalies of the vertical velocity averaged between 5°W and 60°E. The corresponding climatological mean vertical velocity is shown in the same figure (thick contours). Evidently, forcing by warm oceanic anomalies moves the center of the generally ascending motion in the Nordic seas area to the north and makes the vertical circulation deeper. This shift is accompanied by enhanced precipitation in the MIZ and decreased precipitation south of the IBS corridor (Fig. 10).

Fig. 9.
Fig. 9.

Anomalies of vertical velocity (positive downward) averaged between 5°W and 60°E along a meridional section from 47.5° to 90°N in winters 1982/83–2005/06 regressed onto the previous summer AWT index. The CI is 1 × 10−3 Pa s−1 per unit AWT index. The contour and shading colors are explained in the caption to Fig. 2. The thick contours are the climatological mean winter vertical velocity (CI = 5 × 10−3 Pa s−1; downward red and upward blue; zero contour omitted).

Citation: Journal of Climate 27, 23; 10.1175/JCLI-D-13-00763.1

Fig. 10.
Fig. 10.

Precipitation anomalies in the extended Nordic seas area in winters 1982/83–2005/06 regressed onto the previous summer AWT index. The CI is 5 × 10−2 mm day−1 per unit AWT index. The contour and shading colors and the thick yellow line are explained in the caption to Fig. 2.

Citation: Journal of Climate 27, 23; 10.1175/JCLI-D-13-00763.1

The anomalous circulation cell in the vertical–meridional plane is consistent with significant anomalous ageostrophic winds , which at the tropopause emanate from the northern divergences toward the southern convergences (Fig. 6f). However, the full three-dimensional atmospheric response is more complex, as indicated by anomalous ageostrophic winds that at the surface feed the Barents Sea convergence from all directions (Fig. 6e). This complexity and that the anomalous circulation is mainly steered by air–sea interactions in the MIZ are reflected by less significant anomalies of the upper-tropospheric vertical motion south of the IBS corridor than in the Barents Sea MIZ. For instance, local extremes of the correlation with the AWT index for ω′ at 300 hPa in these areas are 0.55 (at 57.5°N, 2.5°E) and −0.74 (at 80°N, 37.5°E), respectively.

The significant anomalies of surface ageostrophic winds suggest that surface friction plays an important role for the atmospheric response to oceanic forcing in the Nordic seas area. The same conclusion was drawn by Schlichtholz (2013) from an analysis of the AWT-associated anomalies of Ekman pumping at the top of the PBL (shown in units of the pressure velocity in Fig. 3d). The prominent role of surface friction will be further analyzed in section 4a.

4. Anomalous vorticity balance

The results shown in the previous section indicate that 1) anomalous heating from below in the Nordic seas marginal ice zone generates hot spots that induce a cyclonic surface geostrophic circulation anomaly around them, 2) the cyclonic wind anomaly extends up to the tropopause while tilting with height from the Barents Sea westward, and 3) this anomaly coexists with an anomalous wind convergence in the PBL and a compensating divergence aloft. To examine dynamic linkages between all these anomalies, we consider the linearized vorticity budget for extratropical atmospheric disturbances on the seasonal time scale, which may be written as follows:
e3
where ∂/∂t is the time derivative, is the anomaly of the horizontal eddy vorticity flux convergence defined in Eq. (2), and F′ is the curl of the anomalous frictional force. Note that F′ is estimated as the residual of the other terms, and therefore includes also computational errors and omitted terms (assumed negligible based on scale analysis). The omitted terms may be derived, for instance, from Eq. (10) in Holopainen (1978). In addition to the nonlinear contribution u′ · ζ′ to the horizontal advection term, the omitted terms include contributions to the absolute vortex-tube stretching term from and , and anomalies of the vortex-tube tipping, the vertical advection of relative vorticity, and the curl of the vertical eddy momentum flux convergence. The anomalous vorticity tendency (∂ζ′/∂t) is negligible but is retained on the lhs of Eq. (3) to emphasize the fact that the terms on the right-hand side represent competitive sources and sinks of the anomalous vorticity. Vorticity tendency anomalies will be shown for the early winter (November–February) season, for which ∂ζ/∂t is calculated by differentiation of the winter (December–March) and late autumn (October–January) vorticity.

a. Role of surface friction

To check the scenario of a frictionally modulated atmospheric response to oceanic forcing proposed by Schlichtholz (2013), Fig. 11 displays the AWT-associated winter pattern of the friction term F′ from Eq. (3) at 1000 hPa in the Nordic seas and adjacent areas. This pattern mirrors the corresponding pattern of wind divergence (Fig. 6e). It exhibits a frictional vorticity sink (F′ < 0) in the area of the vorticity source from the vortex-tube stretching by convergent winds (fω′/∂p = −f · u′ > 0) in the MIZ. It also shows a frictional vorticity source (F′ > 0) in the area of the vorticity sink from the vortex-tube squashing by divergent winds (fω′/∂p < 0) south of the IBS corridor. Not only the patterns but also the magnitudes of the divergence and friction effects are similar. In the center of the Barents Sea convergence, for instance, · u′ reaches −1.1 × 10−6 s−1 per unit AWT index (Fig. 6e). The corresponding estimate of −f · u′ is 1.5 × 10−10 s−2 for f = 1.4 × 10−4 s−1. The magnitude of F′ is exactly the same in this area (Fig. 11). Evidently, the primary vorticity balance in the PBL is between anomalies of the vortex-tube stretching and the curl of the frictional force. The equation for this balance can be written as
e4
where F′ has been reduced to a contribution resulting from the vertical shear of the horizontal stress τ.
Fig. 11.
Fig. 11.

The friction term F′ from Eq. (3) at 1000 hPa in the extended Nordic seas area in winters 1982/83–2005/06 regressed onto the previous summer AWT index. The CI is 2 × 10−11 s−2 per unit AWT index. The contour and shading colors and the thick yellow line are explained in the caption to Fig. 2. The thick magenta line indicates the zonal section at which the terms from Eq. (3) and some related variables are shown in Figs. 12 and 13.

Citation: Journal of Climate 27, 23; 10.1175/JCLI-D-13-00763.1

To link the horizontal and vertical velocity anomalies through the vorticity balance, we first integrate Eq. (4) through the boundary layer with the condition of zero stress at the top of the layer and zero vertical velocity at the surface. The latter condition is justified only for flat surfaces such as oceans and continental plains. It ignores the ascent or descent of air on the slopes of mountains (e.g., Pedlosky 1987). Then, we express the surface stress anomaly through the anomaly of geostrophic velocity near the surface, , using a simple drag law:
e5
where ρa is a constant air density, is the mean thickness of the PBL, and is a friction coefficient (assumed constant). As a result, we obtain a relation between the geostrophic vorticity near the surface, , and the Ekman pumping at the top of the PBL:
e6
Equation (6) yields a reasonable estimate of the PBL thickness. For instance, in the Barents Sea vortex area where s−1 (Fig. 8a) and Pa s−1 (Fig. 3d), should be about 500 m. This estimate is consistent with the squeezing of large convergence anomalies below the 925-hPa level in this area (Fig. 8c).

Equation (6) explains why the cyclonic surface wind perturbation (ζ′ > 0) in the MIZ (Fig. 6c) coexists with anomalous surface wind convergence and pumping of air out of the PBL to the free atmosphere in this area (Figs. 6e and 3d). It also explains why the anticyclonic surface wind perturbation (ζ′ < 0) south of the IBS corridor coexists with anomalous surface wind divergence and suction of air from the free atmosphere to the boundary layer in this area. While the anomalous Ekman pumping in the MIZ appears directly in response to the thermally driven horizontal circulation anomaly, the anomalous Ekman suction south of the IBS corridor closes the quasi-meridional overturning cell described in section 3d. The scenario in which this cell is driven in the MIZ by the frictional convergence in the PBL is supported by similar magnitudes of and the midtropospheric vertical velocity anomaly. In the Barents Sea hot spot, ω′ reaches about −7 × 10−3 Pa s−1 per unit AWT index (Fig. 8b). This anomaly should influence the upper-tropospheric vorticity budget, which will be discussed below.

b. Role of mean winds

The vortex-tube squashing by the anomalous upper-level wind divergence over the Barents Sea hot spot (Fig. 8c) is a vorticity sink that makes the upper-tropospheric extension of the anomalous surface cyclonic vortex tilt with height (Fig. 8a). At the tropopause, where the anomalous divergence reaches 2.5 × 10−7 s−1, −f · u′ is equal to −3.5 × 10−11 s−2 per unit AWT index. This source of anticyclonic vorticity competes with the mean advection of anomalous vorticity, mainly by the westerlies (, where x is the zonal coordinate). At the tropopause, the latter has the same magnitude (3.5 × 10−11 s−2) as the vortex-tube squashing (Fig. 12a). Therefore, the approximate vorticity balance in the anomalous upper-tropospheric divergence over the northern Barents Sea is
e7
Fig. 12.
Fig. 12.

(a) The mean zonal advection of anomalous vorticity and (b) the sum of the total advection and vortex-tube stretching terms from Eq. (3) along a vertical section at 80°N (thick magenta line in Fig. 11) in winters 1982/83–2005/06 regressed onto the previous summer AWT index. The CI is 1 × 10−11 s−2 per unit AWT index. The contour and shading colors are explained in the caption to Fig. 2.

Citation: Journal of Climate 27, 23; 10.1175/JCLI-D-13-00763.1

Deviations from the balance in Eq. (7) occur on the eastern side of the upper-level divergence where the advection effect becomes weak. Moreover, the advection effect is relatively large in its anticyclonic lobe over the Greenland coast where significant anomalies extend from the tropopause down to the surface (Fig. 12a). They are not compensated by the vortex-tube stretching since the wind divergence is not significant in this area (Fig. 8c). This indicates that oceanic forcing leads to an anomalous upper-tropospheric vorticity balance that is generally more complex than suggested by Eq. (7). The anomalous advection of mean absolute vorticity, , contributes to this balance. However, its magnitude is generally smaller than the magnitude of the largest terms. Figure 12b shows the AWT-associated imbalance between the total advection anomaly and the anomalous vortex-tube stretching across the Barents Sea hot spot at 80°N in winter. This imbalance represents the combined effect of friction and eddies but with opposite sign, that is, . Evidently, the atmospheric response to oceanic heat anomalies in the Nordic seas area is affected not only by friction in the PBL but also by dissipation and/or eddy forcing in the upper troposphere. Below we will show that the eddy forcing is not negligible.

c. Role of eddy forcing

In early winter, the AWT-associated cyclonic vorticity anomaly that tilts with height from the Barents Sea MIZ toward Greenland is not yet significant above 500 hPa (Fig. 13a). Its eastern limit does not extend as far east as in winter (Fig. 8a). These features are consistent with a positive AWT-associated vorticity tendency anomaly through early winter (Fig. 13b). This anomaly is smaller by two orders of magnitude than the other terms of the AWT-associated early-winter vorticity budget (Figs. 13c–g). However, it is significant in a cyclonic, westward tilting lobe that is collocated at the surface with the Barents Sea hot spot and reaches the tropopause in Fram Strait. In this lobe, the correlations of ∂ζ′/∂t with the AWT index (thick black contours in Fig. 13b) exceed 0.7 in the PBL and 0.5 throughout the troposphere.

Fig. 13.
Fig. 13.

(a) The vorticity anomaly, (b) vorticity tendency anomaly, (c) mean advection of anomalous vorticity, (d) anomalous advection of mean absolute vorticity, (e) vortex-tube stretching anomaly, (f) anomalous eddy vorticity flux convergence, (g) friction term from Eq. (3), and (h) the sum of (c) and (f) along a vertical section at 80°N (thick magenta line in Fig. 11) in early winters (November–February) of the period 1982/83–2005/06 regressed onto the previous summer AWT index. In (a), (b), and (c)–(h), the CI is 5 × 10−7 s−1, 1 × 10−13 s−2, and 1 × 10−11 s−2 per unit AWT index, respectively. The contour and shading colors are explained in the caption to Fig. 2. In (b), the thick black contours are the correlation coefficients (only contours of |r| ≥ 0.4 are plotted).

Citation: Journal of Climate 27, 23; 10.1175/JCLI-D-13-00763.1

The anomalous early-winter vorticity budget confirms that the cyclonic vorticity anomaly in the PBL over the Barents Sea is driven by anomalous, frictionally driven vortex-tube stretching (Figs. 13e,g). It also shows that the upper-tropospheric extension of this anomaly is driven by anomalous convergence of the eddy vorticity flux (Fig. 13f). This convergence, when combined with the mean advection of anomalous vorticity, exhibits a strong upper-tropospheric core at the Greenwich meridian (Fig. 13h). This combined vorticity source is counteracted by anomalous vorticity dissipation (Fig. 13g); that is,
e8
This shows that the dynamic atmospheric response to oceanic forcing in the Nordic seas area has not only a thermally direct baroclinic component but also an indirect component steered by eddies.

5. Anomalous heat balance

We have shown that oceanically driven vertical motion and eddy forcing play a key role in the anomalous vorticity budget over the Nordic seas. To find whether the anomalous vertical motion and eddy forcing play also a significant role in the corresponding heat budget, we use the following linearized thermodynamic equation:
e9
where is the anomaly of the horizontal eddy heat flux convergence defined in Eq. (2) and J′ is the anomalous diabatic heating, estimated as the residual of the other terms. Following similar studies (e.g., White and Chen 2002), the mean vertical advection of anomalous temperature and the anomaly of the vertical eddy heat flux convergence are assumed negligible and are omitted in Eq. (9).

a. Role of low-level eddy forcing

The AWT-associated terms from Eq. (9) in the lower troposphere (at 925 hPa) over the Nordic seas and adjacent areas in winter are shown in Figs. 14a–e. Consistent with the corresponding surface heat flux anomalies (Fig. 3c), the diabatic heating anomalies are positive in the Barents Sea and Greenland Sea MIZ (Fig. 14e). The role of feedback from the anomalous circulation on the thermal budget, however, is different in these areas. In the Greenland Sea MIZ, the cold advection by northerly and westerly wind anomalies (Fig. 14b) is a primary agent compensating the diabatic warming; that is,
e10
The same balance holds over the open water in the IBS corridor. However, in this area a diabatic cooling appears in response to the warm advection by southwesterly to westerly wind anomalies and results in downward surface heat flux anomalies (Fig. 3c).
Fig. 14.
Fig. 14.

(a) The mean horizontal advection of anomalous temperature, (b) anomalous horizontal advection of mean temperature, (c) anomalous vertical advection of mean temperature, (d) anomalous eddy heat flux convergence, (e) anomalous diabatic heating, and (f) the sum of (d) and (e) at the 925-hPa level in the extended Nordic seas area in winters 1982/83–2005/06 regressed onto the previous summer AWT index. The CI is 0.2°C day−1 per unit AWT index. The contour and shading colors and the thick yellow line are explained in the caption to Fig. 2. The thick magenta line indicates the zonal section at which the terms from Eq. (9) are shown in Fig. 16. In (e), the black box indicates the area used for calculating averages of the terms from Eq. (9) (see Fig. 15).

Citation: Journal of Climate 27, 23; 10.1175/JCLI-D-13-00763.1

The lobe of warm advection by the wind disturbance continues through the Barents Sea MIZ where it is as strong a heat source as the anomalous diabatic heating (Figs. 14b,e). The main heat sink competing with these heat sources is from a divergent eddy heat flux (Fig. 14d). The mean horizontal advection of anomalous temperature is negligible in this area (Fig. 14a), but the cold advection by anomalous ascending motion is not (Fig. 14c). However, the latter is significant west of the significant diabatic heating anomaly where it is mainly balanced with the anomalous horizontal advection of mean temperature. Therefore, the primary anomalous low-level heat balance in the area of large surface heat flux anomalies in the Barents Sea MIZ is
e11

In the early stage of the atmospheric response (late autumn), upward surface heat flux anomalies in the Barents Sea MIZ lead to significant warming of the lower troposphere (Fig. 4a), but the anomalous circulation is not significant yet (Figs. 4b,c). In this stage, the diabatic heating in the Barents Sea MIZ is compensated by the eddy effect . This is shown in Fig. 15, which displays the postsummer evolution of the AWT-associated terms from Eq. (9) at 925 hPa integrated over the northeastern Barents Sea (black box in Fig. 14e). In winter, when strong wind anomalies develop, the combined effect of diabatic warming and eddy heat flux divergence acts as a heat sink in the lower troposphere over the Barents/Kara Sea region (Fig. 14f).

Fig. 15.
Fig. 15.

Time-lagged regression coefficients of the terms from Eq. (9) at 925 hPa averaged over the black box in Fig. 14e regressed onto the summer AWT index in the 1982–2005 period. The anomalies are in degrees Celsius per day per unit AWT index. Filled symbols denote anomalies statistically significant at the 90% confidence level. Positive lags correspond to the AWT index leading the atmospheric variables calculated as 4-month averages with the interval of 1 month. Lag 6 months corresponds to winter (December–March).

Citation: Journal of Climate 27, 23; 10.1175/JCLI-D-13-00763.1

b. Vertical structure of diabatic heating

Figures 16a–e display the AWT-associated terms from Eq. (9) along a vertical section at 77.5°N (marked with a magenta line in Fig. 14) in winter. The section runs through the low-level diabatic heating anomalies in the marginal ice zone on both sides of the Nordic seas (Fig. 14e). Apparently, the diabatic heating anomalies drastically diminish with height in the PBL (Fig. 16e). However, significant vertical extensions of the surface maxima in the Barents Sea and Greenland Sea MIZ reach the upper troposphere. As oceanic forcing strongly affects sea ice extent (see section 3a) and sea ice extent influences cloud cover (e.g., Liu et al. 2012), radiative adjustments owing to anomalous cloudiness may contribute to these extensions. However, the deeper extension in the Barents Sea MIZ should primarily result from anomalous release of latent heat of condensation owing to anomalous vertical motion. This scenario is supported by the significant increase of precipitation during the anomalous ascent of air in this area (Figs. 10 and 9).

Fig. 16.
Fig. 16.

(a) The mean horizontal advection of anomalous temperature, (b) anomalous horizontal advection of mean temperature, (c) anomalous vertical advection of mean temperature, (d) anomalous eddy heat flux convergence, (e) anomalous diabatic heating, and (f) the sum of (d) and (e) along a vertical section at 77.5°N (thick magenta line in Fig. 14) in winters 1982/83–2005/06 regressed onto the previous summer AWT index. The CI is 0.2°C day−1 per unit AWT index. The contour and shading colors are explained in the caption to Fig. 2.

Citation: Journal of Climate 27, 23; 10.1175/JCLI-D-13-00763.1

The upper-tropospheric/lower-stratospheric cooling from the anomalous vertical advection of mean temperature over the Barents Sea is as strong (~1°C day−1 per unit AWT index) as the diabatic warming at the surface (Fig. 16c). This cooling is balanced with the combined warming from the diabatic source (Fig. 16e), a convergent eddy heat flux (Fig. 16d) and, to some extent, anomalous horizontal advection of mean temperature (Fig. 16b). The latter is important at the western edge of the lobe of significant cooling by the anomalous ascending motion. In the core of this lobe at 30°–45°E, the primary anomalous upper-level heat balance is
e12
Since the diabatic warming in the lower troposphere is counteracted by a larger eddy heat flux divergence, the vertical profile of effective forcing by over the Barents Sea in the positive phase of the AWT index is characterized by cooling in the PBL and warming aloft (Fig. 16f). This explains why significant AWT-associated winter air temperature anomalies extend in this area nearly to the tropopause (Fig. 4a).

6. Discussion

a. Comparison with other studies

Classical studies based on linearized steady-state, β-plane models show that anomalous diabatic heating in the atmosphere leads to a direct baroclinic response (e.g., Hoskins and Karoly 1981; Hendon and Hartmann 1982). Details of the response depend on whether the forcing is localized in low or mid latitudes and whether it is deep or shallow. Nonlinear models indicate that the atmospheric equilibrium response to extratropical SST or sea ice cover anomalies depends on transient eddy feedbacks that modify the direct baroclinic response into an equivalent barotropic response extending far beyond the forcing area (e.g., Kushnir et al. 2002; Alexander et al. 2004; Deser et al. 2004). Modeling studies of the transient evolution of the atmospheric circulation to imposed patterns of anomalous surface forcing in the North Atlantic sector (Ferreira and Frankignoul 2005; Deser et al. 2007) show that the initial, linear baroclinic response is fast (~5–10 days) and has a short persistence (2–3 weeks). The equilibrium equivalent barotropic structure of a hemispheric extent is reached in 2–4 months. It has 2–3 times larger amplitude than the initial response, and exhibits only limited influence of diabatic heating near the surface in the vicinity of the forcing area.

Anomalous oceanic forcing in the Nordic seas MIZ leads indeed to an indirect atmospheric response through eddy–mean flow interactions, as deduced here from a significant contribution of the eddy vorticity flux convergence to the anomalous vorticity budget. However, judging from correlations with the AWT index, the baroclinic response in the forcing area is even stronger than the indirect response. It is characterized by surface-intensified geostrophic wind anomalies around the MIZ and a deep vertical flow anomaly over the northern Barents Sea. The latter is sustained by anomalous Ekman pumping at the top of the PBL. Schlichtholz (2013) suggested that the atmospheric response to oceanic forcing in the Nordic seas area is strong and persistent because of a dynamic feedback between anomalous surface winds and ocean currents resulting from coexistence of anomalies in the atmospheric Ekman pumping and Ekman suction at the bottom of the ocean surface frictional layer. As a consequence, persistent anomalous vertical motion should appear in the free atmosphere. As shown here, this motion induces a quasi-meridional overturning circulation anomaly. Such an anomaly is consistent with the interpretation of low-level wind divergences associated with wintertime sea ice extent anomalies in the Nordic seas by Wu et al. (2004). They considered, however, wintertime sea ice extent anomalies in the Nordic seas as being forced by the concurrent large-scale atmospheric circulation rather than the ocean.

The frictionally modulated atmospheric response to oceanic forcing in the Nordic seas area is characterized by a spatial coherence between low-level vorticity and divergence anomalies. This is in contrast with the atmospheric response to surface forcing at lower latitudes where the planetary β effect (related to variations of the Coriolis parameter with latitude) is a key dynamic agent shaping the baroclinic response. At midlatitudes, anomalous low-level vortex-tube stretching is approximately in the Sverdrup balance with the anomalous advection of planetary vorticity (f · u′ ≈ −u′ · f). Consequently, low-level divergence anomalies tend to appear in quadrature with the corresponding vorticity (or geopotential height) anomalies. This is the case, for instance, of the classical study by Hoskins and Karoly (1981). They emphasized generation of sinking motion below the maximum of an imposed positive diabating heating anomaly in the midlatitudes by anomalous meridional winds. In their model, the diabatic heating anomaly was balanced with the anomalous horizontal advection of mean temperature along the prevailing meridional temperature gradient (, where y is the meridional coordinate). Our results show that a similar thermal balance holds at low levels in the area of diabatic heating in the Greenland Sea MIZ. However, the diabatic heating anomaly and the anomalous horizontal advection of mean temperature at low levels in the Barents Sea MIZ are both sources of heat in the positive phase of the AWT index and sinks of heat in the negative phase of the AWT index. The thermal restoring in this area occurs primarily through anomalous eddy heat flux convergence.

Oceanic forcing in the Nordic seas area leads to anomalous ascending motion in the region of positive diabatic heating anomalies. A rising motion associated with positive diabatic heating anomalies is also found in several studies on the midlatitude response to surface forcing. This is the case, for instance, in a study on the relationship between the atmospheric circulation and SST in the northwest Atlantic based on an atmospheric general circulation model and observations by Palmer and Zhaobo (1985). They postulated that particular profiles of mean winds and a particular combination of anomalous diabatic heating with transient eddy heat flux convergence favor ascending motion in the area of positive SST anomalies. In the Barents Sea MIZ, the combined effect of diabatic heating and eddy heat flux convergence acts as a very deep heat source. A maximum heating rate occurs in the upper troposphere/lower stratosphere where the anomalous diabatic and eddy-driven warming counteracts the dynamic cooling due to the anomalous rising motion.

Ascending motion associated with positive diabating heating anomalies was also found by Honda et al. (1999) in their numerical modeling study of the atmospheric response to anomalous sea ice extent in the Sea of Okhotsk. This association was also noted by White and Chen (2002) and White et al. (2004) in their studies of tropospheric response to SST and sea ice extent anomalies in the Antarctic circumpolar wave based on the NCEP–NCAR reanalysis. While the diabatic heating diagnosed by Honda et al. (1999) was very shallow, the diabatic heating diagnosed by White and Chen (2002) and White et al. (2004) was deep. In all of these studies the eddy heat flux convergence was found to be negligible. The eddy vorticity flux convergence played no significant role either. Therefore, the atmospheric response to oceanic forcing in the Nordic seas area is unique in terms of heat and vorticity budgets.

b. Climatic implications

Warm AWT anomalies lead to reduced sea ice cover in the Nordic seas, which is one of the main reasons for the large surface temperature anomalies and the tropospheric response to oceanic forcing (see section 3a). Therefore, the most important features of this response should be significantly linked to sea ice anomalies. Such a link is illustrated in Fig. 17, which shows the time-lagged correlation of the area-averaged upper-tropospheric vertical motion in the Barents Sea MIZ in winter with the total sea ice area (SIA) in the Nordic seas. A maximum correlation (r = 0.73) occurs when SIA leads by 1 month. Interestingly, another maximum (r = 0.66) appears at lag −10 months corresponding to SIA of the previous early spring (February–May). In between, there is a nonsignificant minimum (r = 0.29) at lag −5 months corresponding to SIA during late summer (July–October). This evolution should reflect reemergence of sea ice anomalies. Such a reemergence has been observed and modeled for the total Arctic SIA (e.g., Blanchard-Wrigglesworth et al. 2011). In the Nordic seas, it most likely results from surface reemergence of oceanic heat anomalies that are formed at (or advected toward) the ice edge in the deep mixed layer during the late winter/spring season, survive through summer below the seasonal pycnocline, and are reentrained into the deepening mixed layer in the following autumn/winter season (Schlichtholz 2013). Wintertime sea ice cover is significantly linked to the previous spring SST anomalies in the Greenland Sea as well as in the Barents Sea (Schlichtholz 2011).

Fig. 17.
Fig. 17.

Time-lagged correlation of the anomalies of vertical velocity (positive downward) at 400 hPa in the Barents Sea MIZ (averaged over the black box in Fig. 6f) in winters (December–March) 1982/83–2005/06 with the anomalies of the total sea ice area in the Nordic seas. Filled symbols denote correlations statistically significant at the 90% confidence level. SIA is calculated as 4-month averages of the sea ice concentration integrated over the black box in Fig. 3a with the interval of 1 month. Negative lags correspond to SIA leading the vertical velocity.

Citation: Journal of Climate 27, 23; 10.1175/JCLI-D-13-00763.1

During the most significant atmospheric response to oceanic forcing (in early winter/winter), the surface heat flux anomalies in the Nordic seas MIZ are accompanied by equally large surface heat flux anomalies of opposite sign on the open water side of the climatological ice edge. However, in the early stage (late autumn) and late stage (late winter/spring) of the atmospheric response, the anomalous upward fluxes in the MIZ corresponding to warm oceanic events largely exceed the anomalous downward fluxes in the open water. The difference is ~4 TW when integrated over the Nordic seas and averaged over a year (lags 1–12 months in Fig. 5). This difference corresponds to an area-averaged local surface heat flux anomaly equal to ~1 W m−2. This anomaly would decrease the temperature of a 200-m-thick ocean layer by ~0.04°C in a year. This decrease is an order of magnitude smaller than the magnitude the summer AWT anomaly in the BSO area (~0.4°C). Even though this scaling is very crude, it indicates that not all anomalous ocean heat content is passed to the atmosphere. A substantial part of temperature anomalies staying in the ocean should propagate into the Arctic Ocean. Observations indeed show such a propagation (e.g., Polyakov et al. 2011).

The downward lobes of anomalous surface heat flux in the open water that appear during warm AWT events on the eastern and western sides of the Nordic seas (Fig. 3c) are mainly driven by southwesterly surface wind anomalies in the Iceland–Barents Sea corridor. Southwesterly surface wind anomalies over the Nordic seas and downward turbulent surface heat fluxes south of upward fluxes in the Barents Sea and Greenland Sea MIZ appear also in the patterns of expected changes over the twenty-first century owing to a northward moving ice edge (e.g., Koenigk et al. 2013). This and the strong surface-intensified warming in the Barents Sea MIZ during warm AWT events (Figs. 2 and 4a) suggest that the mechanisms of atmospheric response to oceanic forcing studied here will play an important role during future climate changes.

In future climate projections, sea ice reduction in the Barents/Kara Sea is connected with cold winter events over Eurasia (e.g., Yang and Christensen 2012). This connection is also a characteristic of the present climate change (e.g., Petoukhov and Semenov 2010; Inoue et al. 2012; Francis and Vavrus 2012). This suggests that the local atmospheric response to oceanic forcing in the Nordic seas might have been accompanied by a remote response in recent decades. This scenario will be investigated elsewhere (P. Schlichtholz 2014, unpublished manuscript, “A new perspective on seasonal prediction of winter climate in middle latitudes”).

7. Conclusions

We have conducted a regression analysis between oceanic heat anomalies parameterized by an index of observed summertime Atlantic water temperature (AWT) at the entrance to the Barents Sea introduced by Schlichtholz and Houssais (2011) and atmospheric fields over the Nordic seas and adjacent areas in the following winter from the NCEP–NCAR reanalysis in the period 1982–2006. This analysis shows that the oceanically driven low-level airflow anomaly around the Nordic seas marginal ice zone is intensified at the surface. This intensification results from a thermally direct baroclinic adjustment of the atmospheric circulation to anomalous heating (corresponding to warm AWT anomalies and subsequent anomalous sea ice retreat) or cooling (corresponding to cold AWT anomalies and subsequent anomalous sea ice advance) from below. The atmospheric anomalies in the positive phase of the AWT index are schematically illustrated in Fig. 18.

Fig. 18.
Fig. 18.

Schematic of atmospheric response to warm anomalies of Atlantic water temperature (AWT+). AWT+ leads to negative sea ice cover anomalies (c), warm surface temperature anomalies , upward surface heat flux anomalies , and warm air temperature anomalies (T+) in the marginal ice zone (MIZ) of the Barents and Greenland Seas. A cyclonic wind disturbance (ζ+) develops around the hot spots in the planetary boundary layer (PBL). Strong anomalies of geostrophic flow occur on its southern rim along the Iceland–Barents Sea (IBS) corridor. In this area, warming by wind anomalies is compensated by anomalous diabatic cooling (J) and results in downward surface heat flux anomalies . In the Greenland Sea MIZ, cooling by wind anomalies balances the anomalous diabatic warming (J+) in the PBL. In the Barents Sea MIZ, the sum of J+ and in the PBL is balanced with anomalous cooling by eddies . At the tropopause, J+ plus anomalous warming by eddies are balanced with adiabatic cooling by anomalous ascending motion . The anomalous ascending motion (ω) is forced by upward anomalies of the Ekman pumping resulting from frictional convergence in the cyclonic surface wind disturbance. A quasi-meridional overturning circulation anomaly is closed by southward anomalies of ageostrophic wind () at the tropopause, anomalous descending motion (ω+) south of the IBS corridor, and northward in the PBL. Increases (P+) and decreases (P) in precipitation occur in the area of ascending and descending motion, respectively. Downward anomalies of the Ekman pumping south of the IBS corridor are associated with an anticyclonic wind disturbance (ζ) in the PBL. In the upper troposphere, a broad ζ and a narrow ζ+ are mainly driven by anomalous wind divergence and eddy vorticity flux convergence , respectively.

Citation: Journal of Climate 27, 23; 10.1175/JCLI-D-13-00763.1

The anomalous heating generates hot spots (localized warm temperature anomalies) in the planetary boundary layer (PBL) in the Barents Sea and Greenland Sea marginal ice zone (MIZ). Anomalous air temperature gradients induce a cyclonic wind disturbance that is antiparallel to the thermal wind shear around these hot spots. The weaker Greenland Sea hot spot is buried within the PBL, but the stronger Barents Sea hot spot is significant up to the tropopause.

As the thermally driven geostrophic flow anomaly is extreme at the surface, the frictional convergence in the cyclonic disturbance forces an upward vertical velocity (Ekman pumping) at the top of the PBL. This velocity in turn drives a compensating divergence in the upper troposphere, which in the Barents Sea hot spot is extreme at the tropopause. This divergence feeds a downward vertical motion south of the Iceland–Barents Sea (IBS) corridor. Anomalous Ekman suction at the top of the PBL that coexists with an anticyclonic surface wind disturbance in this area closes a quasi-meridional overturning circulation anomaly driven by the thermal forcing in the MIZ. The ascending and descending branches of this overturning cell sustain positive and negative precipitation anomalies, respectively. The overturning cell is accompanied by equivalent barotropic wind anomalies in the upper troposphere.

The anomalous upper-level divergence in the Barents Sea MIZ is a source of anticyclonic vorticity. It reduces the lateral extent of the anomalous surface-intensified cyclonic vortex and makes it tilt with height toward Greenland. The vortex-tube squashing by the upper-level divergence is principally balanced with the mean advection of anomalous vorticity. The upper-tropospheric extension of the surface cyclonic vortex is mainly driven by eddy–mean flow interactions, as revealed by anomalies in the vorticity budget.

Anomalies in the heat budget show that distinct dynamic feedbacks on the thermal atmospheric response to oceanic forcing occur in different parts of the Nordic seas area. In the Greenland Sea MIZ, the diabatic warming in the PBL corresponding to upward surface heat flux anomalies is balanced with cold advection of mean temperature by anomalous winds. In the Barents Sea MIZ, the anomalies of diabatic heating and horizontal advection of mean temperature have the same sign and approximately the same magnitude in the PBL. The combined warming from these sources is compensated by anomalous eddy heat flux divergence. In the open water, the warm advection by anomalous southwesterlies that develop on the common rim of the cyclonic disturbance in the MIZ and the anticyclonic disturbance south of the IBS corridor is balanced with diabatic cooling leading to downward surface heat flux anomalies.

The strong anomalous ascent of air in the Barents Sea MIZ causes anomalous adiabatic cooling. This cooling has the largest rate in the upper troposphere/lower stratosphere where it is mainly balanced by anomalous diabatic warming (presumably increased condensational heating) and eddy heat flux convergence. As a result, the vertical profile of the combined anomalous diabatic and eddy-driven heating in the Barents Sea MIZ during warm AWT events is characterized by cooling in the PBL and warming aloft.

Comparison of the above results with studies of atmospheric response to surface forcing in other regions indicates that the atmospheric response to oceanic heat anomalies in the Nordic seas area is unique because of the prominent role of surface friction and because of specific vertical profiles of diabatic heating and eddy heat flux convergence over the Barents Sea. Our analysis also shows that oceanic anomalies tend to move characteristic features of the atmospheric circulation in the Nordic seas area both horizontally and vertically. In particular, the anomalous dipole in the surface wind vorticity that during warm AWT events consists of the cyclonic vortex in the MIZ and the anticyclonic vortex south of the IBS corridor reflects a northward shift of the generally cyclonic surface circulation in this area. The anomalous quasi-meridional overturning cell with the deep rising motion in the Barents Sea MIZ during these events corresponds to a northward and upward displacement of a core in the generally rising motion over the Nordic seas. These findings indirectly support the scenario presented in the introduction that the ocean is also a flywheel of the mean wintertime atmospheric circulation in this region.

We have speculated that the mechanisms of air–sea interactions investigated here should be at work during future climate changes. However, one should be aware that our results, though significant based on statistical tests, are afflicted with uncertainty since they are based on a relatively small number of episodes of warm and cold water. Indeed, the AWT index exhibits only four positive phases and three negative phases in the period under study (Schlichtholz and Houssais 2011, Fig. 2). Moreover, the NCEP–NCAR reanalysis data used here have a relatively low resolution. Verification of our results can be made using, for instance, the recently compiled, highly resolved Interim European Centre for Medium-Range Weather Forecasts Re-Analysis (ERA-Interim) data available since 1979–present (Dee et al. 2011). A check may also be made using ocean temperature data from the Norwegian Institute of Marine Research, which has carried out hydrographic measurements in the Barents Sea opening typically four to six times per year in recent decades (e.g., Skagseth 2008).

Acknowledgments

This research was supported by the Institute of Oceanology of the Polish Academy of Sciences, by the Polish–Norwegian Research Programme through the project PAVE (Atlantic Water Pathways to the Arctic: Variability and Effects on Climate and Ecosystems), and by the Centre for Polar Studies (KNOW), Poland. The NOAA/OAR/ESRL PSD, Boulder, Colorado, USA, is acknowledged for providing the NCEP Reanalysis Derived data from their Web site at http://www.esrl.noaa.gov/psd/. Ocean temperature data were provided by the Oceanographic Database of the International Council for the Exploration of the Sea (http://www.ices.dk/ocean/) and the World Ocean Database (WOD05) of the National Oceanographic Data Center of NOAA (http://www.nodc.noaa.gov/). The author thanks anonymous reviewers for their helpful comments.

REFERENCES

  • Aagaard, K., , J. H. Swift, , and E. C. Carmack, 1985: Thermohaline circulation in the Arctic Mediterranean Seas. J. Geophys. Res., 90 (C3), 48334846, doi:10.1029/JC090iC03p04833.

    • Search Google Scholar
    • Export Citation
  • Ådlandsvik, B., , and H. Loeng, 1991: A study of the climatic system in the Barents Sea. Polar Res., 10, 4550, doi:10.1111/j.1751-8369.1991.tb00633.x.

    • Search Google Scholar
    • Export Citation
  • Alexander, M. A., , U. S. Bhatt, , J. E. Walsh, , M. S. Timlin, , J. S. Miller, , and J. D. Scott, 2004: The atmospheric response to realistic Arctic sea ice anomalies in an AGCM during winter. J. Climate, 17, 890905, doi:10.1175/1520-0442(2004)017<0890:TARTRA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Årthun, M., , T. Eldevik, , L. H. Smedsrud, , Ø. Skagseth, , and R. B. Ingvaldsen, 2012: Quantifying the influence of Atlantic heat on Barents Sea ice variability and retreat. J. Climate, 25, 47364743, doi:10.1175/JCLI-D-11-00466.1.

    • Search Google Scholar
    • Export Citation
  • Bengtsson, L., , V. A. Semenov, , and O. M. Johannessen, 2004: The early twentieth-century warming in the Arctic—A possible mechanism. J. Climate, 17, 40454057, doi:10.1175/1520-0442(2004)017<4045:TETWIT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Bitz, C. M., , M. M. Holland, , E. C. Hunke, , and R. E. Moritz, 2005: On the maintenance of sea-ice edge. J. Climate, 18, 29032921, doi:10.1175/JCLI3428.1.

    • Search Google Scholar
    • Export Citation
  • Blanchard-Wrigglesworth, E., , K. C. Armour, , C. M. Bitz, , and E. DeWeaver, 2011: Persistence and inherent predictability of Arctic sea ice in a GCM ensemble and observations. J. Climate, 24, 231250, doi:10.1175/2010JCLI3775.1.

    • Search Google Scholar
    • Export Citation
  • Bretherton, C. S., , M. Widmann, , V. P. Dymnikov, , J. M. Wallace, , and I. Bladé, 1999: The effective number of spatial degrees of freedom of a time-varying field. J. Climate, 12, 19902009, doi:10.1175/1520-0442(1999)012<1990:TENOSD>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Dee, D. P., and et al. , 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553597, doi:10.1002/qj.828.

    • Search Google Scholar
    • Export Citation
  • Deser, C., , G. Magnusdottir, , R. Saravanan, , and A. Philips, 2004: The effects of North Atlantic SST and sea ice anomalies on the winter circulation in CCM3. Part II: Direct and indirect components of the response. J. Climate, 17, 877889, doi:10.1175/1520-0442(2004)017<0877:TEONAS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Deser, C., , R. A. Tomas, , and S. Peng, 2007: The transient atmospheric circulation response to North Atlantic SST and sea ice anomalies. J. Climate, 20, 47514767, doi:10.1175/JCLI4278.1.

    • Search Google Scholar
    • Export Citation
  • Ferreira, D., , and C. Frankignoul, 2005: The transient atmospheric response to midlatitude SST anomalies. J. Climate, 18, 10491067, doi:10.1175/JCLI-3313.1.

    • Search Google Scholar
    • Export Citation
  • Francis, J. A., , and E. Hunter, 2007: Drivers of declining sea ice in the Arctic winter: A tale of two seas. Geophys. Res. Lett., 34, L17503, doi:10.1029/2007GL030995.

    • Search Google Scholar
    • Export Citation
  • Francis, J. A., , and S. J. Vavrus, 2012: Evidence linking Arctic amplification to extreme weather in mid-latitudes. Geophys. Res. Lett., 39, L06801, doi:10.1029/2012GL051000.

    • Search Google Scholar
    • Export Citation
  • Furevik, T., 2001: Annual and interannual variability of Atlantic Water temperatures in the Norwegian and Barents Seas. Deep-Sea Res. I, 48, 383404, doi:10.1016/S0967-0637(00)00050-9.

    • Search Google Scholar
    • Export Citation
  • Hendon, H. H., , and D. L. Hartmann, 1982: Stationary waves on a sphere: Sensitivity to thermal feedback. J. Atmos. Sci., 39, 19061920, doi:10.1175/1520-0469(1982)039<1906:SWOASS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Holopainen, E. O., 1978: On the dynamic forcing of the long-term mean flow by the large-scale Reynolds’ stresses in the atmosphere. J. Atmos. Sci., 35, 15961604, doi:10.1175/1520-0469(1978)035<1596:OTDFOT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Honda, M., , K. Yamazaki, , H. Nakamura, , and K. Takeuchi, 1999: Dynamic and thermodynamic characteristics of atmospheric response to anomalous sea-ice extent in the Sea of Okhotsk. J. Climate, 12, 33473358, doi:10.1175/1520-0442(1999)012<3347:DATCOA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hoskins, B. J., , and D. J. Karoly, 1981: The steady linear response of a spherical atmosphere to thermal and orographic forcing. J. Atmos. Sci., 38, 11791196, doi:10.1175/1520-0469(1981)038<1179:TSLROA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Ikeda, M., 1990: Decadal oscillations of the air–ice–ocean system in the Northern Hemisphere. Atmos.–Ocean, 28, 106139, doi:10.1080/07055900.1990.9649369.

    • Search Google Scholar
    • Export Citation
  • Inoue, J., , M. E. Hori, , and K. Takaya, 2012: The role of Barents Sea ice in the wintertime cyclone track and emergence of a warm-Arctic cold-Siberian anomaly. J. Climate, 25, 25612568, doi:10.1175/JCLI-D-11-00449.1.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and et al. , 1996: The NCEP–NCAR 40-yr Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437471, doi:10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Koenigk, T., , and L. Brodeau, 2014: Ocean heat transport into the Arctic in the twentieth and twenty-first century in EC-Earth. Climate Dyn., 42, 31013120, doi:10.1007/s00382-013-1821-x.

    • Search Google Scholar
    • Export Citation
  • Koenigk, T., , U. Mikolajewicz, , J. H. Jungclaus, , and A. Kroll, 2009: Sea ice in the Barents Sea: Seasonal to interannual variability and climate feedbacks in a global coupled model. Climate Dyn., 32, 11191138, doi:10.1007/s00382-008-0450-2.

    • Search Google Scholar
    • Export Citation
  • Koenigk, T., , L. Brodeau, , R. Graversen, , J. Karlsson, , G. Svensson, , M. Tjernström, , U. Willén, , and K. Wyser, 2013: Arctic climate change in 21st century CMIP5 simulations with EC-Earth. Climate Dyn., 40, 27192743, doi:10.1007/s00382-012-1505-y.

    • Search Google Scholar
    • Export Citation
  • Kushnir, Y., , W. A. Robinson, , I. Bladé, , N. M. Hall, , S. Peng, , and R. Sutton, 2002: Atmospheric GCM response to extratropical SST anomalies: Synthesis and evaluation. J. Climate, 15, 22332256, doi:10.1175/1520-0442(2002)015<2233:AGRTES>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Liu, Y., , J. R. Key, , Z. Liu, , X. Wang, , and S. J. Vavrus, 2012: A cloudier Arctic expected with diminishing sea ice. Geophys. Res. Lett., 39, L05705, doi:10.1029/2012GL051251.

    • Search Google Scholar
    • Export Citation
  • Overland, J. E., , and M. Wang, 2010: Large-scale atmospheric circulation changes are associated with the recent loss of Arctic sea ice. Tellus, 62A, 19, doi:10.1111/j.1600-0870.2009.00421.x.

    • Search Google Scholar
    • Export Citation
  • Palmer, T. N., , and S. Zhaobo, 1985: A modeling and observational study of the relationship between sea surface temperature in the northwest Atlantic and the atmospheric general circulation. Quart. J. Roy. Meteor. Soc., 111, 947975, doi:10.1002/qj.49711147003.

    • Search Google Scholar
    • Export Citation
  • Pedlosky, J., 1987: Geophysical Fluid Dynamics. 2nd ed. Springer-Verlag, 710 pp.

  • Petoukhov, V., , and V. A. Semenov, 2010: A link between reduced Barents-Kara sea ice and cold winter extremes over northern continents. J. Geophys. Res., 115, D21111, doi:10.1029/2009JD013568.

    • Search Google Scholar
    • Export Citation
  • Polyakov, I. V., and et al. , 2011: Fate of early 2000s Arctic warm water pulse. Bull. Amer. Meteor. Soc., 92, 561566, doi:10.1175/2010BAMS2921.1.

    • Search Google Scholar
    • Export Citation
  • Schlichtholz, P., 2011: Influence of oceanic heat variability on sea ice anomalies in the Nordic Seas. Geophys. Res. Lett., 38, L05705, doi:10.1029/2010GL045894.

    • Search Google Scholar
    • Export Citation
  • Schlichtholz, P., 2013: Observational evidence for oceanic forcing of atmospheric variability in the Nordic seas area. J. Climate, 26, 29572975, doi:10.1175/JCLI-D-11-00594.1.

    • Search Google Scholar
    • Export Citation
  • Schlichtholz, P., , and M.-N. Houssais, 2011: Forcing of oceanic heat anomalies by air–sea interactions in the Nordic Seas area. J. Geophys. Res., 116, C01006, doi:10.1029/2009JC005944.

    • Search Google Scholar
    • Export Citation
  • Seager, R., , D. S. Battisti, , J. Yin, , N. Gordon, , N. Naik, , A. C. Clement, , and M. A. Cane, 2002: Is the Gulf Stream responsible for Europe’s mild winters? Quart. J. Roy. Meteor. Soc., 128, 25632586, doi:10.1256/qj.01.128.

    • Search Google Scholar
    • Export Citation
  • Skagseth, Ø., 2008: Recirculation of Atlantic Water in the western Barents Sea. Geophys. Res. Lett., 35, L11606, doi:10.1029/2008GL033785.

    • Search Google Scholar
    • Export Citation
  • White, W. B., , and S.-C. Chen, 2002: Thermodynamic mechanisms responsible for the tropospheric response to SST anomalies in the Antarctic circumpolar wave. J. Climate, 15, 25772596, doi:10.1175/1520-0442(2002)015<2577:TMRFTT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • White, W. B., , P. Glorsen, , and I. Simmonds, 2004: Tropospheric response in the Antarctic circumpolar wave along the sea ice edge. J. Climate, 17, 27652779, doi:10.1175/1520-0442(2004)017<2765:TRITAC>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Wu, B., , J. Wang, , and J. Walsh, 2004: Possible feedback of winter sea ice in the Greenland and Barents Seas on the local atmosphere. Mon. Wea. Rev., 132, 18681876, doi:10.1175/1520-0493(2004)132<1868:PFOWSI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Yang, S., , and J. H. Christensen, 2012: Arctic sea ice reduction and European cold winters in CMIP5 climate change experiments. Geophys. Res. Lett., 39, L20707, doi:10.1029/2012GL053338.

    • Search Google Scholar
    • Export Citation
Save