• Alexander, M. A., , and J. D. Scott, 1997: Surface flux variability over the North Pacific and North Atlantic Oceans. J. Climate, 10, 29632978, doi:10.1175/1520-0442(1997)010<2963:SFVOTN>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Barnston, A. G., , and R. E. Livezey, 1987: Classification, seasonality, and persistence of low-frequency atmospheric circulation patterns. Mon. Wea. Rev., 115, 10831126, doi:10.1175/1520-0493(1987)115<1083:CSAPOL>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Barsugli, J. J., , and D. S. Battisti, 1998: The basic effects of atmosphere–ocean thermal coupling on midlatitude variability. J. Atmos. Sci., 55, 477493, doi:10.1175/1520-0469(1998)055<0477:TBEOAO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Benedict, J. J., , S. Lee, , and S. B. Feldstein, 2004: Synoptic view of the North Atlantic oscillation. J. Atmos. Sci., 61, 121144, doi:10.1175/1520-0469(2004)061<0121:SVOTNA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Bond, N. A., , and M. F. Cronin, 2008: Regional weather patterns during anomalous air–sea fluxes at the Kuroshio Extension Observatory (KEO). J. Climate, 21, 16801697, doi:10.1175/2007JCLI1797.1.

    • Search Google Scholar
    • Export Citation
  • Bretherton, C. S., , C. Smith, , and J. M. Wallace, 1992: An intercomparison of methods for finding coupled patterns in climate data. J. Climate, 5, 541560, doi:10.1175/1520-0442(1992)005<0541:AIOMFF>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Cayan, D. R., , and D. H. Peterson, 1989: The influence of North Pacific atmospheric circulation on streamflow in the west. Aspects of Climate Variability in the Pacific and the Western Americas, Geophys. Monogr., Vol. 35, Amer. Geophys. Union, 375–397.

  • Chang, E. K. M., 2009: Are band-pass variance statistics useful measures of storm track activity? Re-examining storm track variability associated with the NAO using multiple storm track measures. Climate Dyn., 33, 277296, doi:10.1007/s00382-009-0532-9.

    • Search Google Scholar
    • Export Citation
  • Chelton, D. B., , and S.-P. Xie, 2010: Coupled ocean–atmosphere interaction at oceanic mesoscales. Oceanography, 23, 5269, doi:10.5670/oceanog.2010.05.

    • Search Google Scholar
    • Export Citation
  • Chelton, D. B., and et al. , 2001: Observations of coupling between surface wind stress and sea surface temperature in the eastern tropical Pacific. J. Climate, 14, 14791498, doi:10.1175/1520-0442(2001)014<1479:OOCBSW>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Chelton, D. B., , M. G. Schlax, , M. H. Freilich, , and R. F. Milliff, 2004: Satellite measurements reveal persistent small-scale features in ocean winds. Science, 303, 978983, doi:10.1126/science.1091901.

    • Search Google Scholar
    • Export Citation
  • Chhak, K. C., , E. Di Lorenzo, , N. Schneider, , and P. F. Cummins, 2009: Forcing of low-frequency ocean variability in the northeast Pacific. J. Climate, 22, 12551276, doi:10.1175/2008JCLI2639.1.

    • Search Google Scholar
    • Export Citation
  • Colucci, S. J., , and J. C. Davenport, 1987: Rapid surface anticyclogenesis: Synoptic climatology and attendant large-scale circulation changes. Mon. Wea. Rev., 115, 822836, doi:10.1175/1520-0493(1987)115<0822:RSASCA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Compo, G. P., and et al. , 2011: The Twentieth Century Reanalysis Project. Quart. J. Roy. Meteor. Soc., 137, 128, doi:10.1002/qj.776.

    • Search Google Scholar
    • Export Citation
  • Deser, C., , and M. L. Blackmon, 1993: Surface climate variations over the North Atlantic Ocean during winter—1900–1989. J. Climate, 6, 17431753, doi:10.1175/1520-0442(1993)006<1743:SCVOTN>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Di Lorenzo, E., and et al. , 2008: North Pacific Gyre Oscillation links ocean climate and ecosystem change. Geophys. Res. Lett., 35, L08607, doi:10.1029/2007GL032838.

    • Search Google Scholar
    • Export Citation
  • Feldstein, S. B., 2003: The dynamics of NAO teleconnection pattern growth and decay. Quart. J. Roy. Meteor. Soc., 129, 901924, doi:10.1256/qj.02.76.

    • Search Google Scholar
    • Export Citation
  • Frankignoul, C., , and K. Hasselmann, 1977: Stochastic climate models. Part II: Application to sea-surface temperature anomalies and thermocline variability. Tellus, 29, 289305, doi:10.1111/j.2153-3490.1977.tb00740.x.

    • Search Google Scholar
    • Export Citation
  • Frankignoul, C., , N. Sennechael, , Y. O. Kwon, , and M. A. Alexander, 2011: Influence of the meridional shifts of the Kuroshio and the Oyashio Extensions on the atmospheric circulation. J. Climate, 24, 762777, doi:10.1175/2010JCLI3731.1.

    • Search Google Scholar
    • Export Citation
  • Harnik, N., , and E. K. M. Chang, 2003: Storm track variations as seen in radiosonde observations and reanalysis data. J. Climate, 16, 480495, doi:10.1175/1520-0442(2003)016<0480:STVASI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hasselmann, K., 1976: Stochastic climate models: Part I. Theory. Tellus, 28, 473485, doi:10.1111/j.2153-3490.1976.tb00696.x.

  • Hoskins, B. J., , and P. J. Valdes, 1990: On the existence of storm-tracks. J. Atmos. Sci., 47, 18541864, doi:10.1175/1520-0469(1990)047<1854:OTEOST>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hurrell, J. W., , and H. VanLoon, 1997: Decadal variations in climate associated with the North Atlantic oscillation. Climatic Change, 36, 301326, doi:10.1023/A:1005314315270.

    • Search Google Scholar
    • Export Citation
  • Hurrell, J. W., , and C. Deser, 2010: North Atlantic climate variability: The role of the North Atlantic Oscillation. J. Mar. Syst., 79, 231244, doi:10.1016/j.jmarsys.2009.11.002.

    • Search Google Scholar
    • Export Citation
  • Hurrell, J. W., , Y. Kushnir, , G. Ottersen, , and M. Visbeck, 2003: An overview of the North Atlantic oscillation. The North Atlantic Oscillation: Climatic Significance and Environmental Impact, J. W. Hurrell et al., Eds., Amer. Geophys. Union, 1–33.

  • Kalnay, E., and et al. , 1996: The NCEP/NCAR 40-Year 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
  • Kelly, K. A., , R. J. Small, , R. M. Samelson, , B. Qiu, , T. M. Joyce, , Y.-O. Kwon, , and M. F. Cronin, 2010: Western boundary currents and frontal air–sea interaction: Gulf Stream and Kuroshio Extension. J. Climate,23, 5644–5667, doi:10.1175/2010JCLI3346.1.

  • Kushnir, Y., , W. A. Robinson, , I. Bladé, , N. M. J. 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
  • Kwon, Y.-O., , and T. M. Joyce, 2013: Northern Hemisphere winter atmospheric transient eddy heat fluxes and the Gulf Stream and Kuroshio–Oyashio Extension variability. J. Climate, 26, 98399859, doi:10.1175/JCLI-D-12-00647.1.

    • Search Google Scholar
    • Export Citation
  • Kwon, Y.-O., , M. A. Alexander, , N. A. Bond, , C. Frankignoul, , H. Nakamura, , B. Qiu, , and L. A. Thompson, 2010: Role of the Gulf Stream and Kuroshio–Oyashio systems in large-scale atmosphere–ocean interaction: A review. J. Climate,23, 3249–3281, doi:10.1175/2010JCLI3343.1.

  • Linkin, M. E., , and S. Nigam, 2008: The North Pacific Oscillation–west Pacific teleconnection pattern: Mature-phase structure and winter impacts. J. Climate, 21, 19791997, doi:10.1175/2007JCLI2048.1.

    • Search Google Scholar
    • Export Citation
  • Maidens, A., , A. Arribas, , A. A. Scaife, , C. MacLachlan, , D. Peterson, , and J. Knight, 2013: The influence of surface forcings on prediction of the North Atlantic Oscillation regime of winter 2010/11. Mon. Wea. Rev., 141, 38013813, doi:10.1175/MWR-D-13-00033.1.

    • Search Google Scholar
    • Export Citation
  • Maloney, E. D., , and D. B. Chelton, 2006: An assessment of the sea surface temperature influence on surface wind stress in numerical weather prediction and climate models. J. Climate, 19, 27432762, doi:10.1175/JCLI3728.1.

    • Search Google Scholar
    • Export Citation
  • Mantua, N. J., , S. R. Hare, , Y. Zhang, , J. M. Wallace, , and R. C. Francis, 1997: A Pacific interdecadal climate oscillation with impacts on salmon production. Bull. Amer. Meteor. Soc., 78, 10691079, doi:10.1175/1520-0477(1997)078<1069:APICOW>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Marshall, J., and et al. , 2001: North Atlantic climate variability: Phenomena, impacts and mechanisms. Int. J. Climatol., 21, 18631898, doi:10.1002/joc.693.

    • Search Google Scholar
    • Export Citation
  • Miller, A. J., , and N. Schneider, 2000: Interdecadal climate regime dynamics in the North Pacific Ocean: Theories, observations and ecosystem impacts. Prog. Oceanogr., 47, 355379, doi:10.1016/S0079-6611(00)00044-6.

    • Search Google Scholar
    • Export Citation
  • Miller, A. J., , D. R. Cayan, , T. P. Barnett, , N. E. Graham, , and J. M. Oberhuber, 1994: The 1976–77 climate shift of the Pacific Ocean. Oceanography, 7, 2126, doi:10.5670/oceanog.1994.11.

    • Search Google Scholar
    • Export Citation
  • Miller, A. J., , F. Chai, , S. Chiba, , J. R. Moisan, , and D. J. Neilson, 2004: Decadal-scale climate and ecosystem interactions in the North Pacific Ocean. J. Oceanogr., 60, 163188, doi:10.1023/B:JOCE.0000038325.36306.95.

    • Search Google Scholar
    • Export Citation
  • Nakamura, H., 1996: Year-to-year and interdecadal variability in the activity of intraseasonal fluctuations in the Northern Hemisphere wintertime circulation. Theor. Appl. Climatol., 55, 1932, doi:10.1007/BF00864700.

    • Search Google Scholar
    • Export Citation
  • Nakamura, H., , G. Lin, , and T. Yamagata, 1997: Decadal climate variability in the North Pacific during the recent decades. Bull. Amer. Meteor. Soc., 78, 22152225, doi:10.1175/1520-0477(1997)078<2215:DCVITN>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Nakamura, H., , T. Sampe, , Y. Tanimoto, , and A. Shimpo, 2004: Observed associations among storm tracks, jet streams and midlatitude oceanic fronts. Earth’s Climate: The Ocean–Atmosphere Interaction, Geophys. Mongr., Vol. 147, Amer. Geophys. Union, 329–345.

  • Nakamura, H., , T. Sampe, , A. Goto, , W. Ohfuchi, , and S.-P. Xie, 2008: On the importance of midlatitude oceanic frontal zones for the mean state and dominant variability in the tropospheric circulation. Geophys. Res. Lett., 35, L15709, doi:10.1029/2008GL034010.

    • Search Google Scholar
    • Export Citation
  • Nakamura, M., , and S. Yamane, 2009: Dominant anomaly patterns in the near-surface baroclinicity and accompanying anomalies in the atmosphere and oceans. Part I: North Atlantic basin. J. Climate, 22, 880904, doi:10.1175/2008JCLI2297.1.

    • Search Google Scholar
    • Export Citation
  • Nakamura, M., , and S. Yamane, 2010: Dominant anomaly patterns in the near-surface baroclinicity and accompanying anomalies in the atmosphere and oceans. Part II: North Pacific basin. J. Climate, 23, 64456467, doi:10.1175/2010JCLI3017.1.

    • Search Google Scholar
    • Export Citation
  • Newman, M., , G. P. Compo, , and M. A. Alexander, 2003: ENSO-forced variability of the Pacific decadal oscillation. J. Climate, 16, 38533857, doi:10.1175/1520-0442(2003)016<3853:EVOTPD>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • O’Neill, L. W., , D. B. Chelton, , and S. K. Esbensen, 2012: Covariability of surface wind and stress responses to sea surface temperature fronts. J. Climate,25, 5916–5942, doi:10.1175/JCLI-D-11-00230.1.

  • Qiu, B., , S. Chen, , and P. Hacker, 2004: Synoptic-scale air–sea flux forcing in the western North Pacific: Observations and their impact on SST and the mixed layer. J. Phys. Oceanogr., 34, 21482159, doi:10.1175/1520-0485(2004)034<2148:SAFFIT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Rivière, G., , and I. Orlanski, 2007: Characteristics of the Atlantic storm-track eddy activity and its relation with the North Atlantic Oscillation. J. Atmos. Sci., 64, 241266, doi:10.1175/JAS3850.1.

    • Search Google Scholar
    • Export Citation
  • Rogers, J. C., 1981: The North Pacific oscillation. J. Climatol., 1, 3957, doi:10.1002/joc.3370010106.

  • Rogers, J. C., 1990: Patterns of low-frequency monthly sea level pressure variability (1899–1986) and associated wave cyclone frequencies. J. Climate, 3, 13641379, doi:10.1175/1520-0442(1990)003<1364:POLFMS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Rogers, J. C., 1997: North Atlantic storm track variability and its association to the North Atlantic Oscillation and climate variability of northern Europe. J. Climate, 10, 16351647, doi:10.1175/1520-0442(1997)010<1635:NASTVA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Saha, S., and et al. , 2010: The NCEP Climate Forecast System Reanalysis. Bull. Amer. Meteor. Soc., 91, 10151057, doi:10.1175/2010BAMS3001.1.

    • Search Google Scholar
    • Export Citation
  • Sampe, T., , and S.-P. Xie, 2007: Mapping high sea winds from space: A global climatology. Bull. Amer. Meteor. Soc., 88, 19651978, doi:10.1175/BAMS-88-12-1965.

    • Search Google Scholar
    • Export Citation
  • Saravanan, R., 1998: Atmospheric low-frequency variability and its relationship to midlatitude SST variability: Studies using the NCAR climate system model. J. Climate, 11, 13861404, doi:10.1175/1520-0442(1998)011<1386:ALFVAI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Shaman, J., , R. M. Samelson, , and E. Skyllingstad, 2010: Air–sea fluxes over the Gulf Stream region: Atmospheric controls and trends. J. Climate, 23, 26512670, doi:10.1175/2010JCLI3269.1.

    • Search Google Scholar
    • Export Citation
  • Taguchi, B., , H. Nakamura, , M. Nonaka, , and S.-P. Xie, 2009: Influences of the Kuroshio/Oyashio Extensions on air–sea heat exchanges and storm-track activity as revealed in regional atmospheric model simulations for the 2003/04 cold season. J. Climate, 22, 65366560, doi:10.1175/2009JCLI2910.1.

    • Search Google Scholar
    • Export Citation
  • Taguchi, B., , H. Nakamura, , M. Nonaka, , N. Komori, , A. Kuwano-Yoshida, , K. Takaya, , and A. Goto, 2012: Seasonal evolutions of atmospheric response to decadal SST anomalies in the North Pacific subarctic frontal zone: Observations and a coupled model simulation. J. Climate,25, 111–139, doi:10.1175/JCLI-D-11-00046.1.

  • Thompson, D. W. J., , and J. M. Wallace, 1998: The Arctic Oscillation signature in the wintertime geopotential height and temperature fields. Geophys. Res. Lett., 25, 12971300, doi:10.1029/98GL00950.

    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., , and J. W. Hurrell, 1994: Decadal atmosphere–ocean variations in the Pacific. Climate Dyn., 9, 303319, doi:10.1007/BF00204745.

    • Search Google Scholar
    • Export Citation
  • Vallis, G. K., , and E. P. Gerber, 2008: Local and hemispheric dynamics of the North Atlantic Oscillation, annular patterns and the zonal index. Dyn. Atmos. Oceans, 44, 184212, doi:10.1016/j.dynatmoce.2007.04.003.

    • Search Google Scholar
    • Export Citation
  • van Loon, H., , and J. C. Rogers, 1978: The seesaw in winter temperatures between Greenland and northern Europe. Part I: General description. Mon. Wea. Rev., 106, 296310, doi:10.1175/1520-0493(1978)106<0296:TSIWTB>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Walsh, J. E., , A. S. Phillips, , D. H. Portis, , and W. L. Chapman, 2001: Extreme cold outbreaks in the United States and Europe, 1948–99. J. Climate, 14, 26422658, doi:10.1175/1520-0442(2001)014<2642:ECOITU>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Wheeler, D. D., , V. L. Harvey, , D. E. Atkinson, , R. L. Collins, , and M. J. Mills, 2011: A climatology of cold air outbreaks over North America: WACCM and ERA-40 comparison and analysis. J. Geophys. Res., 116, D12107, doi:10.1029/2011JD015711.

    • Search Google Scholar
    • Export Citation
  • Woollings, T., , B. Hoskins, , M. Blackburn, , and P. Berrisford, 2008: A new Rossby wave–breaking interpretation of the North Atlantic Oscillation. J. Atmos. Sci., 65, 609626, doi:10.1175/2007JAS2347.1.

    • Search Google Scholar
    • Export Citation
  • Yu, L., , X. Jin, , and R. A. Weller, 2008: Multidecadel global flux datasets from the Objectively Analyzed Air–Sea Fluxes (OAFlux) Project: Latent and sensible heat fluxes, ocean evaporation, and related surface meteorological variables. OAFlux Project Tech. Rep. OA-2008-01, Woods Hole Oceanographic Institution, 64 pp.

  • View in gallery

    Climatological winter season (NDJFM) mean THF in the (a) KER and (b) GSR based on 61-yr NCEP–NCAR reanalysis (W m−2). Area used to define extreme flux events in the KER and GSR are indicated by the rectangles.

  • View in gallery

    (a) The 61-yr winter season (NDJFM) NCEP–NCAR reanalysis daily THF (black solid), the standard deviation (gray shaded), and the HF80 critical value chosen to define extreme flux events (black dashed) for the KER at top and GSR at bottom. (b) Histogram (number of winters) of number of event days in each winter during 1948–2009, where the number of event days is obtained by counting the days when the area-averaged daily THF exceeds HF80. (c) Histogram (number of winters) of fraction of event-day THF contribution to the total THF in each winter, where the contribution is calculated by integrating the THF on event days divided by the total THF accumulated in all winter days. (d) Histogram (number of extreme flux events) of extreme flux event durations in days, where the extreme flux event duration is defined as the number of continuous days when the THF is above HF80. In all the histogram plots, black is for the GSR in the Atlantic and gray is for the KER in the Pacific.

  • View in gallery

    Event-day composites derived by averaging 1–8-day bandpass filtered THF (W m−2; color shaded and contours, positive being THF into the atmosphere) in the (a) KER and (b) GSR; 2-m air temperature (contours; °C) and specific humidity (color shaded; kg kg−1) in the (c) KER and (d) GSR; and SLP (color shaded; hPa) and 10-m wind (vector; m s−1) in the (e) KER and (f) GSR. All the variables were derived from the NCEP–NCAR reanalysis (1948–2009).

  • View in gallery

    Total boreal winter THF (top plot), non-event-day THF (middle plot), and event-day THF (bottom plot) for 20CRV2 (dashed), NCEP–NCAR reanalysis (black solid), and NCEP CFSR (gray solid) reanalysis in the (a) KER and (b) GSR. Also shown are normalized total THF (black) and cold storm–induced THF (gray) based on NCEP–NCAR reanalysis, in the (c) KER and (d) GSR. The cold storm–induced THF is derived by first regressing the 1–8-day bandpass filtered daily SLP onto the event day composite of SLP and then integrating THF over the days during which the regression coefficient exceeds 1.0.

  • View in gallery

    (a) The first principal component time series and (c) the associated EOF (shaded and contoured with a contour interval of 1 hPa) in the North Pacific derived from NDJFM SLP anomalies using NCEP–NCAR reanalysis. (b) The second principal component time series and (d) the associated EOF (shaded and contoured with a contour interval of 1 hPa) in the North Atlantic.

  • View in gallery

    Spatial patterns of the first SVD mode (squared variance, 27%) between (a) event-day THF and (b) NDJFM SLP anomalies in the North Pacific derived from NCEP–NCAR reanalysis with zero contour shown in dashed. (c) Percentage of contribution from the first THF SVD mode to the total THF variance (shaded and contours). (d) Time series of the first SVD mode for event-day THF anomalies (gray solid) and SLP anomalies (dashed), overlaid with the ALP index (black solid).

  • View in gallery

    As in Fig. 6, but for the second SVD mode (squared variance, 23%) between event-day THF and NDJFM SLP anomalies in the North Atlantic.

  • View in gallery

    Lag composites of daily SLP anomalies (contours, hPa) in the (left) KER and (right) GSR on event days during ALP positive phase derived from 61 winter season (NDJFM) NCEP–NCAR reanalysis. The event days were identified based on the HF80 threshold for daily THF in each NDJFM. (a),(b) Lag = 0 corresponds to the occurrence of extreme events and lag = (c),(d) 1, (e),(f) 2, and (g),(h) 3 correspond to the subsequent 3 days.

  • View in gallery

    Winter season (NDJFM) mean SLP anomalies (contours, hPa) during (a) ALP and (b) EAP positive phase in the North Pacific and Atlantic, respectively, derived from NCEP–NCAR reanalysis. Composites of event-day SLP anomalies by integrating from lag 0 to lag 6 day during the positive phase of (c) ALP and (d) EAP. Composites of non-event-day SLP anomalies by integrating over the rest of days during positive phase of (e) ALP and (f) EAP. The sum of the event-day composite in (c),(d) and non-event-day composite in (e),(f) gives back the winter season mean SLP anomalies in (a),(b).

  • View in gallery

    First SVD mode (left) between NDJFM event-day THF and 300-hPa 〈υυ′〉 (squared variance, 15%) and (right) between NDJFM SLP anomalies and 300-hPa 〈υυ′〉 (squared variance, 34%) in the North Pacific derived from NCEP–NCAR reanalysis. SVD time series are shown with (a) event-day THF and (b) SLP anomalies in gray and 300-hPa 〈υυ′〉 in black. (c),(d) SVD spatial patterns of 300-hPa 〈υυ′〉 (shaded and contours). SVD spatial patterns (e) of event-day THF (shaded and contours) and (f) of SLP anomalies (shaded and contours). For all the spatial patterns, positive values are plotted with solid contours and negative values with dashed contours.

  • View in gallery

    As in Fig. 10, but for the second SVD mode (left) between NDJFM event-day THF and 300-hPa 〈υυ′〉 (squared variance, 15%) and (right) between NDJFM SLP anomalies and 300-hPa 〈υυ′〉 (squared variance, 20%) in the North Atlantic.

  • View in gallery

    Lag composite of 1–8-day bandpass filtered 300-hPa geopotential height for (left) event-day storms and (right) non-event-day storms in the North Pacific at (a),(b) lag −1, (c),(d) lag 0, (e),(f) lag 1, and (g),(h) lag 2. (i),(j) Trajectories of the low pressure center movement from lag −3 day to lag 4 day, superimposed on the low pressure composites at each of the lags. Contour intervals are 10 m for event-day composites and 2 m for non-event-day composites. Event-day storms and non-event-day storms were derived from 61 winter season (NDJFM) NCEP–NCAR reanalysis based on the HF80 threshold.

  • View in gallery

    As in Fig. 12, but for (left) event-day and (right) non-event-day storms in the North Atlantic.

  • View in gallery

    Seasonal mean SST (°C) anomalies regressed onto extreme flux event index in the (left) KER in the North Pacific and (right) GSR in the North Atlantic when SST anomalies (a),(b) lead extreme flux events by 1 month and (c),(d) SST lag by 1 month. Zero contours are in dashed and regression values significant at 95% confidence level based on a Student’s t test are shaded by black dots.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 870 870 301
PDF Downloads 61 61 27

Winter Extreme Flux Events in the Kuroshio and Gulf Stream Extension Regions and Relationship with Modes of North Pacific and Atlantic Variability

View More View Less
  • 1 Department of Oceanography, Texas A&M University, College Station, Texas, and Ocean University of China, Qingdao Collaborative Innovation Center of Marine Science and Technology, Qingdao, China
  • | 2 Department of Oceanography, and Department of Atmospheric Sciences, Texas A&M University, College Station, Texas, and Ocean University of China, Qingdao Collaborative Innovation Center of Marine Science and Technology, Qingdao, China
  • | 3 Department of Atmospheric Sciences, Texas A&M University, College Station, Texas
  • | 4 Ocean University of China, Qingdao Collaborative Innovation Center of Marine Science and Technology, Qingdao, China
© Get Permissions
Full access

Abstract

Boreal winter (November–March) extreme flux events in the Kuroshio Extension region (KER) of the northwestern Pacific and the Gulf Stream region (GSR) of the northwestern Atlantic are analyzed and compared, based on NCEP Climate Forecast System Reanalysis (CFSR), NCEP–NCAR reanalysis, and NOAA Twentieth Century Reanalysis data, as well as the observationally derived OAFlux dataset. These extreme flux events, most of which last less than 3 days, are characterized by cold air outbreaks (CAOs) with an anomalous northerly wind that brings cold and dry air from the Eurasian and North American continents to the KER and GSR, respectively. A close relationship between the extreme flux events over KER (GSR) and the Aleutian low pattern (ALP) [east Atlantic pattern (EAP)] is found with more frequent occurrence of the extreme flux events during a positive ALP (EAP) phase and vice versa. A further lag-composite analysis suggests that the ALP (EAP) is associated with accumulated effects of the synoptic winter storms accompanied by the extreme flux events and shows that the event-day storms tend to have a preferred southeastward propagation path over the North Pacific (Atlantic), potentially contributing to the southward shift of the storm track over the eastern North Pacific (Atlantic) basin during the ALP (EAP) positive phase. Finally, lag-regression analyses indicate a potential positive influence of sea surface temperature (SST) anomalies along the KER (GSR) on the development of the extreme flux events in the North Pacific (Atlantic).

Corresponding author address: Xiaohui Ma, O&M Building, MS3146, Department of Oceanography, Texas A&M University, College Station, TX 77840. E-mail: maxiaohui@tamu.edu

This article is included in the Climate Implications of Frontal Scale Air–Sea Interaction Special Collection.

Abstract

Boreal winter (November–March) extreme flux events in the Kuroshio Extension region (KER) of the northwestern Pacific and the Gulf Stream region (GSR) of the northwestern Atlantic are analyzed and compared, based on NCEP Climate Forecast System Reanalysis (CFSR), NCEP–NCAR reanalysis, and NOAA Twentieth Century Reanalysis data, as well as the observationally derived OAFlux dataset. These extreme flux events, most of which last less than 3 days, are characterized by cold air outbreaks (CAOs) with an anomalous northerly wind that brings cold and dry air from the Eurasian and North American continents to the KER and GSR, respectively. A close relationship between the extreme flux events over KER (GSR) and the Aleutian low pattern (ALP) [east Atlantic pattern (EAP)] is found with more frequent occurrence of the extreme flux events during a positive ALP (EAP) phase and vice versa. A further lag-composite analysis suggests that the ALP (EAP) is associated with accumulated effects of the synoptic winter storms accompanied by the extreme flux events and shows that the event-day storms tend to have a preferred southeastward propagation path over the North Pacific (Atlantic), potentially contributing to the southward shift of the storm track over the eastern North Pacific (Atlantic) basin during the ALP (EAP) positive phase. Finally, lag-regression analyses indicate a potential positive influence of sea surface temperature (SST) anomalies along the KER (GSR) on the development of the extreme flux events in the North Pacific (Atlantic).

Corresponding author address: Xiaohui Ma, O&M Building, MS3146, Department of Oceanography, Texas A&M University, College Station, TX 77840. E-mail: maxiaohui@tamu.edu

This article is included in the Climate Implications of Frontal Scale Air–Sea Interaction Special Collection.

1. Introduction

Modes of climate variability in the North Pacific and North Atlantic have been well documented and widely discussed in the literature (e.g., Hurrell and VanLoon 1997; Mantua et al. 1997; Nakamura et al. 1997; Hurrell et al. 2003). The most dominant patterns of North Pacific and Atlantic climate variability are the Pacific decadal oscillation (PDO) and the North Atlantic Oscillation (NAO), while the importance of other patterns, such as the North Pacific Oscillation (NPO) and the east Atlantic pattern (EAP), has also been recognized (Barnston and Livezey 1987; Linkin and Nigam 2008; Rogers 1981). These extratropical modes of variability are manifested in changes in large-scale atmospheric pressure, temperature, and precipitation patterns (Cayan and Peterson 1989; Hurrell and Deser 2010; Marshall et al. 2001; Miller et al. 2004; Rogers 1997; Trenberth and Hurrell 1994). They have also been linked to changes in storm track activity and atmospheric blocking activity in the North Pacific and Atlantic sectors (e.g., Chang 2009; Maidens et al. 2013; Nakamura 1996; Taguchi et al. 2012; Woollings et al. 2008). Understanding of these extratropical modes of climate variability has important implications for the understanding of climate predictability in the North Pacific and Atlantic, which is a key objective of the international Climate and Ocean: Variability, Predictability and Change (CLIVAR) program (http://www.clivar.org).

Dynamical mechanisms governing variability of midlatitude climate system are still not fully understood. Earlier studies suggested that decadal and longer time-scale sea surface temperature (SST) variability can simply arise from a red-noise ocean mixed-layer response to white-noise atmospheric forcing (Frankignoul and Hasselmann 1977; Hasselmann 1976). Later studies focused on how and to what extent midlatitude SST anomalies can feed back onto the atmosphere. Some studies indicate that there is active ocean–atmosphere coupling in the midlatitude (Miller and Schneider 2000). Other studies argue that the atmospheric response to midlatitude SST anomalies is weak compared to variability generated by its internal dynamics (Saravanan 1998). A review of studies on atmospheric response to midlatitude SST anomalies can be found in Kushnir et al. (2002). According to Kushnir et al. (2002), the weak SST-forced atmospheric variability can be achieved through two mechanisms: 1) a linear mechanism where passive thermodynamic feedbacks between surface heat fluxes and ocean mixed layer temperature act to reduce thermal damping, enhancing low-frequency variability of the atmosphere, and 2) a nonlinear mechanism where an eddy-mediated process links SST forcing to storm track activity. The former can be considered as a passive coupling between the atmosphere and ocean and can be understood as an extension of Hasselmann climate model (Hasselmann 1976) in the framework of a coupled atmospheric energy balance model and a slab ocean mixed layer model, as elegantly demonstrated by Barsugli and Battisti (1998). The latter involves complex interactions between atmospheric storm tracks and SST fronts along intense ocean western boundary current (WBCs) (Nakamura et al. 2004, 2008), which are less understood.

Many recent studies on midlatitude atmosphere–ocean interactions focus on the latter mechanism. Special attention has been paid to the Kuroshio Extension region (KER) in the North Pacific and the Gulf Stream region (GSR) in the North Atlantic. These strong WBCs are known to carry vast amount of oceanic heat poleward from the tropics and release it to the atmosphere through turbulent heat fluxes (THFs), destabilizing the atmosphere and favoring the development of storm tracks (Kelly et al. 2010; Kwon et al. 2010; Nakamura et al. 2004, 2008; Nakamura and Yamane 2009, 2010). Nakamura et al. (2004) argued that the sharp SST gradients along the Kuroshio and Gulf Stream fronts play a fundamental role in anchoring the North Pacific and North Atlantic storm track through their influence on the low-level baroclinicity of the atmosphere. In a suite of AGCM simulations, Taguchi et al. (2009) demonstrated that a smoothed SST gradient in the KER leads to a weak storm activity. Evidence of atmosphere–ocean coupling along the Kuroshio and Gulf Stream has also been presented using high-resolution satellite data (e.g., Chelton et al. 2001, 2004; Sampe and Xie 2007). By spatially filtering satellite measured wind speed and SST, a remarkable positive correlation between small-scale wind speed and underlying oceanic mesoscale eddies in eddy-rich regions, such as the KER and GSR, has been revealed (Chelton and Xie 2010; Maloney and Chelton 2006; O’Neill et al. 2012), suggesting an influence of mesoscale SST on the overlying atmosphere in these eddy-rich regions. A more recent coupled GCM study of atmospheric response to decadal SST anomalies in the North Pacific by Taguchi et al. (2012) further suggests that the PDO may also be influenced by SST anomalies along the subarctic frontal zone. Collectively, these recent studies point to the potentially important role of mesoscale and frontal-scale atmosphere–ocean interactions along the Kuroshio and Gulf Stream in large-scale climate variability in the North Pacific and Atlantic. Yet, dynamic linkages between them are still poorly understood.

A key process in mesoscale and frontal-scale atmosphere–ocean interactions along the Kuroshio and Gulf Stream is surface latent and sensible heat flux exchange; together they determine total turbulent heat exchanges between the atmosphere and ocean. It is well known that synoptic storm systems can be vital in regulating surface latent and sensible heat fluxes in these regions (Hoskins and Valdes 1990; Kwon and Joyce 2013; Taguchi et al. 2009). Shaman et al. (2010, hereafter SH10) recently showed, based on 60 years of NCEP–NCAR reanalysis data, that in the GSR a major proportion of wintertime surface THF is attributed to a relatively small number of high flux events associated with extreme synoptic storms. An earlier study of a 17-yr daily surface heat flux record from an AGCM simulation (Alexander and Scott 1997) revealed that in winter season surface heat flux variability on time scales less than 30 days can explain more than half of the total intraseasonal heat flux variance in both the North Pacific and Atlantic Oceans. Alexander and Scott (1997) further showed that, within the 30-day frequency band, the 3–10-day synoptic storms make the most important contribution to the intraseasonal heat flux variance and their activity affects the low-level atmospheric circulation variability, suggesting a linkage between storm activity and large-scale atmospheric circulation variability. A study of cold air outbreaks (CAOs) in the North Atlantic showed that enhanced CAOs appear to be associated with negative phases of the NAO (Walsh et al. 2001). All these studies point to the importance of extreme THF associated with winter synoptic storms in affecting the air–sea coupling in the WBC regimes and a potential linkage between the synoptic storms in these regions and large-scale atmospheric circulation.

Motivated by these previous studies, we attempt to advance our understanding of frontal-scale and mesoscale atmosphere–ocean interactions in the KER and GSR and their role in climate variability by focusing on two aspects. First, we will extend the extreme flux event analysis presented by SH10 to the KER and compare the results in the KER to those of SH10 in the GSR. Given that both these regions are considered to be the most intensive air–sea interaction regions in the extratropics (Kelly et al. 2010), it is of considerable interest to contrast differences and similarities between extreme flux events in these regions and their respective role in large-scale climate variability in the North Pacific and Atlantic sector. Second, we will investigate the relationship among variability of extreme flux events in the KER and GSR, North Pacific and Atlantic storm tracks, and modes of climate variability in North Pacific and Atlantic sector. Furthermore, we will examine the relationship between the extreme flux variability and SST variability in North Pacific and Atlantic, respectively.

The paper is organized as follows. Section 2 describes the datasets, the criteria to define extreme flux events, and the methodologies used to perform the analysis. Section 3 illustrates the general characteristics of extreme flux events in the KER and GSR and contrasts their similarities and differences in the two regions. Section 4 examines the relationship among KER and GSR extreme flux variability, North Pacific and Atlantic storm tracks, and modes of climate variability in North Pacific and Atlantic sector. Section 5 discusses the relationship between extreme flux variability and SST variability in the KER and GSR. Finally, section 6 summarizes and discusses the major findings of the study.

2. Data and analysis method

The study of surface turbulence heat flux variability in the WBC regions and its relation with midlatitude storms requires high-resolution data both temporally and spatially. The datasets we used are 6-hourly the National Centers for Environmental Prediction (NCEP) Climate Forecast System Reanalysis (CFSR, 1979–2009; Saha et al. 2010), NCEP–National Center for Atmospheric Research (NCAR) reanalysis (1948–2009; Kalnay et al. 1996), and NOAA Twentieth Century Reanalysis, version 2 (20CRV2, 1870–2009; Compo et al. 2011), as well as the daily Objectively Analyzed Air–Sea Fluxes (OAFlux, 1985–2009) dataset (Yu et al. 2008). Throughout this study, we analyzed all these datasets to examine and compare the variability of extreme flux events in the KER and GSR. The majority of the results discussed in this paper, however, are primarily based on the NCEP–NCAR reanalysis, as many previous studies have shown that the NCEP–NCAR reanalysis captures realistic storm variability and extreme flux events through validation of the reanalysis against direct observations (Bond and Cronin 2008; Harnik and Chang 2003; Qiu et al. 2004). To further assure that the NCEP–NCAR reanalysis surface turbulent heat fluxes are sufficiently realistic, we compared the daily reanalysis flux time series to those derived from measurements at the mooring sites of the Kuroshio Extension System Study (KESS; http://uskess.whoi.edu) and the CLIVAR Mode Water Dynamic Experiment (CLIMODE; http://www.climode.org) for the overlapping periods (not shown). The results reveal a correlation between the two turbulent heat flux time series with r = 0.91 in KER and r = 0.83 in GSR, respectively, suggesting that the reanalysis fluxes are adequate to carry out the extreme flux events analysis.

We focus on the examination of boreal winter season [November–March (NDJFM)] extreme flux events in the KER and GSR. The extreme flux events are defined using an area-averaged heat flux index over a rectangular area of 30° × 10° along the KER and GSR as shown in Fig. 1, where the largest winter mean surface latent and sensible heat fluxes occur. We note the area in the Atlantic is different from the so-called CLIMODE region described in SH10. We chose these regions because they are aligned more closely with the most significant THF variability, and thus likely include all extreme flux events in the North Pacific and Atlantic. In SH10, extreme flux events are defined by area-averaged daily sensible/latent heat flux, which exceeds the 80th percentile value (a chosen threshold) of the entire winter sensible/latent heat flux variability. The days when extreme flux events occur are referred to as event days and the rest of the winter days are referred to as non-event days. They also showed that the basic results of the study are not sensitive to the chosen threshold. Our own analysis confirms this finding and shows that the results in both the KER and GSR are not particularly sensitive to either the choice of the analysis regions or the threshold value. Therefore, following SH10, we chose the 80th percentile as a threshold (hereafter referred as HF80) to define extreme flux events in the KER and GSR for all the datasets used in this study. Figure 2 illustrates the selection criterion for the extreme events and compares their statistical properties in the KER and GSR, which will be further discussed in the next section.

Fig. 1.
Fig. 1.

Climatological winter season (NDJFM) mean THF in the (a) KER and (b) GSR based on 61-yr NCEP–NCAR reanalysis (W m−2). Area used to define extreme flux events in the KER and GSR are indicated by the rectangles.

Citation: Journal of Climate 28, 12; 10.1175/JCLI-D-14-00642.1

Fig. 2.
Fig. 2.

(a) The 61-yr winter season (NDJFM) NCEP–NCAR reanalysis daily THF (black solid), the standard deviation (gray shaded), and the HF80 critical value chosen to define extreme flux events (black dashed) for the KER at top and GSR at bottom. (b) Histogram (number of winters) of number of event days in each winter during 1948–2009, where the number of event days is obtained by counting the days when the area-averaged daily THF exceeds HF80. (c) Histogram (number of winters) of fraction of event-day THF contribution to the total THF in each winter, where the contribution is calculated by integrating the THF on event days divided by the total THF accumulated in all winter days. (d) Histogram (number of extreme flux events) of extreme flux event durations in days, where the extreme flux event duration is defined as the number of continuous days when the THF is above HF80. In all the histogram plots, black is for the GSR in the Atlantic and gray is for the KER in the Pacific.

Citation: Journal of Climate 28, 12; 10.1175/JCLI-D-14-00642.1

To identify storms responsible for the extreme flux events, we first applied a 1–8-day bandpass filter to the 6-hourly reanalysis data, so that synoptic-scale storm activity is isolated from large-scale, low-frequency atmospheric variability. We then divide the bandpass filtered data into two groups according to event days and non-event days selected using the HF80 threshold. Finally, the event-day and non-event-day data are averaged, respectively, to form a composite for each group. Only event-day composites for various bandpass filtered variables are shown in Fig. 3 since the non-event-day composite reveals nearly exact opposite patterns to those of the cold storms by construction, as discussed in the following section.

Fig. 3.
Fig. 3.

Event-day composites derived by averaging 1–8-day bandpass filtered THF (W m−2; color shaded and contours, positive being THF into the atmosphere) in the (a) KER and (b) GSR; 2-m air temperature (contours; °C) and specific humidity (color shaded; kg kg−1) in the (c) KER and (d) GSR; and SLP (color shaded; hPa) and 10-m wind (vector; m s−1) in the (e) KER and (f) GSR. All the variables were derived from the NCEP–NCAR reanalysis (1948–2009).

Citation: Journal of Climate 28, 12; 10.1175/JCLI-D-14-00642.1

Finally, to explore relationships between extreme flux events and modes of large-scale atmospheric variability, we performed empirical orthogonal function (EOF) analysis and singular value decomposition (SVD) analysis on various relevant variables. SVD analysis is effective in identifying covarying oceanic and atmospheric patterns (Bretherton et al. 1992). Before EOF and SVD are applied, the anomalies of relevant variables are obtained by removing the climatological annual cycle, weighted by the square root of the cosine of latitude and detrended at each grid point. The covariance field of two related variables is computed when carrying out SVD analysis. Additionally, we carried out lag-regression analyses to examine relationships between SST and extreme flux events. These analyses are based on seasonal mean anomalies and the results will be discussed in section 4.

3. Event-day and non-event-day fluxes and the associated storms

a. Characteristics of extreme flux events

Figure 2a shows the NDJFM daily climatology of THF and the associated standard deviation for the KER (top graph) and GSR (bottom graph), respectively, based on the 61 years of NCEP–NCAR reanalysis. Similar results are obtained using other reanalysis datasets (i.e., NCEP CFSR and 20CRV2), as well as the OAFlux dataset (not shown), indicating the robustness of the finding. The THF climatology value is considerably higher in the KER (~250 W m−2) than in the GSR (~220 W m−2) and displays a stronger seasonal variation in the Pacific than in the Atlantic, consistent with the stronger seasonal variability of the Kuroshio path than the Gulf Stream path (not shown). Despite the larger THF climatological value in the KER, its standard deviation is smaller than that of the GSR, suggesting that the storm activity in the GSR has stronger variability than in KER. This may be attributed to the difference in land–ocean distribution between the KER and GSR, with the latter being closer to the continent and thus experiencing stronger CAOs. Figure 2a also shows that the daily THF climatologies in both regions are below the HF80 threshold value, indicating that the HF80 is an objectively justifiable criterion to define the extreme flux events.

The histograms shown in Fig. 2 summarize the general characteristics of the extreme flux events. The average number of extreme flux event days (Fig. 2b) for the 61 winters in NCEP–NCAR reanalysis is about 30 for each winter in both the Pacific and Atlantic, which accounts for only 20% of the entire winter (NDJFM) days, whereas the average contribution of extreme flux events to the total THF (Fig. 2c) reaches 31% for the Pacific and 33% for the Atlantic. Extreme flux events in the KER have a broader range of variability with higher median value in terms of occurrences and contribution to the total THF in each winter than in the GSR (Figs. 2b,c). It is also clear from Fig. 2d that the majority of the extreme flux events last less than 3 days with 85% (88%) of extreme flux events in the KER (GSR) having a life cycle shorter than 3 days. The average duration of the less-than-3-day extreme flux events in the KER (GSR) is about 1.75 (1.63) days. It is worth noting that almost all the extreme flux events identified in the 61-yr NCEP–NCAR reanalysis have a life span of less than 8 days in both the KER and GSR, consistent with the notion that these extreme flux events are closely associated with synoptic-scale storms. This justifies the use of a 1–8-day bandpass filter before composite analysis is performed, as described in the previous section. Our findings are also consistent with the results of SH10 and support the conclusion that a relatively small number of extreme flux events make a significant contribution to the total surface turbulent heat exchange in the winter season in the WBC regimes and that these extreme flux events are closely related to synoptic-scale atmospheric variability.

b. Synoptic storms associated with extreme flux events

To further test the idea that the extreme flux events are closely related to CAOs, we performed a composite analysis using atmospheric surface variables, including THF, surface air temperature, specific humidity, sea level pressure (SLP), and surface winds. Different from the composite analysis in SH10, which was based on the unfiltered daily variables, we first applied a 1–8-day bandpass filter to the data to isolate synoptic-scale storms from large-scale atmospheric flows and then performed composite analyses on the filtered data as described in section 2. In this way we hope to build a direct link between synoptic-scale storms to the extreme flux events.

The event-day composite of the filtered data exhibits a clear wave pattern with an intense heat loss near the KER and GSR regions (Figs. 3a,b). The strong heat loss is accompanied by a cold and dry air mass anomaly in the regions during event days (Figs. 3c,d). The event-day composite of SLP and surface winds shows a dipole-like pattern centered near the KER and GSR regions with an anomalous cyclone to the east of the regions and an anticyclones to the west, in between which are strong northwesterlies (Figs. 3e,f). The strongest winds collocate with the largest pressure gradient resulting from geostrophy. Overall, the resultant patterns bear a close resemblance to the classic patterns of CAOs in both ocean basins (Colucci and Davenport 1987; Wheeler et al. 2011).

However, there are obvious differences between the CAOs in the Pacific and Atlantic. For the Pacific CAOs, the cold and dry surface air carried by the northwesterlies originating from the northeastern part of the Eurasian continent first encounters the relatively warm ocean water in the marginal seas west of Japan before reaching the KER. The air–sea exchange over the marginal seas causes latent and sensible heat loss from the ocean to the atmosphere and reduces the strength of the Pacific CAOs over the KER. In contrast, the Atlantic CAOs directly carry cold and dry Canadian air mass to the warm Gulf Stream, resulting in a more intense air–sea exchange over the GSR than in the KER. As a result, the Atlantic CAOs have higher intensity than its Pacific counterparts, as shown in Figs. 3a and 3b. We refer to the CAOs as “cold storms” in the following discussion. Additionally, in the Pacific the most significant THF loss occurs along the Kuroshio Extension after the separation of the Kuroshio from Japan, whereas the THF loss in the Atlantic is tilted along a southwest–northeast direction with the maximum heat loss located farther north, closer to the continent. These THF differences are closely linked to the differences in the underlying WBCs (Kelly et al. 2010). The Kuroshio flows roughly parallel along 35°N after it encounters the Oyashio current from the north, whereas the Gulf Stream, after its separation from Cape Hatteras, flows northeastward. We note that these results from the composite analysis are not sensitive to the region where the extreme flux time series is computed.

The non-event-day composite reveals nearly exact opposite patterns to those of the cold storms by construction, albeit with a much reduced amplitude (not shown). In particular, southerly winds prevail over the region of maximum THF anomaly with a low SLP anomaly to west and a high SLP anomaly to the east. Instead of transporting the cold and dry continental air over the warm currents, the southerly winds bring the warm and moist air from the subtropics to the regions, reducing the latent and sensible heat release from the warm currents. We refer to these synoptic disturbances as “non-event days.”

c. Variability of extreme flux events

To determine the contribution from the extreme flux events to the total winter mean THF variability in the KER and GSR, we divided the total THF variability into two components: 1) one induced by extreme flux events and 2) the other induced by nonextreme flux events. The former, referred to as event-day THF, was computed by integrating sensible and latent heat fluxes during event days in each winter season, while the latter, non-event-day THF, was computed by integrating sensible and latent heat fluxes during non-event days. Obviously, the sum of the event-day and non-event-day THF gives the total THF in each winter.

Figures 4a and 4b show the total, non-event-day, and event-day THF over the KER and GSR derived from the three different reanalysis datasets. Overall, the three datasets show reasonable agreement in the overlapping periods. In particular, the correlation between NCEP CFSR and NCEP–NCAR reanalysis THF is above 0.9 for both the KER and GSR over the 31-yr period from 1979 to 2009. The correlation between 20CRV2 and NCEP–NCAR reanalysis over the 62-yr period from 1948 to 2009 is about 0.7, demonstrating that the interannual and long-term time scale variability of the KER and GSR THF is well captured by all three reanalysis products. An examination of the non-event-day and event-day THF clearly indicates that the latter contains much higher variance than the former. The standard deviation for the event-day THF is 4.0 × 108 and 3.3 × 108 J m−2 for the KER and GSR, respectively, compared to 2.0 × 108 and 1.7 × 108 J m−2 for the non-event-day THF based on NCEP–NCAR reanalysis data. Furthermore, the event-day THF is highly correlated with the total THF with a correlation value of approximately 0.9 for both the KER and GSR, although part of this high correlation could be explained as a property of the non-Gaussian distribution of THF. The percentage of total THF variance explained by the event-day THF is 73%, 79%, and 83% for NCEP CFSR, NCEP–NCAR reanalysis, and 20CRV2 data, respectively, in the KER. The corresponding values for the GSR are 85%, 75%, and 70%, respectively. In contrast, the correlation between the total THF and non-event-day THF is much lower (~0.3) for all the reanalysis data in both regions, which is statistically insignificant at a 95% confidence level based on a Student’s t test. These results strongly suggest that CAO-induced extreme flux events practically determine the winter season turbulent heat exchange between the atmosphere and ocean in both the KER and GSR. The conclusion is entirely consistent with the finding of SH10 based on NCEP–NCAR reanalysis, pointing to the critical role of synoptic-scale storms in the climate system in the WBC regimes.

Fig. 4.
Fig. 4.

Total boreal winter THF (top plot), non-event-day THF (middle plot), and event-day THF (bottom plot) for 20CRV2 (dashed), NCEP–NCAR reanalysis (black solid), and NCEP CFSR (gray solid) reanalysis in the (a) KER and (b) GSR. Also shown are normalized total THF (black) and cold storm–induced THF (gray) based on NCEP–NCAR reanalysis, in the (c) KER and (d) GSR. The cold storm–induced THF is derived by first regressing the 1–8-day bandpass filtered daily SLP onto the event day composite of SLP and then integrating THF over the days during which the regression coefficient exceeds 1.0.

Citation: Journal of Climate 28, 12; 10.1175/JCLI-D-14-00642.1

To further validate that it is the cold storms that are mainly responsible for interannual and longer time scale THF variability, we performed the following analysis to identify the occurrence of CAOs in the KER and GSR. First we projected the 1–8-day bandpass filtered daily SLP onto the event-day SLP composites (Figs. 3e,f) that exhibit cold storm structure. We then examined the regression time series and integrated the THF for the period when the regression time series is above 1.0 in each winter. Note that since the regression time series are dimensionless, the value of 1.0 should give a regressed pattern that is identical to the composite cold storm, assuming that the residual is negligibly small. Figure 4c and 4d show the normalized time series of the integrated THF in the KER and GSR for the past 61 winters based on NCEP–NCAR reanalysis, overlaid by the normalized total THF time series derived by the area-averaged method shown in Figs. 4a and 4b. As expected, the two time series are highly correlated with r = 0.6 in the KER and r = 0.7 in the GSR. This further verifies the finding that the cold storm activity essentially determines the year-to-year variability of the THF in the KER and GSR.

In summary, the analysis of the KER and GSR THF indicates that even though the non-event-day fluxes contribute more dominantly to the winter mean THF in these regions, the decadal and longer time scale variability of the THF is almost entirely determined by the event-day fluxes that are closely related to winter storm activity, particularly CAO activity, in the regions. Next, we investigate the relationship between event-day flux variability and modes of climate variability in the North Pacific and Atlantic.

4. Relationship between extreme flux events and modes of North Pacific and Atlantic variability

a. Relationship between extreme flux events and ALP/EAP

It has been well documented that the leading modes of variability in SLP are the Aleutian low pattern (ALP) that typically accompanies with the PDO defined by the first EOF of SST and the NPO in the North Pacific sector (Trenberth and Hurrell 1994; Mantua et al. 1997; Rogers 1981) and the NAO and the EAP in the North Atlantic sector (Barnston and Livezey 1987; van Loon and Rogers 1978). These modes of variability can be described in terms of EOF analysis on winter mean SLP anomalies (Chhak et al. 2009; Di Lorenzo et al. 2008; Hurrell and Deser 2010; Newman et al. 2003). The ALP and NPO are typically captured by the first and second EOF of winter mean SLP anomalies in the North Pacific sector (Di Lorenzo et al. 2008; Newman et al. 2003), while the NAO and EAP are represented by the two leading EOFs of the SLP anomalies in North Atlantic sector (Rogers 1990; Thompson and Wallace 1998). In this study, we focus on the ALP and the EAP, as we will demonstrate that these two modes of variability are closely linked to the event-day flux variability in the KER and GSR, respectively.

The leading EOF of the North Pacific winter SLP anomaly (Fig. 5c) captures the ALP in the eastern North Pacific. Figure 5d shows the EAP as the second EOF of the North Atlantic winter SLP anomaly. The ALP explains approximately 45% of winter SLP anomaly variance in the North Pacific between 20° and 70°N, while the EAP explains 20% of the North Atlantic winter SLP variance in the same latitudinal band. The corresponding principal component (PC) time series are shown in Figs. 5a and 5b. We use the standardized PCs (normalized by the corresponding standard deviation) to define an ALP and EAP index, respectively.

Fig. 5.
Fig. 5.

(a) The first principal component time series and (c) the associated EOF (shaded and contoured with a contour interval of 1 hPa) in the North Pacific derived from NDJFM SLP anomalies using NCEP–NCAR reanalysis. (b) The second principal component time series and (d) the associated EOF (shaded and contoured with a contour interval of 1 hPa) in the North Atlantic.

Citation: Journal of Climate 28, 12; 10.1175/JCLI-D-14-00642.1

To investigate the relationship between extreme flux event activity in the KER (GSR) and modes of climate variability in the North Pacific and Atlantic, we performed an SVD analysis between event-day THF and winter mean SLP anomalies in the North Pacific and Atlantic, respectively. The event-day THF anomalies are defined as accumulated THF anomalies exceeding HF80 threshold in each winter season. Figures 6a and 6b show spatial patterns of the first SVD mode of event-day THF and SLP in the North Pacific, which explains 26.93% of the total squared covariance. The resultant SLP SVD (Fig. 6b) depicts an intensified Aleutian low that resembles a typical SLP pattern in PDO positive phase (Fig. 5c). The corresponding time series of the first SLP SVD is nearly identical to the ALP index with a correlation coefficient between the two as high as 0.99, indicating that the first SVD mode reproduces PDO variability extremely well (Fig. 6d). The first THF SVD displays, as expected, significant loading in the KER where the extreme flux events were defined (Fig. 6a). The corresponding THF time series is highly correlated with the ALP index with a correlation coefficient of 0.59 and is even more highly correlated to the detrended extreme THF time series shown in Fig. 4a with a correlation coefficient of 0.92, indicating that the SVD depicts the covariability between the extreme THF events and ALP (Fig. 6d). In the following discussion, we will use the detrended extreme THF time series shown in Figs. 4a and 4b to gauge the interannual-to-decadal variability of the extreme flux event activity in the KER and GSR, respectively, and refer to these time series as the detrended extreme flux indices. The broad patterns of SLP and THF derived from the SVD analysis are consistent with the results shown by Miller et al. (1994), who studied the 1976/77 climate shift in the Pacific basin. To quantify the percentage of total winter THF variance explained by the first SVD mode, we linearly regressed the winter THF anomaly onto the THF SVD and found that over 40% of total THF variance can be explained by the first SVD mode in the region where extreme flux events are defined (Fig. 6c).

Fig. 6.
Fig. 6.

Spatial patterns of the first SVD mode (squared variance, 27%) between (a) event-day THF and (b) NDJFM SLP anomalies in the North Pacific derived from NCEP–NCAR reanalysis with zero contour shown in dashed. (c) Percentage of contribution from the first THF SVD mode to the total THF variance (shaded and contours). (d) Time series of the first SVD mode for event-day THF anomalies (gray solid) and SLP anomalies (dashed), overlaid with the ALP index (black solid).

Citation: Journal of Climate 28, 12; 10.1175/JCLI-D-14-00642.1

The first leading SVD mode between event-day THF and winter mean SLP anomalies in the North Atlantic, which explains 29.8% of the squared variance, captures the NAO. This mode of variability is not related to extreme THF in the GSR but is more related to the extreme flux events in the Labrador Sea region (not shown). Since the focus of this study is on extreme THF in KER and GSR, we concentrate the discussion on the second leading SVD mode that shows a close relationship to extreme THF in the GSR. Figures 7a and 7b show the second leading SVD mode between event-day THF and winter mean SLP anomalies in the North Atlantic, which explains 22.9% of the squared covariance. The SLP SVD (Fig. 7b) shows a nearly identical pattern to that of EAP (Fig. 5d) and the corresponding SVD time series is highly correlated with the EAP index at 0.99, indicating that the SVD captures the full spectrum of EAP variability (Fig. 7d). The THF SVD shows large loading in the GSR (Fig. 7a) and the THF time series is correlated with the EAP index at 0.75 and with the Atlantic detrended extreme flux indices at 0.93 (Fig. 7d). The total variance of THF explained by this second SVD is over 50%, although the center of action is downstream of the GSR (Fig. 7c). These results confirm the linkage between extreme flux events in the GSR and EAP variability. Therefore, ALP and EAP share a commonality that their variability is connected to the occurrence of extreme flux events in the corresponding WBC regions.

Fig. 7.
Fig. 7.

As in Fig. 6, but for the second SVD mode (squared variance, 23%) between event-day THF and NDJFM SLP anomalies in the North Atlantic.

Citation: Journal of Climate 28, 12; 10.1175/JCLI-D-14-00642.1

The above analysis reveals a close relationship between extreme flux events in the KER (GSR) and ALP (EAP) in the North Pacific (Atlantic). To shed some light on plausible causality, we selected all the positive ALP (EAP) years based on the ALP (EAP) index (Figs. 5a,b). A lag-composite was then performed using raw daily SLP anomalies on extreme flux event days during the positive ALP/EAP years based on the HF80 threshold. The left panels of Fig. 8 show the evolution of the composite SLP anomalies starting from the onset of the extreme event (lag 0; Fig. 8a) to the subsequent 3 days (lag 1, 2, and 3; Figs. 8c,e,g) in the Pacific. A typical extreme flux event originating in the KER is accompanied by a weak high SLP anomaly to the west of the extreme THF anomaly and a strong low SLP anomaly to the east, consistent with the CAO storm patterns shown in Figs. 3e and 3f. As the storm develops, the SLP anomalies, particularly the low SLP anomalies, move eastward while decreasing in strength, leaving negative SLP anomalies deposited on the eastern side of the basin, which contribute to the intensified Aleutian low for the positive ALP phase (Fig. 5c). Similar evolution patterns are also found in the Atlantic sector (Fig. 8, right panels). The negative SLP anomalies associated with extreme flux events in the GSR propagate downstream, contributing to the low pressure anomaly associated with the EAP (Fig. 5d). In fact, if we simply sum up the SLP anomalies from lag 0 to lag 6 in the Pacific and Atlantic [the total number of these storm composite days (from lag 0 to 6) accounts for 23% and 28% of the entire winter days during PDO positive phase in the Pacific and EAP positive phase in the Atlantic, respectively] and divide by the count of all winter days, the resultant SLP anomalies bear a striking resemblance to the winter mean SLP anomalies during the positive phase of the ALP and EAP, respectively (Figs. 9a–d). On the other hand, the SLP anomalies averaged for the rest of the winter days in these years reveal patterns that are nearly orthogonal to the ALP and EAP SLP anomalies (Figs. 9e,f). These results suggest that the SLP anomalies during positive ALP and EAP phases may be explained, to a large extent, as an accumulated effect of the synoptic CAO storms in the North Pacific and Atlantic. This explanation is in line with the previous studies that emphasize the importance of synoptic scale storms in changing the low frequency mode of climate variability (e.g., Benedict et al. 2004; Feldstein 2003; Rivière and Orlanski 2007; Vallis and Gerber 2008).

Fig. 8.
Fig. 8.

Lag composites of daily SLP anomalies (contours, hPa) in the (left) KER and (right) GSR on event days during ALP positive phase derived from 61 winter season (NDJFM) NCEP–NCAR reanalysis. The event days were identified based on the HF80 threshold for daily THF in each NDJFM. (a),(b) Lag = 0 corresponds to the occurrence of extreme events and lag = (c),(d) 1, (e),(f) 2, and (g),(h) 3 correspond to the subsequent 3 days.

Citation: Journal of Climate 28, 12; 10.1175/JCLI-D-14-00642.1

Fig. 9.
Fig. 9.

Winter season (NDJFM) mean SLP anomalies (contours, hPa) during (a) ALP and (b) EAP positive phase in the North Pacific and Atlantic, respectively, derived from NCEP–NCAR reanalysis. Composites of event-day SLP anomalies by integrating from lag 0 to lag 6 day during the positive phase of (c) ALP and (d) EAP. Composites of non-event-day SLP anomalies by integrating over the rest of days during positive phase of (e) ALP and (f) EAP. The sum of the event-day composite in (c),(d) and non-event-day composite in (e),(f) gives back the winter season mean SLP anomalies in (a),(b).

Citation: Journal of Climate 28, 12; 10.1175/JCLI-D-14-00642.1

b. Relationship between extreme flux events and storm tracks

The next question is then to determine how extreme flux event variability in the KER and GSR relates to Pacific and Atlantic storm track variability, respectively. To examine this question, we applied SVD analysis to the winter season event-day THF anomaly and 1–8-day bandpass filtered 300-hPa meridional wind (υ) product, that is, 〈υυ′〉, which gives a measure of storm tracks. The leading SVD and the corresponding time series for the North Pacific sector are shown in Fig. 10 (left panels). The THF SVD (Fig. 10e) again shows a large loading of THF signal in the KER similar to that shown in Fig. 6a. The covarying storm track pattern (Fig. 10c) shows a dipole-like structure with a positive anomaly in the southeastern North Pacific and a negative anomaly in the northern part of North Pacific, suggesting a southward shift of the upper-level storm track. The most significant change in the upper-level storm track occurs downstream of the KER in the southeastern North Pacific basin close to the North American continent. A similar pattern of Pacific storm track changes is obtained when the SVD analysis was applied to 〈υυ′〉 and winter mean SLP anomalies over the North Pacific sector, as shown by Fig. 10 (right panels) in which the leading SLP SVD bears a close resemblance to the ALP (Fig. 5c). The two leading SVD time series of the storm track variability are also highly correlated with each other with a correlation coefficient of 0.92. Furthermore, the SLP and THF SVD time series are correlated with the ALP index and the Pacific detrended extreme flux index at 0.99 and 0.94, respectively. These results point to a close relationship among the extreme flux events in the KER, the upper-level storm track in the North Pacific, and the ALP. A positive ALP corresponds to enhanced occurrence of extreme flux events in the KER and a southward shift of the upper-level Pacific storm track. Opposite relationships tend to occur during a negative ALP phase.

Fig. 10.
Fig. 10.

First SVD mode (left) between NDJFM event-day THF and 300-hPa 〈υυ′〉 (squared variance, 15%) and (right) between NDJFM SLP anomalies and 300-hPa 〈υυ′〉 (squared variance, 34%) in the North Pacific derived from NCEP–NCAR reanalysis. SVD time series are shown with (a) event-day THF and (b) SLP anomalies in gray and 300-hPa 〈υυ′〉 in black. (c),(d) SVD spatial patterns of 300-hPa 〈υυ′〉 (shaded and contours). SVD spatial patterns (e) of event-day THF (shaded and contours) and (f) of SLP anomalies (shaded and contours). For all the spatial patterns, positive values are plotted with solid contours and negative values with dashed contours.

Citation: Journal of Climate 28, 12; 10.1175/JCLI-D-14-00642.1

Similar SVD analyses were performed for the North Atlantic sector. Figure 11 shows the second leading SVD mode between 〈υυ′〉 and event-day THF anomalies and between 〈υυ′〉 and winter mean SLP anomalies over the North Atlantic sector. In comparison with the Pacific counterparts, the explained squared covariance by these second SVDs is lower and in the range of 15%–20%. The two resultant SVDs for the upper storm track show similar structure, consisting of an enhanced storm activity over the eastern North Atlantic basin and reduced storm activity in the GSR and Greenland–Iceland–Norwegian seas region, corresponding to enhanced extreme flux events in the GSR and a positive phase of the EAP. The two SVD time series of storm track are highly correlated at 0.96, suggesting that they represent the same mode of variability. Given that the THF and SLP SVD time series are significantly correlated with the GSR detrended extreme flux index and the EAP index at 0.84 and 0.94, respectively, the resultant two SVD pairs suggest that the extreme flux event activity in the GSR, the Atlantic storm track, and the EAP variability are closely related.

Fig. 11.
Fig. 11.

As in Fig. 10, but for the second SVD mode (left) between NDJFM event-day THF and 300-hPa 〈υυ′〉 (squared variance, 15%) and (right) between NDJFM SLP anomalies and 300-hPa 〈υυ′〉 (squared variance, 20%) in the North Atlantic.

Citation: Journal of Climate 28, 12; 10.1175/JCLI-D-14-00642.1

c. Event-day and non-event-day storm paths

To probe deeper into the relationship between synoptic storm activity in the KER and GSR and North Pacific and Atlantic storm tracks, we performed the following composite analysis. First, winter season storms were divided into two groups: event-day storms and non-event-day storms according to the HF80 criterion, and then lag composites of 1–8-day bandpass filtered 300-hPa geopotential height field were made for the two groups of storm events using the respective THF time series in the KER and GSR from 4 days before to 4 days after the maximum THF. Figure 12 compares the evolution of the two groups of storms in the North Pacific sector from lag −1 day to lag 2 day. Originating from the northeastern Eurasian continent, both groups of storms gradually grow and reach their peak around lag 0–1 day when passing through the KER. There is a clear asymmetry in storm strength about lag 0 day with larger storm amplitude at lag 1 day than at lag −1 day, suggesting that the storms gain strength as they pass through the KER. A visual inspection also indicates a seemingly different storm path between the event-day and non-event-day storm composites with the former having a more southward path than the latter, particularly in the eastern half of the basin. This difference can be seen more clearly by tracking the centers of the low-pressure systems that pass through the Kuroshio at lag 0 day (Figs. 12c,d). Figures 12i and 12j show trajectories of the low pressure center movement from lag −3 day to lag 4 day, superimposed on the low pressure composites at each of the lags. It is apparent that the event-day storms tend to follow a zonal path, whereas the non-event-day storms prefer a northeastward path. However, large uncertainties exist in tracking storm path and the significance of these storm path differences needs to be further scrutinized. Nevertheless, since the event-day (non-event-day) storms are associated with stronger (weaker) THF near the KER, which correspond to a southward (northward) shift of the Pacific storm track as shown by the SVD analysis (Fig. 10), the southeastward (northeastward) path of the event-day (non-event-day) storms is consistent with the storm track change during a positive (negative) ALP phase, suggesting that the shift of the storm track may be attributed to different downstream development for event-day and non-event-day storms.

Fig. 12.
Fig. 12.

Lag composite of 1–8-day bandpass filtered 300-hPa geopotential height for (left) event-day storms and (right) non-event-day storms in the North Pacific at (a),(b) lag −1, (c),(d) lag 0, (e),(f) lag 1, and (g),(h) lag 2. (i),(j) Trajectories of the low pressure center movement from lag −3 day to lag 4 day, superimposed on the low pressure composites at each of the lags. Contour intervals are 10 m for event-day composites and 2 m for non-event-day composites. Event-day storms and non-event-day storms were derived from 61 winter season (NDJFM) NCEP–NCAR reanalysis based on the HF80 threshold.

Citation: Journal of Climate 28, 12; 10.1175/JCLI-D-14-00642.1

Figure 13 illustrates similar storm composites and paths for the North Atlantic sector. It again shows that the event-day (non-event-day) storms prefer a southeastward (northeastward) downstream propagation path, consistent with the increase (decrease) in storm track activity in the eastern North Atlantic and decrease (increase) in storm track activity in the Greenland–Iceland–Norwegian seas region. Different from the storm growth in the KER, both the event-day and non-event-day storms are well developed at lag −1 day and reach their peaks at lag 0 day as they pass the GSR and then gradually decay.

Fig. 13.
Fig. 13.

As in Fig. 12, but for (left) event-day and (right) non-event-day storms in the North Atlantic.

Citation: Journal of Climate 28, 12; 10.1175/JCLI-D-14-00642.1

In summary, the above analyses suggest a cohesive interrelationship among extreme flux events induced by CAOs in the WBC regimes, storm track activity, and basin-scale modes of climate variability in both the North Pacific and Atlantic. Since the extreme flux events play a vital role in determining local air–sea exchange in the KER and GSR as shown in the previous section, it is natural to ask whether the variability of the extreme flux events is affected by SST variability along the Kuroshio and the Gulf Stream. The following section addresses this question.

5. Relationship between extreme flux events and SST

We performed lag-regression analyses between the extreme flux event time series and SST anomalies in the North Pacific and Atlantic. Here, the extreme flux event time series are derived using the HF80 threshold for each month of NDJFM in each winter season and SST anomalies are calculated in corresponding months leading/lagging the flux certain time. For example, lag 1 month means that SST anomalies were averaged over each month of DJFMA. There is no overlapping period between the flux and SST time series.

Figures 14a and 14c show regressed seasonal mean SST anomalies onto the extreme flux event time series at lag −1 and lag 1 month over the North Pacific. When the SST leads the extreme flux events by one month (i.e., lag −1), there are statistically significant positive regression values forming along the Kuroshio. The positive regression values along the Kuroshio persist from lag −3 month to −1 month (not shown), suggesting that the SST anomalies along the Kuroshio may help fuel the development of extreme flux events in the region. Moving downstream of the Kuroshio, a broad area of statistically significant negative regression values exists in the central and eastern North Pacific, suggesting a possible negative feedback between the downstream SST and extreme flux event variability. When the SST lags the extreme flux events by one month (i.e., lag 1), the significant positive regressions along the Kuroshio fade away, while the negative regressions downstream increase, suggesting that the downstream development of the extreme flux events tends to cool the SST in the central and eastern North Pacific. The negative regression values also dominate for lag 2 and lag 3 months (not shown). A 3-month period is a typical time scale for the ocean mixed layer to respond to changes in the atmospheric surface fluxes. A recent study by Frankignoul et al. (2011) also reported positive atmospheric response to the sharp SST front in the KER. Therefore, these results are consistent with the notion that the sharp SST front along the Kuroshio Extension may interact positively with the development of synoptic atmospheric systems in the region during the boreal winter, while the SST variability in the central and eastern North Pacific is largely forced by the atmosphere.

Fig. 14.
Fig. 14.

Seasonal mean SST (°C) anomalies regressed onto extreme flux event index in the (left) KER in the North Pacific and (right) GSR in the North Atlantic when SST anomalies (a),(b) lead extreme flux events by 1 month and (c),(d) SST lag by 1 month. Zero contours are in dashed and regression values significant at 95% confidence level based on a Student’s t test are shaded by black dots.

Citation: Journal of Climate 28, 12; 10.1175/JCLI-D-14-00642.1

Figures 14b and 14d show similar regression maps for the North Atlantic sector. As shown in the Pacific sector, statistically significant positive SST regressions are found along the GSR when SST lead extreme flux events by 1 month (lag −1) and persist in other months (lag −2 and lag −3; not shown), confirming the influence of local SST on the extreme flux events in the GSR. Also, significant negative SST anomalies in the GSR and positive SST anomalies south of the GSR are found when the SST lags extreme flux events by 1 month. This points to an oceanic response to large-scale atmospheric forcing through change of surface wind and turbulent heat fluxes, as noted by previous studies (e.g., Deser and Blackmon 1993).

In summary, a robust large-scale negative SST response is identified downstream of the KER (GSR) in the central (northern) North Pacific (Atlantic) basins. Also, significant positive regressions are revealed in the KER (GSR) when SST anomalies lead extreme flux events, suggesting a fueling effect from the underlying SST. It is worth pointing out that it is also possible that the positive regressions are due to the high persistence of SST anomalies in the frontal regions (Frankignoul et al. 2011). Further studies are needed to confirm the positive influence of SST in the KER (GSR).

6. Conclusions and discussion

Based on NCEP CFSR, NCEP–NCAR reanalysis, 20CRV2, and OAFlux datasets, extreme flux events associated with CAOs in boreal winter in the KER and GSR are analyzed and compared in terms of their statistics, spatial structure, and interannual-to-decadal variability. The relationship between extreme flux events in the KER (GSR) and modes of climate variability in the North Pacific (Atlantic) are also investigated, as well as the associated storm track response. Finally, potential influences from the underlying ocean on the extreme flux events are discussed.

Extreme flux events associated with synoptic-scale atmosphere storms dominate the total surface turbulent heat exchange between the atmosphere and ocean during boreal winter in both the KER and GSR. The statistical characteristics of the extreme flux events reveal that the average accumulated number of the extreme flux event days, which typically (>85%) last fewer than 3 days, occupies only 20% of the winter period, but the events contribute over 30% to the total THF during the entire winter season and explain more than 80% of the total variance of THF in the regions. The short duration of the extreme flux events suggests that they are associated with synoptic-scale atmospheric variability.

Composite analysis reveals that the extreme flux events are closely related to the winter season CAOs. The extreme flux events in the Pacific differ somewhat from those in the Atlantic in terms of magnitude, location, and orientation. The latter tend to have stronger intensity and shorter duration than the former.

SVD analyses show that the extreme flux events in the KER (GSR) are closely related to the ALP (EAP) in the North Pacific (Atlantic). Lag composite analyses suggest that both the ALP and EAP are linked to the evolution of CAOs. When extreme flux events occur in the KER (GSR), the associated storms form upstream of the ALP (EAP) and then propagate eastward, eventually evolving into patterns similar to SLP anomalies associated with the ALP and EAP. Indeed, the anomalous SLP patterns during the positive phase of the ALP and EAP can be practically reconstructed using only the SLP anomalies after the onset and during the decay phase of the CAO storms, suggesting that the ALP and EAP are connected to the accumulated effects of the synoptic winter storms accompanied with the extreme flux events. This explanation is consistent with the modeling results represented by Feldstein (2003), Benedict et al. (2004), Rivière and Orlanski (2007), and Vallis and Gerber (2008).

Composites of upper tropospheric geopotential height anomalies indicate that the storms associated with the extreme flux events tend to propagate southeastward compared to non-event-day storms downstream propagation path, consistent with the negative SLP anomalies corresponding to a positive ALP/EAP phase. The extreme flux events over the GSR are closely linked to the EAP instead of the NAO, although the NAO is the first mode of climate variability in the North Atlantic. In fact, we found that it is the extreme flux events in the Labrador Sea region that are related to the NAO. A detailed discussion of this relationship is beyond the scope of this paper.

Finally, lag-regression analyses of SST anomalies and extreme flux event index show some evidence of a positive SST influence on the occurrence of extreme flux events in the KER/GSR, suggesting that the variability of extreme flux events may have some influences from the underlying SST along the KER/GSR. The analyses also present unequivocal evidence for atmospheric forcing of ocean in the central basin of the North Pacific and Atlantic when the extreme flux events lead SST.

Based on the results presented in this study, we hypothesize that extreme winter storm activity in the KER and GSR is crucial to understanding the ALP (and PDO) in the North Pacific and the EAP in the North Atlantic. To a large extent, the low SLP anomalies occurring during positive phase of the ALP and EAP may be considered to be a cumulative effect of the extreme storm variability in the KER and GSR. The North Pacific and Atlantic storm track variability can also be directly linked to the extreme storm variability in the KER and GSR. Further studies are needed to understand the potential role of frontal-scale and mesoscale air–sea interactions along the KER and GSR in the development of extreme flux events and associated storm systems, as it is possible that SST impacts on the ALP and EAP may be achieved through its effects on the extreme heat flux variability in the KER and GSR.

Acknowledgments

This research is partially supported by the China National Global Change Major Research Project 2013CB956204; U.S. National Science Foundation Grants AGS-1067937, AGS-1462127, and AGS-1347808; Department of Energy Grants DE-SC0004966 and DE-SC0006824; and National Oceanic and Atmospheric Administration Grant NA11OAR4310154. We also acknowledge the support from the China National Program on Key Basic Research Project (973 Program) 2014CB745000. The Texas Advanced Computing Center (TACC) at The University of Texas at Austin and the Texas A&M Supercomputing Facility provided high performance computing resources that contributed to the research results reported in this paper.

REFERENCES

  • Alexander, M. A., , and J. D. Scott, 1997: Surface flux variability over the North Pacific and North Atlantic Oceans. J. Climate, 10, 29632978, doi:10.1175/1520-0442(1997)010<2963:SFVOTN>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Barnston, A. G., , and R. E. Livezey, 1987: Classification, seasonality, and persistence of low-frequency atmospheric circulation patterns. Mon. Wea. Rev., 115, 10831126, doi:10.1175/1520-0493(1987)115<1083:CSAPOL>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Barsugli, J. J., , and D. S. Battisti, 1998: The basic effects of atmosphere–ocean thermal coupling on midlatitude variability. J. Atmos. Sci., 55, 477493, doi:10.1175/1520-0469(1998)055<0477:TBEOAO>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Benedict, J. J., , S. Lee, , and S. B. Feldstein, 2004: Synoptic view of the North Atlantic oscillation. J. Atmos. Sci., 61, 121144, doi:10.1175/1520-0469(2004)061<0121:SVOTNA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Bond, N. A., , and M. F. Cronin, 2008: Regional weather patterns during anomalous air–sea fluxes at the Kuroshio Extension Observatory (KEO). J. Climate, 21, 16801697, doi:10.1175/2007JCLI1797.1.

    • Search Google Scholar
    • Export Citation
  • Bretherton, C. S., , C. Smith, , and J. M. Wallace, 1992: An intercomparison of methods for finding coupled patterns in climate data. J. Climate, 5, 541560, doi:10.1175/1520-0442(1992)005<0541:AIOMFF>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Cayan, D. R., , and D. H. Peterson, 1989: The influence of North Pacific atmospheric circulation on streamflow in the west. Aspects of Climate Variability in the Pacific and the Western Americas, Geophys. Monogr., Vol. 35, Amer. Geophys. Union, 375–397.

  • Chang, E. K. M., 2009: Are band-pass variance statistics useful measures of storm track activity? Re-examining storm track variability associated with the NAO using multiple storm track measures. Climate Dyn., 33, 277296, doi:10.1007/s00382-009-0532-9.

    • Search Google Scholar
    • Export Citation
  • Chelton, D. B., , and S.-P. Xie, 2010: Coupled ocean–atmosphere interaction at oceanic mesoscales. Oceanography, 23, 5269, doi:10.5670/oceanog.2010.05.

    • Search Google Scholar
    • Export Citation
  • Chelton, D. B., and et al. , 2001: Observations of coupling between surface wind stress and sea surface temperature in the eastern tropical Pacific. J. Climate, 14, 14791498, doi:10.1175/1520-0442(2001)014<1479:OOCBSW>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Chelton, D. B., , M. G. Schlax, , M. H. Freilich, , and R. F. Milliff, 2004: Satellite measurements reveal persistent small-scale features in ocean winds. Science, 303, 978983, doi:10.1126/science.1091901.

    • Search Google Scholar
    • Export Citation
  • Chhak, K. C., , E. Di Lorenzo, , N. Schneider, , and P. F. Cummins, 2009: Forcing of low-frequency ocean variability in the northeast Pacific. J. Climate, 22, 12551276, doi:10.1175/2008JCLI2639.1.

    • Search Google Scholar
    • Export Citation
  • Colucci, S. J., , and J. C. Davenport, 1987: Rapid surface anticyclogenesis: Synoptic climatology and attendant large-scale circulation changes. Mon. Wea. Rev., 115, 822836, doi:10.1175/1520-0493(1987)115<0822:RSASCA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Compo, G. P., and et al. , 2011: The Twentieth Century Reanalysis Project. Quart. J. Roy. Meteor. Soc., 137, 128, doi:10.1002/qj.776.

    • Search Google Scholar
    • Export Citation
  • Deser, C., , and M. L. Blackmon, 1993: Surface climate variations over the North Atlantic Ocean during winter—1900–1989. J. Climate, 6, 17431753, doi:10.1175/1520-0442(1993)006<1743:SCVOTN>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Di Lorenzo, E., and et al. , 2008: North Pacific Gyre Oscillation links ocean climate and ecosystem change. Geophys. Res. Lett., 35, L08607, doi:10.1029/2007GL032838.

    • Search Google Scholar
    • Export Citation
  • Feldstein, S. B., 2003: The dynamics of NAO teleconnection pattern growth and decay. Quart. J. Roy. Meteor. Soc., 129, 901924, doi:10.1256/qj.02.76.

    • Search Google Scholar
    • Export Citation
  • Frankignoul, C., , and K. Hasselmann, 1977: Stochastic climate models. Part II: Application to sea-surface temperature anomalies and thermocline variability. Tellus, 29, 289305, doi:10.1111/j.2153-3490.1977.tb00740.x.

    • Search Google Scholar
    • Export Citation
  • Frankignoul, C., , N. Sennechael, , Y. O. Kwon, , and M. A. Alexander, 2011: Influence of the meridional shifts of the Kuroshio and the Oyashio Extensions on the atmospheric circulation. J. Climate, 24, 762777, doi:10.1175/2010JCLI3731.1.

    • Search Google Scholar
    • Export Citation
  • Harnik, N., , and E. K. M. Chang, 2003: Storm track variations as seen in radiosonde observations and reanalysis data. J. Climate, 16, 480495, doi:10.1175/1520-0442(2003)016<0480:STVASI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hasselmann, K., 1976: Stochastic climate models: Part I. Theory. Tellus, 28, 473485, doi:10.1111/j.2153-3490.1976.tb00696.x.

  • Hoskins, B. J., , and P. J. Valdes, 1990: On the existence of storm-tracks. J. Atmos. Sci., 47, 18541864, doi:10.1175/1520-0469(1990)047<1854:OTEOST>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Hurrell, J. W., , and H. VanLoon, 1997: Decadal variations in climate associated with the North Atlantic oscillation. Climatic Change, 36, 301326, doi:10.1023/A:1005314315270.

    • Search Google Scholar
    • Export Citation
  • Hurrell, J. W., , and C. Deser, 2010: North Atlantic climate variability: The role of the North Atlantic Oscillation. J. Mar. Syst., 79, 231244, doi:10.1016/j.jmarsys.2009.11.002.

    • Search Google Scholar
    • Export Citation
  • Hurrell, J. W., , Y. Kushnir, , G. Ottersen, , and M. Visbeck, 2003: An overview of the North Atlantic oscillation. The North Atlantic Oscillation: Climatic Significance and Environmental Impact, J. W. Hurrell et al., Eds., Amer. Geophys. Union, 1–33.

  • Kalnay, E., and et al. , 1996: The NCEP/NCAR 40-Year 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
  • Kelly, K. A., , R. J. Small, , R. M. Samelson, , B. Qiu, , T. M. Joyce, , Y.-O. Kwon, , and M. F. Cronin, 2010: Western boundary currents and frontal air–sea interaction: Gulf Stream and Kuroshio Extension. J. Climate,23, 5644–5667, doi:10.1175/2010JCLI3346.1.

  • Kushnir, Y., , W. A. Robinson, , I. Bladé, , N. M. J. 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
  • Kwon, Y.-O., , and T. M. Joyce, 2013: Northern Hemisphere winter atmospheric transient eddy heat fluxes and the Gulf Stream and Kuroshio–Oyashio Extension variability. J. Climate, 26, 98399859, doi:10.1175/JCLI-D-12-00647.1.

    • Search Google Scholar
    • Export Citation
  • Kwon, Y.-O., , M. A. Alexander, , N. A. Bond, , C. Frankignoul, , H. Nakamura, , B. Qiu, , and L. A. Thompson, 2010: Role of the Gulf Stream and Kuroshio–Oyashio systems in large-scale atmosphere–ocean interaction: A review. J. Climate,23, 3249–3281, doi:10.1175/2010JCLI3343.1.

  • Linkin, M. E., , and S. Nigam, 2008: The North Pacific Oscillation–west Pacific teleconnection pattern: Mature-phase structure and winter impacts. J. Climate, 21, 19791997, doi:10.1175/2007JCLI2048.1.

    • Search Google Scholar
    • Export Citation
  • Maidens, A., , A. Arribas, , A. A. Scaife, , C. MacLachlan, , D. Peterson, , and J. Knight, 2013: The influence of surface forcings on prediction of the North Atlantic Oscillation regime of winter 2010/11. Mon. Wea. Rev., 141, 38013813, doi:10.1175/MWR-D-13-00033.1.

    • Search Google Scholar
    • Export Citation
  • Maloney, E. D., , and D. B. Chelton, 2006: An assessment of the sea surface temperature influence on surface wind stress in numerical weather prediction and climate models. J. Climate, 19, 27432762, doi:10.1175/JCLI3728.1.

    • Search Google Scholar
    • Export Citation
  • Mantua, N. J., , S. R. Hare, , Y. Zhang, , J. M. Wallace, , and R. C. Francis, 1997: A Pacific interdecadal climate oscillation with impacts on salmon production. Bull. Amer. Meteor. Soc., 78, 10691079, doi:10.1175/1520-0477(1997)078<1069:APICOW>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Marshall, J., and et al. , 2001: North Atlantic climate variability: Phenomena, impacts and mechanisms. Int. J. Climatol., 21, 18631898, doi:10.1002/joc.693.

    • Search Google Scholar
    • Export Citation
  • Miller, A. J., , and N. Schneider, 2000: Interdecadal climate regime dynamics in the North Pacific Ocean: Theories, observations and ecosystem impacts. Prog. Oceanogr., 47, 355379, doi:10.1016/S0079-6611(00)00044-6.

    • Search Google Scholar
    • Export Citation
  • Miller, A. J., , D. R. Cayan, , T. P. Barnett, , N. E. Graham, , and J. M. Oberhuber, 1994: The 1976–77 climate shift of the Pacific Ocean. Oceanography, 7, 2126, doi:10.5670/oceanog.1994.11.

    • Search Google Scholar
    • Export Citation
  • Miller, A. J., , F. Chai, , S. Chiba, , J. R. Moisan, , and D. J. Neilson, 2004: Decadal-scale climate and ecosystem interactions in the North Pacific Ocean. J. Oceanogr., 60, 163188, doi:10.1023/B:JOCE.0000038325.36306.95.

    • Search Google Scholar
    • Export Citation
  • Nakamura, H., 1996: Year-to-year and interdecadal variability in the activity of intraseasonal fluctuations in the Northern Hemisphere wintertime circulation. Theor. Appl. Climatol., 55, 1932, doi:10.1007/BF00864700.

    • Search Google Scholar
    • Export Citation
  • Nakamura, H., , G. Lin, , and T. Yamagata, 1997: Decadal climate variability in the North Pacific during the recent decades. Bull. Amer. Meteor. Soc., 78, 22152225, doi:10.1175/1520-0477(1997)078<2215:DCVITN>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Nakamura, H., , T. Sampe, , Y. Tanimoto, , and A. Shimpo, 2004: Observed associations among storm tracks, jet streams and midlatitude oceanic fronts. Earth’s Climate: The Ocean–Atmosphere Interaction, Geophys. Mongr., Vol. 147, Amer. Geophys. Union, 329–345.

  • Nakamura, H., , T. Sampe, , A. Goto, , W. Ohfuchi, , and S.-P. Xie, 2008: On the importance of midlatitude oceanic frontal zones for the mean state and dominant variability in the tropospheric circulation. Geophys. Res. Lett., 35, L15709, doi:10.1029/2008GL034010.

    • Search Google Scholar
    • Export Citation
  • Nakamura, M., , and S. Yamane, 2009: Dominant anomaly patterns in the near-surface baroclinicity and accompanying anomalies in the atmosphere and oceans. Part I: North Atlantic basin. J. Climate, 22, 880904, doi:10.1175/2008JCLI2297.1.

    • Search Google Scholar
    • Export Citation
  • Nakamura, M., , and S. Yamane, 2010: Dominant anomaly patterns in the near-surface baroclinicity and accompanying anomalies in the atmosphere and oceans. Part II: North Pacific basin. J. Climate, 23, 64456467, doi:10.1175/2010JCLI3017.1.

    • Search Google Scholar
    • Export Citation
  • Newman, M., , G. P. Compo, , and M. A. Alexander, 2003: ENSO-forced variability of the Pacific decadal oscillation. J. Climate, 16, 38533857, doi:10.1175/1520-0442(2003)016<3853:EVOTPD>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • O’Neill, L. W., , D. B. Chelton, , and S. K. Esbensen, 2012: Covariability of surface wind and stress responses to sea surface temperature fronts. J. Climate,25, 5916–5942, doi:10.1175/JCLI-D-11-00230.1.

  • Qiu, B., , S. Chen, , and P. Hacker, 2004: Synoptic-scale air–sea flux forcing in the western North Pacific: Observations and their impact on SST and the mixed layer. J. Phys. Oceanogr., 34, 21482159, doi:10.1175/1520-0485(2004)034<2148:SAFFIT>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Rivière, G., , and I. Orlanski, 2007: Characteristics of the Atlantic storm-track eddy activity and its relation with the North Atlantic Oscillation. J. Atmos. Sci., 64, 241266, doi:10.1175/JAS3850.1.

    • Search Google Scholar
    • Export Citation
  • Rogers, J. C., 1981: The North Pacific oscillation. J. Climatol., 1, 3957, doi:10.1002/joc.3370010106.

  • Rogers, J. C., 1990: Patterns of low-frequency monthly sea level pressure variability (1899–1986) and associated wave cyclone frequencies. J. Climate, 3, 13641379, doi:10.1175/1520-0442(1990)003<1364:POLFMS>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Rogers, J. C., 1997: North Atlantic storm track variability and its association to the North Atlantic Oscillation and climate variability of northern Europe. J. Climate, 10, 16351647, doi:10.1175/1520-0442(1997)010<1635:NASTVA>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Saha, S., and et al. , 2010: The NCEP Climate Forecast System Reanalysis. Bull. Amer. Meteor. Soc., 91, 10151057, doi:10.1175/2010BAMS3001.1.

    • Search Google Scholar
    • Export Citation
  • Sampe, T., , and S.-P. Xie, 2007: Mapping high sea winds from space: A global climatology. Bull. Amer. Meteor. Soc., 88, 19651978, doi:10.1175/BAMS-88-12-1965.

    • Search Google Scholar
    • Export Citation
  • Saravanan, R., 1998: Atmospheric low-frequency variability and its relationship to midlatitude SST variability: Studies using the NCAR climate system model. J. Climate, 11, 13861404, doi:10.1175/1520-0442(1998)011<1386:ALFVAI>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Shaman, J., , R. M. Samelson, , and E. Skyllingstad, 2010: Air–sea fluxes over the Gulf Stream region: Atmospheric controls and trends. J. Climate, 23, 26512670, doi:10.1175/2010JCLI3269.1.

    • Search Google Scholar
    • Export Citation
  • Taguchi, B., , H. Nakamura, , M. Nonaka, , and S.-P. Xie, 2009: Influences of the Kuroshio/Oyashio Extensions on air–sea heat exchanges and storm-track activity as revealed in regional atmospheric model simulations for the 2003/04 cold season. J. Climate, 22, 65366560, doi:10.1175/2009JCLI2910.1.

    • Search Google Scholar
    • Export Citation
  • Taguchi, B., , H. Nakamura, , M. Nonaka, , N. Komori, , A. Kuwano-Yoshida, , K. Takaya, , and A. Goto, 2012: Seasonal evolutions of atmospheric response to decadal SST anomalies in the North Pacific subarctic frontal zone: Observations and a coupled model simulation. J. Climate,25, 111–139, doi:10.1175/JCLI-D-11-00046.1.

  • Thompson, D. W. J., , and J. M. Wallace, 1998: The Arctic Oscillation signature in the wintertime geopotential height and temperature fields. Geophys. Res. Lett., 25, 12971300, doi:10.1029/98GL00950.

    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., , and J. W. Hurrell, 1994: Decadal atmosphere–ocean variations in the Pacific. Climate Dyn., 9, 303319, doi:10.1007/BF00204745.

    • Search Google Scholar
    • Export Citation
  • Vallis, G. K., , and E. P. Gerber, 2008: Local and hemispheric dynamics of the North Atlantic Oscillation, annular patterns and the zonal index. Dyn. Atmos. Oceans, 44, 184212, doi:10.1016/j.dynatmoce.2007.04.003.

    • Search Google Scholar
    • Export Citation
  • van Loon, H., , and J. C. Rogers, 1978: The seesaw in winter temperatures between Greenland and northern Europe. Part I: General description. Mon. Wea. Rev., 106, 296310, doi:10.1175/1520-0493(1978)106<0296:TSIWTB>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Walsh, J. E., , A. S. Phillips, , D. H. Portis, , and W. L. Chapman, 2001: Extreme cold outbreaks in the United States and Europe, 1948–99. J. Climate, 14, 26422658, doi:10.1175/1520-0442(2001)014<2642:ECOITU>2.0.CO;2.

    • Search Google Scholar
    • Export Citation
  • Wheeler, D. D., , V. L. Harvey, , D. E. Atkinson, , R. L. Collins, , and M. J. Mills, 2011: A climatology of cold air outbreaks over North America: WACCM and ERA-40 comparison and analysis. J. Geophys. Res., 116, D12107, doi:10.1029/2011JD015711.

    • Search Google Scholar
    • Export Citation
  • Woollings, T., , B. Hoskins, , M. Blackburn, , and P. Berrisford, 2008: A new Rossby wave–breaking interpretation of the North Atlantic Oscillation. J. Atmos. Sci., 65, 609626, doi:10.1175/2007JAS2347.1.

    • Search Google Scholar
    • Export Citation
  • Yu, L., , X. Jin, , and R. A. Weller, 2008: Multidecadel global flux datasets from the Objectively Analyzed Air–Sea Fluxes (OAFlux) Project: Latent and sensible heat fluxes, ocean evaporation, and related surface meteorological variables. OAFlux Project Tech. Rep. OA-2008-01, Woods Hole Oceanographic Institution, 64 pp.

Save