• Alexander, M. A., , I. Bladé, , M. Newman, , J. R. Lazante, , N-C. Lau, , and J. D. Scott, 2002: The atmospheric bridge: The influence of ENSO on air–sea interaction over the global oceans. J. Climate, 15 , 22052231.

    • Search Google Scholar
    • Export Citation
  • Ashok, K., , S. K. Behera, , S. A. Rao, , H. Weng, , and T. Yamagata, 2007: El Niño Modoki and its possible teleconnection. J. Geophys. Res., 112 , C11007. doi:10.1029/2006JC003798.

    • Search Google Scholar
    • Export Citation
  • Baldwin, M. P., 2001: Annular modes in global daily surface pressure. Geophys. Res. Lett., 28 , 41154118.

  • Bjerknes, J., 1969: Atmospheric teleconnections from the equatorial Pacific. Mon. Wea. Rev., 97 , 163172.

  • 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.

    • Search Google Scholar
    • Export Citation
  • Czaja, A., , and C. Frankignoul, 2002: Observed impact of Atlantic SST anomalies on the North Atlantic oscillation. J. Climate, 15 , 606623.

    • Search Google Scholar
    • Export Citation
  • Deser, D., , and M. L. Blackmon, 1995: On the relationship between tropical and North Pacific sea surface temperature variations. J. Climate, 8 , 16771680.

    • Search Google Scholar
    • Export Citation
  • Frankignoul, C., , and E. Kestenare, 2005: Air–sea interactions in the tropical Atlantic: A view based on lagged rotated maximum covariance analysis. J. Climate, 18 , 38743890.

    • Search Google Scholar
    • Export Citation
  • Hoerling, M. P., , and A. Kumar, 2002: Atmospheric response patterns associated with tropical forcing. J. Climate, 15 , 21842203.

  • Horel, J. D., , and J. M. Wallace, 1981: Planetary-scale atmospheric phenomena associated with the interannual variability of sea surface temperature in the equatorial Pacific. Mon. Wea. Rev., 109 , 813829.

    • Search Google Scholar
    • Export Citation
  • Hoskins, B. J., , and D. J. Karoly, 1981: The steady linear response of a spherical atmosphere to thermal and orographic forcing. J. Atmos. Sci., 38 , 11791196.

    • Search Google Scholar
    • Export Citation
  • Huang, B., , and J. Shukla, 2005: Ocean–atmosphere interactions in the tropical and subtropical Atlantic Ocean. J. Climate, 18 , 16521672.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77 , 437471.

  • Kao, H. Y., , and J. Y. Yu, 2009: Contrasting eastern Pacific and central Pacific types of ENSO. J. Climate, 22 , 615632.

  • King, M. P., , and F. Kucharski, 2006: Observed low-frequency covariabilities between the tropical oceans and the North Atlantic oscillation in the twentieth century. J. Climate, 19 , 10321041.

    • Search Google Scholar
    • Export Citation
  • Kushnir, Y., 1994: Interdecadal variations in North Atlantic sea surface temperature and associated atmospheric conditions. J. Climate, 7 , 141157.

    • Search Google Scholar
    • Export Citation
  • Larkin, N. K., , and D. E. Harrison, 2005: Global seasonal temperature and precipitation anomalies during El Niño autumn and winter. Geophys. Res. Lett., 32 , L16705. doi:10.1029/2005GL022860.

    • Search Google Scholar
    • Export Citation
  • Lau, N-C., , and M. J. Nath, 1994: A modeling study of the relative roles of tropical and extratropical SST anomalies in the variability of the global atmosphere–ocean system. J. Climate, 7 , 11841207.

    • Search Google Scholar
    • Export Citation
  • Lau, N-C., , and M. J. Nath, 1996: The role of the “atmospheric bridge” in linking tropical Pacific ENSO events to extratropical SST anomalies. J. Climate, 9 , 20362057.

    • Search Google Scholar
    • Export Citation
  • Peng, S., , W. A. Robinson, , S. Li, , and M. P. Hoerling, 2005: Tropical Atlantic SST forcing of coupled North Atlantic seasonal responses. J. Climate, 18 , 480496.

    • Search Google Scholar
    • Export Citation
  • Rasmusson, E. M., , and T. H. Carpenter, 1982: Variation in tropical sea surface temperature and surface wind fields associated with Southern Oscillation/El Niño. Mon. Wea. Rev., 110 , 354384.

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

  • Smith, T. M., , R. W. Reynolds, , R. E. Livezy, , and D. Stokes, 1996: Reconstruction of historical sea surface temperatures using empirical orthogonal functions. J. Climate, 9 , 14031420.

    • Search Google Scholar
    • Export Citation
  • 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.

    • Search Google Scholar
    • Export Citation
  • Thompson, D. W. J., , and J. M. Wallace, 2000: Annular modes in the extratropical circulation. Part I: Month-to-month variability. J. Climate, 13 , 10001016.

    • Search Google Scholar
    • Export Citation
  • Thompson, D. W. J., , and D. J. Lorenz, 2004: The signature of the annular modes in the tropical troposphere. J. Climate, 17 , 43304342.

    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., , and D. P. Stepaniak, 2001: Indices of El Niño evolution. J. Climate, 14 , 16971701.

  • Trenberth, K. E., , G. W. Branstator, , D. Karoly, , A. Kumar, , N-C. Lau, , and C. Ropelewski, 1998: Progress during TOGA in understanding and modeling global teleconnections associated with tropical sea surface temperatures. J. Geophys. Res., 103 , 1429114324.

    • Search Google Scholar
    • Export Citation
  • Vimont, D. J., , J. M. Wallace, , and D. S. Battisti, 2003: The seasonal footprinting mechanism in the Pacific: Implications for ENSO. J. Climate, 16 , 26682675.

    • Search Google Scholar
    • Export Citation
  • Walker, G. T., , and E. W. Bliss, 1932: World weather V. Mem. Roy. Meteor. Soc., 4 , 5384.

  • Weng, H., , K. Ashok, , S. K. Behera, , S. A. Rao, , and T. Yamagata, 2007: Impacts of recent El Niño Modoki on dry/wet conditions in the Pacific rim during boreal summer. Climate Dyn., 29 , 113129. doi:10.1007/s00382-007-0234-0.

    • Search Google Scholar
    • Export Citation
  • Zhang, Y. Z., , J. M. Wallace, , and D. S. Battisti, 1997: ENSO-like interdecadal variability: 1900–93. J. Climate, 10 , 10041020.

  • View in gallery

    The (a),(d) SC, (b),(e) coefficient correlation, and (c),(f) SCF associated with the (left) first and (right) second MCA modes between anomaly fields of the extratropical Z500 (20°–90°N) and the global tropical SST (20°S–20°N) as a function of seasons and lags. SST leads Z500 at negative lags indicated (in months) on the y axis, while the x axis denotes the months assigned to Z500. The shaded area indicates where the SC is statistically significant at the 5% (dark shading) and 10% level (light shading).

  • View in gallery

    (left) Homogeneous Z500 and (right) heterogeneous SST covariance maps of DJF Z500 and SST anomalies in the second MCA mode at lags from −2 to +3 months. Contour interval is 10 m for Z500 and 0.05 K for SST. Negative contours are dashed and the zero line is omitted. The correlation coefficient r, cross-validated correlation (r*) between the SST and Z500 MCA time series, the SCF, and the SC of the mode are given for each lag. The percentages in parentheses for r and SC give their estimated significance level.

  • View in gallery

    Normalized (a) MCA–Z500 time series and (b) MCA–SST time series for the second MCA mode at lag +1 when Z500 is fixed on DJF. Each year is separated by a blank interval. Correlation maps of DJF anomaly fields of (c) SLP and (d) surface wind at each grid point in the region of 30°S–90°N with the MCA–Z500 time series in (a), and of anomaly fields of tropical–subtropical surface wind in (e) FMA and (f) AMJ with the MCA–SST time series in (b). Only correlations with amplitude ≥0.2 are indicated for SLP, with increments of 0.1, and negative contours are dashed. In (d)–(f), wind stress are shown only where surface wind speed are significantly correlated with at the corresponding coefficient time series at the 90% confidence levels.

  • View in gallery

    The cross correlation between the Niño-3.4 SST index from the previous SON to the following DJF season and the MCA–SST time series in the lagged MCA when Z500 is fixed in (a) DJF and (b) JFM, and SST lags from 0 to +3 months. Contours show the correlation between the Niño-3.4 in the month shown on the ordinate and the MCA–SST in the month shown on the abscissa. Only correlations with amplitude ≥0.3, which are significant at the 95% confidence level, are indicated, with increments of 0.1.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 21 21 4
PDF Downloads 18 18 5

Forcing of Tropical SST Anomalies by Wintertime AO-like Variability

View More View Less
  • 1 School of Meteorology, University of Oklahoma, Norman, Oklahoma
© Get Permissions
Full access

Abstract

A lagged maximum covariance analysis (MCA) is utilized to investigate large-scale patterns of covariability between sea surface temperature (SST) in the global tropics and 500-mb geopotential height (Z500) in the extratropics at monthly to interannual time scales distinct from the conventional El Niño–Southern Oscillation (ENSO) signal during the Northern Hemisphere (NH) winter. The first MCA mode indicates a strong impact of tropical SST anomalies associated with ENSO on the extratropical atmosphere. The second MCA mode corresponds with coupling between Arctic Oscillation (AO)-like atmospheric variations and tropical SST anomalies. An AO-like MCA mode appears to depict an atmosphere-to-ocean forcing, in which the tropical ocean responds to the higher extratropical AO-like atmospheric anomalies with an intraseasonal time lag. In winter, AO-like atmospheric variability is associated with the northern tropical Atlantic mode and the tropical Pacific ENSO Modoki mode through enhanced or weakened trade winds.

The above forced SST anomalies by the AO-like variability may play a role in the subsequent evolution of the conventional ENSO phenomena.

Corresponding author address: Qigang Wu, Suite 5900, 120 David L. Boren Blvd., School of Meteorology, University of Oklahoma, Norman, OK 73072. Email: wuqig@rossby.metr.ou.edu

Abstract

A lagged maximum covariance analysis (MCA) is utilized to investigate large-scale patterns of covariability between sea surface temperature (SST) in the global tropics and 500-mb geopotential height (Z500) in the extratropics at monthly to interannual time scales distinct from the conventional El Niño–Southern Oscillation (ENSO) signal during the Northern Hemisphere (NH) winter. The first MCA mode indicates a strong impact of tropical SST anomalies associated with ENSO on the extratropical atmosphere. The second MCA mode corresponds with coupling between Arctic Oscillation (AO)-like atmospheric variations and tropical SST anomalies. An AO-like MCA mode appears to depict an atmosphere-to-ocean forcing, in which the tropical ocean responds to the higher extratropical AO-like atmospheric anomalies with an intraseasonal time lag. In winter, AO-like atmospheric variability is associated with the northern tropical Atlantic mode and the tropical Pacific ENSO Modoki mode through enhanced or weakened trade winds.

The above forced SST anomalies by the AO-like variability may play a role in the subsequent evolution of the conventional ENSO phenomena.

Corresponding author address: Qigang Wu, Suite 5900, 120 David L. Boren Blvd., School of Meteorology, University of Oklahoma, Norman, OK 73072. Email: wuqig@rossby.metr.ou.edu

1. Introduction

A well-known atmospheric teleconnection pattern exists in the NH midlatitudes, forced by changes in tropical Pacific SSTs associated with ENSO (Bjerknes 1969; Horel and Wallace 1981). The dynamic link between the extratropical circulation changes and tropical heating sources can be explained by linear Rossby wave theory (Trenberth et al. 1998; Hoskins and Karoly 1981). The ENSO signature of atmosphere variability has been studied extensively and is believed to function as an “atmospheric bridge” that links interannual SST fluctuations in the tropical Pacific with oceanic variations at higher latitudes (Alexander et al. 2002; Lau and Nath 1996). Additional studies have shown that components of the large-scale extratropical atmospheric variability in the NH wintertime can influence the tropical atmospheric and/or SST variations. Thompson and Wallace (2000) demonstrate that the north annular mode [NAM; sometimes called the Arctic Oscillation (AO; Thompson and Wallace 1998)] is characterized by fluctuations in the strength of the trade winds throughout the NH subtropics. Baldwin (2001) finds that the NAM pattern is associated with variations in daily sea pressure that extend to the Southern Hemispheric tropics. Using lagged maximum covariance analysis [MCA; also known as singular value decomposition (SVD) analysis; see, e.g., Bretherton et al. (1992)], Czaja and Frankignoul (2002) show that the North Atlantic Oscillation (NAO) accounts for a substantial fraction of tropical Atlantic SST variability. Thompson and Lorenz (2004) conclude that the NAM strongly links the tropical circulation during the NH cold-season months, especially for the cold phase of the ENSO cycle. The most pronounced tropical anomalies lag the NAM index by 2 weeks over the eastern tropical Pacific. Thompson and Lorenz suggest that anomalies in the eddy momentum flux convergence at tropical latitudes associated with the NAM act to reinforce the changes of atmospheric circulation there. In addition, Thompson and Lorenz (2004) provide evidence that the recent trend in the NAM is linearly congruent with a ∼0.1-K cooling of the tropical troposphere from 1979 to 1999 during the NH winter season.

The leading mode of an empirical orthogonal function (EOF) analysis performed with tropical Pacific Ocean SST yields the well-known El Niño pattern with peak SST anomalies in the eastern Pacific (e.g., Rasmusson and Carpenter 1982). This mode accounts for most of the total variance of Pacific Ocean SST, and is usually identified as the canonical ENSO mode. The second EOF of SSTs, recently referred to as the “El Niño Modoki (pseudo–El Niño)” by Ashok et al. (2007), is characterized by warm SST anomalies in the central equatorial Pacific and cool SST anomalies on both the eastern and western regions of the basin. The entity, including its opposite phase “La Niña Modoki,” is referred as ENSO Modoki. The El Niño Modoki has been proposed as independent of the traditional El Niño or ENSO (Ashok et al. 2007; Weng et al. 2007) and represents a new type of El Niño event. Similar but not identical terms describing departures from conventional ENSO include “date line ENSO” (Larkin and Harrison 2005), the “trans-Niño” oscillation (Trenberth and Stepaniak 2001), and “central Pacific ENSO” (Kao and Yu 2009). During an El Niño Modoki event, there are two anomalous Walker circulation cells over the tropical Pacific, instead of the single-celled pattern of the conventional El Niño. El Niño Modoki and its climate impacts are very different from those of canonical El Niño (Larkin and Harrison 2005; Ashok et al. 2007; Weng et al. 2007). All of the above studies indicate that ENSO needs at least two modes to be represented.

Although Czaja and Frankignoul (2002, and other studies) have shown that the AO/NAO causes variations of tropical Atlantic SST, it is not clear whether the tropical Pacific SST variations associated with the ENSO Modoki also are influenced by the extratropical atmospheric variability. The main purpose here is to investigate the linear covariability between the extratropical atmospheric circulation in the NH and the tropical SST (20°S–20°N) in observations independent of the canonical ENSO signal during NH winter using lagged MCA. The lagged MCA has been increasingly utilized to examine the lag association between two fields as a means of ascertaining cause and effect (Czaja and Frankignoul 2002; Frankignoul and Kestenare 2005; King and Kucharski 2006). We extend studies by Thompson and Wallace (2000) and Thompson and Lorenz (2004) to examine the possible impact of NH wintertime NAM anomalies on the tropical SST variations exclusive of the conventional ENSO signal on monthly and longer time scales. The rest of this paper is arranged as follows: section 2 describes the data sources and analysis techniques, the MCA results are presented in section 3, and a short summary is provided in section 4.

2. Datasets and methodology

The key datasets used in this study are the global optimum-interpolated SST (Smith et al. 1996), National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR)-reanalyzed geopotential heights at 500 mb (Z500), and sea level pressure (SLP) and surface wind fields obtained from the National Oceanic and Atmospheric Administration (NOAA)/Climate Diagnostic Center (CDC). The NCEP–NCAR-reanalyzed data are on a 2.5° latitude × 2.5° longitude horizontal grid (Kalnay et al. 1996). We have aggregated Z500 into averages for 5° latitude × 5° longitude grid cells. The 60-yr period from 1948 to 2007 over the region of 20°–90°N is used. The global tropical SST field over the region of 20°S–20°N is analyzed at 4° resolution for the same period. Our results are not sensitive to the precise definitions of the domains. The primary analysis tools used in this study are EOF analysis and lagged MCA. To reduce the influence of trends and low-frequency changes, a second-order polynomial was removed from the monthly anomalies by a least squares fit over the full period considered. Note that the ENSO teleconnection is not removed from either the SST or atmospheric datasets. In both EOF and MCA, area weighting is accomplished by multiplying the Z500 by the square root of the cosine of latitude before computing the covariance matrix. The patterns associated with Z500 have been divided by the square root of the cosine of latitude before plotting.

We use the lagged MCA to investigate covariability of extratropical Z500 in the NH and SST in the global tropics in lead and lag conditions. A separate MCA with fixed 3-month Z500 fields is carried out for each time lag between −6 months (SST leading) and +6 months (SST lagging), as in Czaja and Frankignoul (2002). The lagged covariance matrix is estimated with monthly anomalies binned into groups of 3 months for Z500 on eight seasons from September–November (SON), October–December (OND), November–January (NDJ), December–February (DJF), January–March (JFM), February–April (FMA), March–May (MAM), to April–June (AMJ). For each season and lag, the MCA is based on 59 yr and the length of the time series is 177 months. We evaluate the significance of statistics in the MCA, the squared covariance (SC), and the temporal correlation (r) between the expansion coefficients of Z500 and SST with the exact same Monte Carlo approach described in Czaja and Frankignoul (2002). One hundred of the ensembles of MCA between the scrambled Z500 and original SST are performed for each observed MCA. At last, the predictability of the atmospheric (or SST) signal was found by cross validation, removing successive sets of 3 yr of Z500 and SST from the original anomalous fields before the MCA, and then using the MCA patterns from the 56-yr datasets to determine their amplitude in the middle year that was removed.

Various atmosphere and SST indices in DJF and other seasons are defined. The leading EOF of DJF (and other seasons) monthly Z500 anomalies is readily identifiable as being the AO pattern (Thompson and Wallace 1998). The corresponding standardized principal component (PC) time series are defined as the AO index. The northern tropical Atlantic mode is defined as the first leading mode of rotated EOF analysis of DJF (and other seasons) monthly SST in the tropical Atlantic (20°S–20°N, 70°W–60°E), and the associated standardized PC time series is defined as the northern tropical Atlantic SST index (NATL). The conventional ENSO index is defined as the averaged SST over the Niño-3.4 region (5°S–5°N, 170–120°W). To characterize the second mode of ENSO, the standardized PC time series of the second EOF (El Niño Modoki) of the tropical Pacific SST anomalies (20°S–20°N, 120°E–80°W) are defined as the ENSO Modoki index (EMI): positive (negative) values of the index correspond to El Niño Modoki (La Niña Modoki) events.

3. Results

Figure 1 presents statistics, including the SC, r, and the squared covariance fraction (SCF), associated with the leading two MCA modes in the analyses between the SST in the global tropics and Z500 in the NH extratropics as a function of lag and seasonality. Negative (positive) lags indicate that SST leads (lags) Z500. Throughout the seasons considered, significant SCs are found associated with the first MCA mode when SST leads Z500 by up to 6 months, which suggests a strong SST impact on the Z500 variability. Significant r associated with the first MCA mode is mostly found, and is not dependent on lag and seasons. When SST leads Z500, about 80%–85% of the TSC is explained by the first MCA mode. The coupled patterns associated with the first MCA mode when SST and Z500 is assigned at DJF and JFM (not shown) indicate the forcing of atmosphere in the extratropics by the SST associated with conventional ENSO events in the tropical Pacific. The ENSO signature of extratropical atmospheric variability is very similar to that in the SVD analyses conducted between the seasonal mean Z500 and Pacific SSTs (e.g., Lau and Nath 1994; Zhang et al. 1997). The simultaneous correlation between the MCA–SST time series at all lags and the corresponding Niño-3.4 SST index is very high (0.90–0.99), which also proves that the first MCA mode corresponds well with canonical ENSO events. Figure 1 shows that SC, r, and SCF are strongest when tropical SST leads the atmosphere by 1 month and then decays slowly at longer negative and positive lags. This indicates a persistent ENSO-like SST influence on the atmosphere associated with the first MCA mode during NH wintertime. Hereafter, the first MCA mode will be referred as the ENSO-like MCA mode.

For the second MCA mode (Figs. 1d–f), significant SCs are mostly found for lags 0 to +6 months for all seasons considered. On the other hand, significant SCs are only found when SST leads Z500 by 1–3 months during fall and winter (primarily OND, NDJ, and DJF), and 3–5 months during spring (primarily MAM and AMJ). When Z500 leads SST, about 10%–25% of the TSC is explained by the second MCA mode. Figure 1 shows that the SCs and r associated with the second MCA mode are strongest when the atmosphere leads the SST by 1 month and then decays slowly at longer positive lags, but decays quickly at larger negative lags. Such asymmetry of statistics suggests that the association is strongest when the Z500 leads the SST by 1 month and the dominant air–sea interaction associated with the second MCA mode is the forcing of SST in the tropics by the atmosphere in the NH extratropics.

Figure 2 presents the coupled covariance patterns at lags from −2 to +3 months that are 10% significant in SC and r when Z500 is fixed in DJF. Cross validation suggests that the correlation (r*) is robust for each lag. Each pair of patterns is formed from the so-called heterogeneous covariance for SST and homogeneous pattern for Z500. For all lags, the 500-mb-height patterns correspond with the dominant Z500 EOF mode and are virtually identical to the AO pattern (Thompson and Wallace 1998), and the anomalous SST maximum is predominantly located over the tropical Atlantic and Pacific. The maximum covariance patterns of Z500 vary little with lag, and the expansion coefficients of Z500 associated with the second MCA mode are highly correlated with the AO index, with correlations about 0.90 for lags 0–3 months.

The AO-like atmospheric variability has its greatest impact on SST in the tropical Atlantic between 5° and 20°N and is related to the northern tropical Atlantic mode (Huang and Shukla 2005). At lag 0 and +1, the MCA–SST time series associated with the second MCA mode have high correlations (about 0.60) with the corresponding NATL indices. The above results are consistent with the NAO-like atmospheric forcing to the northern tropical Atlantic SST in Czaja and Frankignoul (2002). In the tropical Pacific, the SST structure is qualitatively similar to the La Niña Modoki (Ashok et al. 2007) or the “central Pacific ENSO” (Kao and Yu 2009). The MCA–SST time series shows a significant negative temporal correlation (about −0.70 for lags 0 and 1 month and −0.60 for lags 2 and 3 months) with the EMI, indicating that the winter- and springtime SST variability associated with ENSO Modoki are significantly correlated with the wintertime AO. Because the SC is strongest when the Z500 leads the SST by 1 month and then decays at longer positive lags (Fig. 1), the spatial patterns in Fig. 2 and the statistics in Fig. 1 associated with the second MCA mode for the positive lags suggest that the dominant air–sea interaction is the persistent forcing of SST fluctuations in the tropics by the AO-like atmospheric variability during NH wintertime. Hereafter, the second MCA mode will be referred as the AO-like MCA mode.

The coupled SST and Z500 patterns associated with the second MCA mode when Z500 is assigned to JFM and SST lags from 0 to +3 months also indicate the forcing of the JFM AO-like atmospheric forcing to the northern tropical Atlantic SST mode and the ENSO Modoki mode (not shown). The MCA–Z500 time series are highly correlated with the AO index, having correlations of roughly 0.80 for lags of 0–3 months, and the MCA–SST time series are highly correlated to NATL (correlations about 0.50 for lags 0–1 months), and EMI (correlation of about −0.80 for lag 0 and −0.66 for lag 1).

To investigate possible atmosphere-to-ocean forcing mechanisms related to the AO-like MCA mode, we generate correlation maps of DJF monthly anomaly fields of SLP and surface wind. Because the maximum covariance (∼16.0), correlation (∼0.56), and cross-validation correlation (0.51) in the lagged MCA for the fixed DJF Z500 occurs when Z500 leads SST by 1 month, the maps for these variables are formed by correlating the MCA–Z500 time series at lag +1 with grid points from 30°S to 90°N in Fig. 3. The correlation between the MCA–Z500 (MCA–SST) time series in Fig. 3a (Fig. 3b) and the AO index is about 0.90 (0.43). For SLP in Fig. 3c, the most pronounced feature is found at NH high and subtropical latitudes, but a robust feature also is evident at tropical latitudes. SLP variability displays a significant basinwide positive pressure anomaly in the North Pacific and Atlantic Oceans, each extending from 5° to 10°N latitude. The anomalous anticyclone circulation is also depicted by the vector winds shown in Fig. 3d. The enhanced northeasterly trade winds associated with the AO found in Thompson and Wallace (2000) are clearly evident in the tropical region. Such results agree with Thompson and Lorenz (2004) and Baldwin (2001) in that the AO should be viewed as structures that extend deep into the tropics. The distribution of the tropical SST anomaly field in JFM (Fig. 2c) displays a consistent relationship with the anomalies in the surface circulation shown in Fig. 3d, suggesting that the tropical SST anomalies associated with the second MCA mode are driven by wind-induced mechanisms (Deser and Blackmon 1995; Kushnir 1994).

Figures 1 and 2 demonstrate that tropical SST anomalies exert a significant influence on the AO-like atmospheric variability in early and middle NH winter at short lead times. For lag −1 and −2 months in Fig. 2, the correlations between the MCA–Z500 (SST) time series and the AO (EMI) index are about 0.8 (−0.6). This suggests that the ENSO Modoki could generate atmospheric changes in the NH extratropics through teleconnections (Ashok et al. 2007). Tropical heating away from the western Pacific also contribute to the formation of the annular-like pattern, as demonstrated in previous studies (e.g., Hoerling and Kumar 2002; Peng et al. 2005). The cross-validated correlations for the second MCA mode are 0.27 and 0.18 for lag −1 and −2, respectively (Fig. 2). This implies that about 3%–7% of AO-like atmospheric variance is influenced by the tropical SST anomalies beyond the ENSO signal at monthly to interannual time scales.

The MCA isolates pairs of spatial patterns and their associated time series by performing a singular value decomposition of the covariance matrix between Z500 and SST. Both Z500 and SST are expanded into orthogonal patterns that maximize their covariance, with the time series being orthogonal to one another between these two fields. Because the first MCA mode represents the coupled variability between wintertime Z500 and SST associated with the canonical ENSO mode, interaction in Fig. 2 associated with the second MCA mode is not dependent on the conventional ENSO, and the MCA–SST time series is uncorrelated with the Niño-3.4 index. The cross-correlation between the Niño-3.4 index and the MCA–SST time series is insignificant when the former leads the latter up to the preceding fall when Z500 are fixed at DJF and JFM and the SST lags from 0 to +3 months (Figs. 4a,b). Instead, Fig. 4 demonstrates that the MCA–SST time series for lags +2 and +3 months for DJF and JFM Z500 are significantly correlated with Niño-3.4 SST anomalies in the following seasons. Therefore, the forced SST anomalies by the AO-like variability may play a role in the subsequent evolution of the conventional ENSO phenomena.

The above link may be related to the seasonal footprinting mechanisms in Vimont et al. (2003). Vimont et al. (2003) found that winter SLP anomalies associated with the North Pacific Oscillation (NPO; Walker and Bliss 1932; Rogers 1981) generate SST anomalies in the tropics and subtropics that persist into the summer seasons and cause zonal wind stress anomalies along the equator. These equatorial zonal wind stress anomalies that are generated through the seasonal footprinting mechanism are a source of stochastic forcing for tropical ENSO variability. The SLP signature in the North Pacific in Fig. 3c bears considerable similarity to the negatively polarized NPO-like SLP anomalies in Vimont et al. (2003, their Fig. 3a); the structural similarity between the negative phase of the forced SST pattern in the northern tropical and subtropical Pacific in Fig. 2c, and the corresponding SST footprint forced by the NPO in Vimont et al. (2003, their Fig. 4a), is remarkable. Significant surface wind stress anomalies in the tropics are found to follow forced SST variability associated with the AO-like MCA mode. Figures 3e,f show the lag regression patterns of the surface wind in FMA and AMJ seasons over the tropics and subtropics on the MCA–SST time series in Fig. 3b. Significant westerly anomalies are maintained in the equator from spring to summer. To highlight the wind anomalies that are associated with the AO-like MCA mode, components associated with the developing canonical ENSO event are removed from the wind fields prior to generating the correlation maps (Figs. 3e,f) using a regression against the Niño-3.4 SST anomalies of the preceding months. The regression coefficient is selected as the maximum regression coefficient within the preceding 6 months. Both the AO-like midlatitude atmospheric variability in our study and the negatively polarized NPO can impact the Pacific trade winds because their southern lobes are located in the northern subtropics, underscoring the potential roles of the wintertime midlatitude atmospheric variability in ocean–atmosphere interaction in the tropics/subtropics in the following seasons.

4. Summary

In this study, a lagged MCA is performed to investigate the linear covariability between the global SST in the tropics and hemispheric atmosphere in the extratropics during NH wintertime at monthly to interannual time scales. The large-scale modes of coupled ocean–atmosphere variability are found to be associated with the leading two MCA modes. The first MCA mode indicates that the tropical SST anomalies associated with conventional ENSO force the atmospheric circulation in the NH extratropics. Such an ENSO signature of Z500 originates from the tropics, so that the first ENSO mode signal leads the atmosphere (Bjerknes 1969; Horel and Wallace 1981; Lau and Nath 1996). In contrast, the second AO-like MCA mode in our study mostly arises from the midlatitude intrinsic atmospheric variability, so that the atmospheric signal leads SST variability associated with the second ENSO mode and northern tropical Atlantic mode. Variations in the extratropical AO-like atmospheric variability thus may provide useful information for predicting tropical SST variability.

In NH wintertime, the AO-like atmospheric variability appears to force the SST anomalies associated with the northern tropical Atlantic mode and the ENSO Modoki mode in the tropical Pacific. Such forced SST variability reflects the Atlantic and Pacific signature of the warming–cooling of the tropical belt associated with a remote forcing by AO-like climate variability in the NH midlatitudes. The impact of AO on the tropical Pacific SST explains why the El Niño Modoki mode has much stronger variability in boreal winter than in summer (Ashok et al. 2007) and why the “central Pacific ENSO” tends to reach its peak intensity around December–January (Kao and Yu 2009). The above results, together with that found by Baldwin (2001) and Thompson and Lorenz (2004), reveal the tropical component of the AO-like atmospheric variability that forces the tropical SST during the NH wintertime.

Our results suggest that wintertime AO-like midlatitude atmospheric variability is affecting the development of the canonical ENSO mode in the following spring and summer. ENSO Modoki-like SST anomalies forced by the wintertime AO-like variability may persist into spring and early summer, and cause equatorial zonal wind stress anomalies, which is a mechanism related to the seasonal footprinting mechanisms in Vimont et al. (2003).

Acknowledgments

QW is supported by a Gary Comer Science and Education Foundation Fellowship and NSF Grant ATM-0555326. The author wishes to thank the two anonymous reviewers for providing comprehensive, constructive reviews, which were very helpful for the revision of the manuscript.

REFERENCES

  • Alexander, M. A., , I. Bladé, , M. Newman, , J. R. Lazante, , N-C. Lau, , and J. D. Scott, 2002: The atmospheric bridge: The influence of ENSO on air–sea interaction over the global oceans. J. Climate, 15 , 22052231.

    • Search Google Scholar
    • Export Citation
  • Ashok, K., , S. K. Behera, , S. A. Rao, , H. Weng, , and T. Yamagata, 2007: El Niño Modoki and its possible teleconnection. J. Geophys. Res., 112 , C11007. doi:10.1029/2006JC003798.

    • Search Google Scholar
    • Export Citation
  • Baldwin, M. P., 2001: Annular modes in global daily surface pressure. Geophys. Res. Lett., 28 , 41154118.

  • Bjerknes, J., 1969: Atmospheric teleconnections from the equatorial Pacific. Mon. Wea. Rev., 97 , 163172.

  • 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.

    • Search Google Scholar
    • Export Citation
  • Czaja, A., , and C. Frankignoul, 2002: Observed impact of Atlantic SST anomalies on the North Atlantic oscillation. J. Climate, 15 , 606623.

    • Search Google Scholar
    • Export Citation
  • Deser, D., , and M. L. Blackmon, 1995: On the relationship between tropical and North Pacific sea surface temperature variations. J. Climate, 8 , 16771680.

    • Search Google Scholar
    • Export Citation
  • Frankignoul, C., , and E. Kestenare, 2005: Air–sea interactions in the tropical Atlantic: A view based on lagged rotated maximum covariance analysis. J. Climate, 18 , 38743890.

    • Search Google Scholar
    • Export Citation
  • Hoerling, M. P., , and A. Kumar, 2002: Atmospheric response patterns associated with tropical forcing. J. Climate, 15 , 21842203.

  • Horel, J. D., , and J. M. Wallace, 1981: Planetary-scale atmospheric phenomena associated with the interannual variability of sea surface temperature in the equatorial Pacific. Mon. Wea. Rev., 109 , 813829.

    • Search Google Scholar
    • Export Citation
  • Hoskins, B. J., , and D. J. Karoly, 1981: The steady linear response of a spherical atmosphere to thermal and orographic forcing. J. Atmos. Sci., 38 , 11791196.

    • Search Google Scholar
    • Export Citation
  • Huang, B., , and J. Shukla, 2005: Ocean–atmosphere interactions in the tropical and subtropical Atlantic Ocean. J. Climate, 18 , 16521672.

    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77 , 437471.

  • Kao, H. Y., , and J. Y. Yu, 2009: Contrasting eastern Pacific and central Pacific types of ENSO. J. Climate, 22 , 615632.

  • King, M. P., , and F. Kucharski, 2006: Observed low-frequency covariabilities between the tropical oceans and the North Atlantic oscillation in the twentieth century. J. Climate, 19 , 10321041.

    • Search Google Scholar
    • Export Citation
  • Kushnir, Y., 1994: Interdecadal variations in North Atlantic sea surface temperature and associated atmospheric conditions. J. Climate, 7 , 141157.

    • Search Google Scholar
    • Export Citation
  • Larkin, N. K., , and D. E. Harrison, 2005: Global seasonal temperature and precipitation anomalies during El Niño autumn and winter. Geophys. Res. Lett., 32 , L16705. doi:10.1029/2005GL022860.

    • Search Google Scholar
    • Export Citation
  • Lau, N-C., , and M. J. Nath, 1994: A modeling study of the relative roles of tropical and extratropical SST anomalies in the variability of the global atmosphere–ocean system. J. Climate, 7 , 11841207.

    • Search Google Scholar
    • Export Citation
  • Lau, N-C., , and M. J. Nath, 1996: The role of the “atmospheric bridge” in linking tropical Pacific ENSO events to extratropical SST anomalies. J. Climate, 9 , 20362057.

    • Search Google Scholar
    • Export Citation
  • Peng, S., , W. A. Robinson, , S. Li, , and M. P. Hoerling, 2005: Tropical Atlantic SST forcing of coupled North Atlantic seasonal responses. J. Climate, 18 , 480496.

    • Search Google Scholar
    • Export Citation
  • Rasmusson, E. M., , and T. H. Carpenter, 1982: Variation in tropical sea surface temperature and surface wind fields associated with Southern Oscillation/El Niño. Mon. Wea. Rev., 110 , 354384.

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

  • Smith, T. M., , R. W. Reynolds, , R. E. Livezy, , and D. Stokes, 1996: Reconstruction of historical sea surface temperatures using empirical orthogonal functions. J. Climate, 9 , 14031420.

    • Search Google Scholar
    • Export Citation
  • 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.

    • Search Google Scholar
    • Export Citation
  • Thompson, D. W. J., , and J. M. Wallace, 2000: Annular modes in the extratropical circulation. Part I: Month-to-month variability. J. Climate, 13 , 10001016.

    • Search Google Scholar
    • Export Citation
  • Thompson, D. W. J., , and D. J. Lorenz, 2004: The signature of the annular modes in the tropical troposphere. J. Climate, 17 , 43304342.

    • Search Google Scholar
    • Export Citation
  • Trenberth, K. E., , and D. P. Stepaniak, 2001: Indices of El Niño evolution. J. Climate, 14 , 16971701.

  • Trenberth, K. E., , G. W. Branstator, , D. Karoly, , A. Kumar, , N-C. Lau, , and C. Ropelewski, 1998: Progress during TOGA in understanding and modeling global teleconnections associated with tropical sea surface temperatures. J. Geophys. Res., 103 , 1429114324.

    • Search Google Scholar
    • Export Citation
  • Vimont, D. J., , J. M. Wallace, , and D. S. Battisti, 2003: The seasonal footprinting mechanism in the Pacific: Implications for ENSO. J. Climate, 16 , 26682675.

    • Search Google Scholar
    • Export Citation
  • Walker, G. T., , and E. W. Bliss, 1932: World weather V. Mem. Roy. Meteor. Soc., 4 , 5384.

  • Weng, H., , K. Ashok, , S. K. Behera, , S. A. Rao, , and T. Yamagata, 2007: Impacts of recent El Niño Modoki on dry/wet conditions in the Pacific rim during boreal summer. Climate Dyn., 29 , 113129. doi:10.1007/s00382-007-0234-0.

    • Search Google Scholar
    • Export Citation
  • Zhang, Y. Z., , J. M. Wallace, , and D. S. Battisti, 1997: ENSO-like interdecadal variability: 1900–93. J. Climate, 10 , 10041020.

Fig. 1.
Fig. 1.

The (a),(d) SC, (b),(e) coefficient correlation, and (c),(f) SCF associated with the (left) first and (right) second MCA modes between anomaly fields of the extratropical Z500 (20°–90°N) and the global tropical SST (20°S–20°N) as a function of seasons and lags. SST leads Z500 at negative lags indicated (in months) on the y axis, while the x axis denotes the months assigned to Z500. The shaded area indicates where the SC is statistically significant at the 5% (dark shading) and 10% level (light shading).

Citation: Journal of Climate 23, 10; 10.1175/2009JCLI2749.1

Fig. 2.
Fig. 2.

(left) Homogeneous Z500 and (right) heterogeneous SST covariance maps of DJF Z500 and SST anomalies in the second MCA mode at lags from −2 to +3 months. Contour interval is 10 m for Z500 and 0.05 K for SST. Negative contours are dashed and the zero line is omitted. The correlation coefficient r, cross-validated correlation (r*) between the SST and Z500 MCA time series, the SCF, and the SC of the mode are given for each lag. The percentages in parentheses for r and SC give their estimated significance level.

Citation: Journal of Climate 23, 10; 10.1175/2009JCLI2749.1

Fig. 3.
Fig. 3.

Normalized (a) MCA–Z500 time series and (b) MCA–SST time series for the second MCA mode at lag +1 when Z500 is fixed on DJF. Each year is separated by a blank interval. Correlation maps of DJF anomaly fields of (c) SLP and (d) surface wind at each grid point in the region of 30°S–90°N with the MCA–Z500 time series in (a), and of anomaly fields of tropical–subtropical surface wind in (e) FMA and (f) AMJ with the MCA–SST time series in (b). Only correlations with amplitude ≥0.2 are indicated for SLP, with increments of 0.1, and negative contours are dashed. In (d)–(f), wind stress are shown only where surface wind speed are significantly correlated with at the corresponding coefficient time series at the 90% confidence levels.

Citation: Journal of Climate 23, 10; 10.1175/2009JCLI2749.1

Fig. 4.
Fig. 4.

The cross correlation between the Niño-3.4 SST index from the previous SON to the following DJF season and the MCA–SST time series in the lagged MCA when Z500 is fixed in (a) DJF and (b) JFM, and SST lags from 0 to +3 months. Contours show the correlation between the Niño-3.4 in the month shown on the ordinate and the MCA–SST in the month shown on the abscissa. Only correlations with amplitude ≥0.3, which are significant at the 95% confidence level, are indicated, with increments of 0.1.

Citation: Journal of Climate 23, 10; 10.1175/2009JCLI2749.1

Save