• An, S.-I., , and F.-F. Jin, 2001: Collective role of thermocline and zonal advective feedbacks in the ENSO mode. J. Climate, 14, 34213432.

    • Search Google Scholar
    • Export Citation
  • An, S.-I., , and F.-F. Jin, 2004: Nonlinearity and asymmetry of ENSO. J. Climate, 17, 23992412.

  • An, S.-I., , F.-F. Jin, , and I.-S. Kang, 1999: The role of zonal advection feedback in phase transition and growth of ENSO in the Cane-Zebiak model. J. Meteor. Soc. Japan, 77, 11511160.

    • Search Google Scholar
    • Export Citation
  • Ashok, H., , and T. Yamagata, 2009: Climate change: The El Niño with a difference. Nature, 461, 481484, doi:10.1038/461481a.

  • 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
  • Battisti, D. S., , and A. C. Hirst, 1989: Interannual variability in a tropical atmosphere–ocean model: Influence of the basic state, ocean geometry, and nonlinearity. J. Atmos. Sci., 46, 16871712.

    • Search Google Scholar
    • Export Citation
  • Behringer, D. W., , and Y. Xue, 2004: Evaluation of the global ocean data assimilation system at NCEP: The Pacific Ocean. Preprints, Eighth Symp. on Integrated Observing and Assimilation Systems for Atmosphere, Oceans, and Land Surface, Seattle, WA, Amer. Meteor. Soc., 2.3. [Available online at https://ams.confex.com/ams/pdfpapers/70720.pdf.]

  • Bejarano, L., , and F.-F. Jin, 2008: Coexistence of equatorial coupled modes of ENSO. J. Climate, 21, 30513067.

  • Cane, M. A., , M. Münnich, , and S. E. Zebiak, 1990: A study of self-excited oscillations of the tropical ocean–atmosphere system. Part I: Linear analysis. J. Atmos. Sci., 47, 15621577.

    • Search Google Scholar
    • Export Citation
  • Compo, G. P., , J. S. Whitaker, , and P. D. Sardeshmukh, 2006: Feasibility of a 100-year reanalysis using only surface pressure data. Bull. Amer. Meteor. Soc., 87, 175190.

    • Search Google Scholar
    • Export Citation
  • Dewitte, B., , J. Choi, , S.-I. An, , and S. Thual, 2012: Vertical structure variability and equatorial waves during central Pacific and eastern Pacific El Niños in a coupled general circulation model. Climate Dyn., 38, 22752289, doi:10.1007/s00382-011-1215-x.

    • Search Google Scholar
    • Export Citation
  • Giese, B. S., , and S. Ray, 2011: El Niño variability in simple ocean data assimilation (SODA), 1871–2008. J. Geophys. Res., 116, C02024, doi:10.1029/2010JC006695.

    • Search Google Scholar
    • Export Citation
  • Ham, Y.-G., , and J.-S. Kug, 2012: How well do current climate models simulate two types of El Niño? Climate Dyn., 39, 383398, doi:10.1007/s00382-011-1157-3.

    • Search Google Scholar
    • Export Citation
  • Jin, F.-F., 1996: Tropical ocean-atmosphere interaction, the Pacific cold tongue, and the El Niño-Southern Oscillation. Science, 274, 7678.

    • Search Google Scholar
    • Export Citation
  • Jin, F.-F., 1997a: An equatorial ocean recharge paradigm for ENSO. Part I: Conceptual model. J. Atmos. Sci., 54, 811829.

  • Jin, F.-F., 1997b: An equatorial ocean recharge paradigm for ENSO. Part II: A stripped-down coupled model. J. Atmos. Sci., 54, 830847.

    • Search Google Scholar
    • Export Citation
  • Jin, F.-F., , and J. D. Neelin, 1993: Modes of interannual tropical ocean-atmosphere interaction—A unified view. Part I: Numerical results. J. Atmos. Sci., 50, 34773503.

    • Search Google Scholar
    • Export Citation
  • Jin, F.-F., , and S.-I. An, 1999: Thermocline and zonal advective feedbacks within the equatorial ocean recharge oscillator model for ENSO. Geophys. Res. Lett., 26, 29892992.

    • Search Google Scholar
    • Export Citation
  • Jin, F.-F., , S.-I. An, , A. Timmermann, , and J. Zhao, 2003: Strong El Niño events and nonlinear dynamical heating. Geophys. Res. Lett., 30, 1120, doi:10.1029/2002GL016356.

    • Search Google Scholar
    • Export Citation
  • Jin, F.-F., , S. T. Kim, , and L. Bejarano, 2006: A coupled-stability index for ENSO. Geophys. Res. Lett., 33, L23708, doi:10.1029/2006GL027221.

    • Search Google Scholar
    • Export Citation
  • Kang, I.-S., , S.-I. An, , and F.-F. Jin, 2001: The role of zonal advection feedback in phase transition and growth of ENSO in the Cane-Zebiak model. J. Meteor. Soc. Japan, 79, 110.

    • Search Google Scholar
    • Export Citation
  • Kao, H.-Y., , and J.-Y. Yu, 2009: Contrasting eastern-Pacific and central-Pacific types of El Niño. J. Climate, 22, 615632.

  • Kim, H.-M., , P. J. Webster, , and J. A. Curry, 2009: Impact of shifting patterns of Pacific Ocean warming on North Atlantic tropical cyclones. Science, 325, 7780, doi:10.1126/science.1174062.

    • Search Google Scholar
    • Export Citation
  • Kug, J.-S., , I.-S. Kang, , and S.-I. An, 2003: Symmetric and antisymmetric mass exchanges between the equatorial and off-equatorial Pacific associated with ENSO. J. Geophys. Res., 108, 3284, doi:10.1029/2002JC001671.

    • Search Google Scholar
    • Export Citation
  • Kug, J.-S., , F.-F. Jin, , and S.-I. An, 2009: Two types of El Niño events: Cold tongue El Niño and warm pool El Niño. J. Climate, 22, 14991515.

    • Search Google Scholar
    • Export Citation
  • Kug, J.-S., , J. Choi, , S.-I. An, , F.-F. Jin, , and A. T. Wittenberg, 2010: Warm pool and cold tongue El Niño events as simulated by the GFDL 2.1 coupled GCM. J. Climate, 23, 12261239.

    • Search Google Scholar
    • Export Citation
  • Larkin, N. K., , and D. E. Harrison, 2005a: On the definition of El Niño and associated seasonal average U.S. weather anomalies. Geophys. Res. Lett., 32, L13705, doi:10.1029/2005GL022738.

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

    • Search Google Scholar
    • Export Citation
  • Mason, S. J., , and L. Goddard, 2001: Probabilistic precipitation anomalies associated with ENSO. Bull. Amer. Meteor. Soc., 82, 619638.

    • Search Google Scholar
    • Export Citation
  • McPhaden, M. J., 2012: A 21st century shift in the relationship between ENSO SST and warm water volume anomalies. Geophys. Res. Lett., 39, L09706, doi:10.1029/2012GL051826.

    • Search Google Scholar
    • Export Citation
  • Meinen, C. S., , and M. J. McPhaden, 2000: Observations of warm water volume changes in the equatorial Pacific and their relationship to El Niño and La Niña. J. Climate, 13, 35513559.

    • Search Google Scholar
    • Export Citation
  • Picaut, J., , F. Masia, , and Y. du Penhoat, 1997: An advective-reflective conceptual model for the oscillatory nature of ENSO. Science, 277, 663666.

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

    • Search Google Scholar
    • Export Citation
  • Ren, H.-L., , and F.-F. Jin, 2011: Niño indices for two types of ENSO. Geophys. Res. Lett., 38, L04704, doi:10.1029/2010GL046031.

  • Ropelewski, C. F., , and M. S. Halpert, 1987: Global and regional scale precipitation associated with El Niño/Southern Oscillation. Mon. Wea. Rev., 115, 16061626.

    • Search Google Scholar
    • Export Citation
  • Smith, R. D., , J. K. Dukowicz, , and R. C. Malone, 1992: Parallel ocean general circulation modeling. Physica D, 60, 3861, doi:10.1016/0167-2789(92)90225-C.

    • Search Google Scholar
    • Export Citation
  • Smith, T. M., , R. W. Reynolds, , T. C. Peterson, , and J. Lawrimore, 2008: Improvements to NOAA's historical merged land–ocean surface temperature analysis (1880–2006). J. Climate, 21, 22832296.

    • Search Google Scholar
    • Export Citation
  • Suarez, M. J., , and P. S. Schopf, 1988: A delayed action oscillator for ENSO. J. Atmos. Sci., 45, 32833287.

  • Trenberth, K. E., 1997: The definition of El Niño. Bull. Amer. Meteor. Soc., 78, 27712777.

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

  • Wang, W., , and M. J. McPhaden, 2001: Surface layer temperature balance in the equatorial Pacific during the 1997–98 El Niño and 1998–99 La Niña. J. Climate, 14, 33933407.

    • Search Google Scholar
    • Export Citation
  • Weisberg, R. H., , and C. Wang, 1997: A western Pacific oscillator paradigm for the El Niño-Southern Oscillation. Geophys. Res. Lett., 24, 779782.

    • Search Google Scholar
    • Export Citation
  • Weng, H., , K. Ashok, , S. Behera, , A. S. 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
  • Weng, H., , S. K. Behera, , and T. Yamagata, 2009: Anomalous winter climate conditions in the Pacific Rim during recent El Niño Modoki and El Niño events. Climate Dyn., 32, 663674, doi:10.1007/s00382-008-0394-6.

    • Search Google Scholar
    • Export Citation
  • Whitaker, J. S., , G. P. Compo, , X. Wei, , and T. M. Hamill, 2004: Reanalysis without radiosondes using ensemble data assimilation. Mon. Wea. Rev., 132, 11901200.

    • Search Google Scholar
    • Export Citation
  • Yeh, S.-W., , J.-S. Kug, , B. Dewitte, , M.-H. Kwon, , B. P. Kirtman, , and F.-F. Jin, 2009: El Niño in a changing climate. Nature, 461, 511514, doi:10.1038/nature08316.

    • Search Google Scholar
    • Export Citation
  • Yu, J.-Y., , and S. T. Kim, 2010a: Three evolution patterns of central-Pacific El Niño. Geophys. Res. Lett., 37, L08706, doi:10.1029/2010GL042810.

    • Search Google Scholar
    • Export Citation
  • Yu, J.-Y., , and S. T. Kim, 2010b: Identification of central-Pacific and eastern-Pacific types of ENSO in CMIP3 models. Geophys. Res. Lett., 37, L15705, doi:10.1029/2010GL044082.

    • Search Google Scholar
    • Export Citation
  • Yu, J.-Y., , F. Sun, , and H.-Y. Kao, 2009: Contributions of Indian Ocean and monsoon biases to the excessive biennial ENSO in CCSM3. J. Climate, 22, 18501858.

    • Search Google Scholar
    • Export Citation
  • Yu, J.-Y., , H.-Y. Kao, , and T. Lee, 2010: Subtropics-related interannual sea surface temperature variability in the equatorial central Pacific. J. Climate, 23, 28692884.

    • Search Google Scholar
    • Export Citation
  • Zhang, Q., , A. Kumar, , Y. Xue, , W. Wang, , and F.-F. Jin, 2007: Analysis of the ENSO cycle in the NCEP coupled forecast model. J. Climate, 20, 12651284.

    • Search Google Scholar
    • Export Citation
  • Zhang, W., , F.-F. Jin, , J. Li, , and H.-L. Ren, 2011: Contrasting impacts of two-type El Niño over the western north Pacific during boreal autumn. J. Meteor. Soc. Japan, 89, 563569.

    • Search Google Scholar
    • Export Citation
  • Zhang, W., , F.-F. Jin, , H.-L. Ren, , J. Li, , and J.-X. Zhao, 2012: Differences in teleconnection over the North Pacific and rainfall shift over the USA associated with two types of El Niño during boreal autumn. J. Meteor. Soc. Japan, 90, 535552.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    (a) Zonal-mean equatorial indices of SSTA (line with circles, K) and SSHA (solid line, 0.025 m) and recharge oscillkator index (RDI) (dashed line, 1 × 10−8) for the 2009–10 case. The indices are generated over 5°S–5°N. The SSTA index is averaged over 160°E–110°W and the SSHA index over 120°E–110°W. RDI is defined as the average of zonal gradient of SSHA over 120°E–90°W, positively proportional to meridional geostrophic current and represents the intensity of recharge–discharge of equatorial upper-ocean heat content. SSH is from the GODAS data and the climatology is defined in 1980–2010. (b),(c) As in (a) but for the QQ and QB modes, based on Fig. 5 of Bejarano and Jin (2008), where the green and blue lines are thermocline depth anomaly index (unit: 10 m) and its derived RDI (unit: 2.5 × 10−6) that are averaged over 140°E–90°W and the SSTA indices over 180°–90°W.

  • View in gallery

    Normalized indices: (a) CPI, WPI, and EMI; (b) CTI and EPI; and (c) HF CTI and WPI. All indices are subject to 3-month running mean after the normalization. Yellow lines correspond to one standard deviation of the indices. Gray number pairs in (a) or (b) are correlation coefficients between the indices with and without the decadal, respectively, while the gray number in (c) is correlation between the detrended indices. All of correlations are obtained by using the indices before the running mean.

  • View in gallery

    Evolution of SSTA (K) (shading) and zonal wind stress anomalies (vectors, 0.01 N m−2) regressed (top) upon the detrended original (left) CTI and (right) WPI and (middle) by their HF indices. Both fields are averaged over 5°S–5°N. Ordinates are from lead 24 months to lag 24 months to event peak. Vectors smaller than 10% of the reference size are masked out. (bottom) SSTA (solid lines) and zonal wind stress anomalies (dashed lines) at lag 0 from the middle panels are shown. SODA data are used for wind stress.

  • View in gallery

    Regressions of SSHA (shading, 5 × 10−3 m) and Ug anomalies (vectors, 5 × 10−3 m s−1) upon the detrended (a) CTI; (b) WPI; and (c),(d) HF WPI, where (a)–(c) are made by using SODA data and (d) GODAS data. An average over 5°S–5°N is used. Vectors smaller than 10% of the reference size are masked out; ordinates are lag months.

  • View in gallery

    As in Fig. 4 but for SSHA (shading, 5 × 10−3 m) and Vg anomalies (vectors, 5 × 10−3 m s−1), where a zonal-mean over 130°E–90°W is used. Abscissas are lag months.

  • View in gallery

    Time evolutions of the zonal-mean SSHA (solid lines, 2 × 10−3 m), anomalous Ug (solid dotted lines, 1 × 10−4 m s−1) and Vg (dot–dashed lines, 2 × 10−5 m s−1) indices, RDIs (dashed lines, 0.5 × 10−9). (a)–(d) correspond to those in Figs. 4 and 5, respectively. All indices are calculated from 5°S–5°N, 130°E–90°W, where Vg indices are obtained by subtracting the south from north of equator. Ordinates are lag months.

  • View in gallery

    Phase evolutions of the regressed SSTA (contours, 0.1 K), SSHA (shading, 5 × 10−3 m) and geostrophic current anomalies (black vectors; Ug 5 × 10−3 m s−1, Vg 2.5 × 10−3 m s−1) at eight different lag months for (a) CT ENSO and (b) WP ENSO by using the detrended HF CTI and HF WPI, respectively. Amplification factors are used to recover amplitude of fields at different lags with e-folding time of 20 months for CT and WP ENSOs. SODA data are used.

  • View in gallery

    Evolution of temperature tendency (0.01 K month−1) of the four dynamical feedback terms for (a),(c) CT ENSO and (b),(d) WP ENSO by using (a),(b) SODA and (c),(d) GODAS datasets. Fields are averaged over 5°S–5°N. Ordinates denote lag months. A zonal 31 (5) point running mean is used for SODA (GODAS) data.

  • View in gallery

    Zonal distributions of (left) ZA and (right) TH feedback terms (0.01 K month−1) at the peak phase averaged from lag −2 to 2 months in Fig. 8 for CT ENSO (dashed lines) and WP ENSO (solid lines) based on the (a) SODA and (b) GODAS datasets.

  • View in gallery

    Time evolution of symmetric (red lines) and asymmetric (blue lines) tendency parts of ZA and TH feedback terms (0.01 K month−1) averaged over the Niño-3 (solid lines) and Niño-4 (dashed lines) regions in Fig. 8 for (left) CT ENSO and (right) WP ENSO based on the (a) SODA and (b) GODAS datasets. Abscissas are lag months.

  • View in gallery

    Schematic diagrams for (a)–(d) CT ENSO and (e)–(h) WP ENSO recharge oscillator mechanisms in the (a),(e) warm, (b),(f) warm-to-cold, (c),(h) cold, and (d),(g) cold-to-warm phases. The red (blue) shadings on the top planes representing the sea surface denote positive (negative) SST anomalies, and those on the inclined planes representing the climatic thermocline denote positive (negative) subsurface temperature anomalies. The dark gray arrows denote mean upwelling,; the black arrows represent the zonal and meridional upper-ocean geostrophic current anomalies, and the solid yellow arrows stand for wind stress anomalies: W, C, H, and L denote warm, cold, high, and low, respectively.

  • View in gallery

    Metrics as a function of α (abscissa) for defining the CT and WP indices using three kinds of unified cost functions (ordinate).

  • View in gallery

    WP indices (gray lines) and their mean (black line) using different α with an internal of 0.01 from 0.3 to 0.5.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 251 251 38
PDF Downloads 196 196 27

Recharge Oscillator Mechanisms in Two Types of ENSO

View More View Less
  • 1 Department of Meteorology, School of Ocean and Earth Sciences and Technology, University of Hawaii at Manoa, Honolulu, Hawaii, and Laboratory for Climate Studies, National Climate Center, China Meteorological Administration, Beijing, China
  • | 2 Department of Meteorology, School of Ocean and Earth Sciences and Technology, University of Hawaii at Manoa, Honolulu, Hawaii
© Get Permissions
Full access

Abstract

The El Niño–Southern Oscillation (ENSO) tends to behave arguably as two different “types” or “flavors” in recent decades. One is the canonical cold-tongue-type ENSO with major sea surface temperature anomalies (SSTA) positioned over the eastern Pacific. The other is a warm-pool-type ENSO with SSTA centered in the central Pacific near the edge of the warm pool. In this study, the basic features and main feedback processes of these two types of ENSO are examined. It is shown that the interannual variability of upper-ocean heat content exhibits recharge–discharge processes throughout the life cycles of both the cold tongue (CT) and warm pool (WP) ENSO types. Through a heat budget analysis with focus on the interannual frequency band, the authors further demonstrate that the thermocline feedback plays a dominant role in contributing to the growth and phase transitions of both ENSO types, whereas the zonal advective feedback contributes mainly to their phase transitions. The westward shift of the SSTA center of the WP ENSO and the presence of significant surface easterly wind anomalies over the far eastern equatorial Pacific during its mature warm phase are the two main factors that lead to a reduced positive feedback for the eastern Pacific SSTA. Nevertheless, both the WP and CT ENSO can be understood to a large extent by the recharge oscillator mechanism.

Corresponding author address: Dr. Hong-Li Ren, School of Ocean and Earth Sciences and Technology, University of Hawaii at Manoa, 2525 Correa Rd. HIG350, Honolulu, HI 96822. E-mail: honglir@hawaii.edu

Abstract

The El Niño–Southern Oscillation (ENSO) tends to behave arguably as two different “types” or “flavors” in recent decades. One is the canonical cold-tongue-type ENSO with major sea surface temperature anomalies (SSTA) positioned over the eastern Pacific. The other is a warm-pool-type ENSO with SSTA centered in the central Pacific near the edge of the warm pool. In this study, the basic features and main feedback processes of these two types of ENSO are examined. It is shown that the interannual variability of upper-ocean heat content exhibits recharge–discharge processes throughout the life cycles of both the cold tongue (CT) and warm pool (WP) ENSO types. Through a heat budget analysis with focus on the interannual frequency band, the authors further demonstrate that the thermocline feedback plays a dominant role in contributing to the growth and phase transitions of both ENSO types, whereas the zonal advective feedback contributes mainly to their phase transitions. The westward shift of the SSTA center of the WP ENSO and the presence of significant surface easterly wind anomalies over the far eastern equatorial Pacific during its mature warm phase are the two main factors that lead to a reduced positive feedback for the eastern Pacific SSTA. Nevertheless, both the WP and CT ENSO can be understood to a large extent by the recharge oscillator mechanism.

Corresponding author address: Dr. Hong-Li Ren, School of Ocean and Earth Sciences and Technology, University of Hawaii at Manoa, 2525 Correa Rd. HIG350, Honolulu, HI 96822. E-mail: honglir@hawaii.edu

1. Introduction

The phenomena of El Niño–Southern Oscillation (ENSO) are recognized to play a crucial role in global climate variability (Rasmusson and Carpenter 1982; Ropelewski and Halpert 1987; Mason and Goddard 2001). In general, the canonical El Niño has its major center of sea surface temperatures anomalies (SSTAs) in the equatorial Pacific cold-tongue (CT) region. Recently, a number of studies reported that, in addition to this canonical El Niño type, a different type of El Niño with its major SSTA center shifted to the central Pacific by the warm-pool (WP) edge region, is becoming a common occurrence during the past 20 years (Larkin and Harrison 2005a,b; Ashok et al. 2007; Kao and Yu 2009; Kug et al. 2009). There is evidence that this ENSO type may emerge even more frequently in a warming climate as projected by the Intergovernmental Panel on Climate Change (IPCC) Fourth Assessment Report (AR4) simulations (Yeh et al. 2009). This El Niño type is accompanied by a distinct tropical atmospheric circulation pattern (Ashok and Yamagata 2009) and exhibits significantly different global climate impacts compared to CT El Niños through tropical–extratropical teleconnections (e.g., Weng et al. 2007, 2009; Kim et al. 2009; Zhang et al. 2011, 2012). So far, various definitions and nomenclatures, such as date line El Niño (Larkin and Harrison 2005a,b), El Niño Modoki (Ashok et al. 2007), central Pacific ENSO (Kao and Yu 2009; Yeh et al. 2009), and WP El Niño (Kug et al. 2009; Ren and Jin 2011, hereafter RJ11), have been given to this type of ENSO. We adopt the terminology of WP and CT El Niño/ENSO in this study to highlight the fact that the WP and CT ENSO SSTAs overlay over very different climatological background sea surface temperatures.

A number of attempts have been made to examine the dynamical processes responsible for the generation and maintenance of the WP El Niño type. Ashok et al. (2007) stressed the important role of wind-induced tropical Pacific thermocline variability in the evolution of WP El Niño events. Kug et al. (2009) accentuated that the zonal-current-driven advective feedback plays a key role in the developing phase of the WP El Niño type, whereas the thermocline feedback may be of less importance. Kao and Yu (2009) reported that phase reversal signatures are unclear for WP El Niños, whereas Yu et al. (2009, 2010) emphasized initial extratropical triggers or preconditions for the generation of the WP-type ENSO events. Further, Yu and Kim (2010a) suggested that there are three possible phase transition pathways for this El Niño type under different preconditions. However, most WP El Niño events occurred in the early 1990s and in the 2000s when the tropical Pacific decadal variability was in warm phases. This decadal modulation of the background state might have interfered, to some extent, with the basic characteristics of the WP ENSO in all previous studies. Kug et al. (2010), based on model output analyses, indicated that the dynamical feedbacks may not be crucial for the phase transition of the WP El Niño owing to a weak discharge of equatorial heat content so that a WP El Niño event is rarely followed by a cold (La Niña) event. However, most coupled general circulation models still fail to simulate both ENSO types due to relatively large model biases (Yu and Kim 2010b; Ham and Kug 2012). The basic features and dynamical processes associated with the WP ENSO flavor remain elusive and need to be further studied.

Bejarano and Jin (2008), in their theoretical study of the ENSO regime dependence on climate mean state changes, showed that there are two leading ENSO-like modes coexisting under the current climate conditions. One of the modes, termed the quasi-quadrennial (QQ) mode, has its SSTA pattern centered over the eastern equatorial Pacific (see their Fig. 7i), similar to the observed SSTA pattern of CT ENSO. The other mode, termed the quasi-biennial (QB) mode, has its SSTA center shifted westward (see their Fig. 8i), which is similar to the observed SSTA pattern of WP ENSO. Although the zonal advective feedback plays a more important role in the phase transition for the second mode, the dynamics of the two modes can be both understood to some extent by the recharge oscillator mechanism (Jin 1996, 1997a; Jin and An 1999).

Motivated by the study of Bejarano and Jin (2008), we give evidence that the 2009–10 event, the strongest WP El Niño event in recorded history, is reminiscent to their QB mode (Fig. 1). As seen in Fig. 1, this WP El Niño event exhibits a clear recharge–discharge process of upper-ocean heat content [referred to as the recharge–discharge index (RDI), defined as the zonal-mean zonal gradient of equatorial thermocline anomalies] and a fast phase transition indicated by the zonal-mean thermocline variability. In this study, we will examine the recharge oscillator mechanisms for the two ENSO types and the associated dynamical feedback in depth.

Fig. 1.
Fig. 1.

(a) Zonal-mean equatorial indices of SSTA (line with circles, K) and SSHA (solid line, 0.025 m) and recharge oscillkator index (RDI) (dashed line, 1 × 10−8) for the 2009–10 case. The indices are generated over 5°S–5°N. The SSTA index is averaged over 160°E–110°W and the SSHA index over 120°E–110°W. RDI is defined as the average of zonal gradient of SSHA over 120°E–90°W, positively proportional to meridional geostrophic current and represents the intensity of recharge–discharge of equatorial upper-ocean heat content. SSH is from the GODAS data and the climatology is defined in 1980–2010. (b),(c) As in (a) but for the QQ and QB modes, based on Fig. 5 of Bejarano and Jin (2008), where the green and blue lines are thermocline depth anomaly index (unit: 10 m) and its derived RDI (unit: 2.5 × 10−6) that are averaged over 140°E–90°W and the SSTA indices over 180°–90°W.

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00601.1

To capture the basic spatiotemporal features of the two ENSO types, we will utilize the so-called WP Niño index (WPI) and CT Niño index (CTI) developed in our recent study (RJ11). By removing the decadal signal, we will focus on interannual variability features of the two ENSO types and examine the dynamical feedback processes to determine their contributions to growth and phase transitions. This paper will be organized as follows. The utilized datasets are described in section 2. The indices and patterns for the two ENSO types are given in section 3. We examine the recharge–discharge processes of upper-ocean heat content associated with both ENSO types in section 4 and compare contributions of the different dynamical feedback processes to the growth and phase transitions in section 5. We conclude with a summary and discussion in section 6.

2. Data and ENSO indices

We use the improved extended reconstructed SST version 3b (ERSST V3b) (Smith et al. 2008) dataset from the National Climate Data Center of the National Oceanic and Atmospheric Administration. This SST data consists of monthly 2° spatial superobservations, which are defined as individual observations averaged onto a 2° horizontal grid. We will examine the period from January 1950 to February 2011.

The interior ocean temperature, current, and sea surface height variables are primarily taken from the Simple Ocean Data Assimilation (SODA) reanalysis version 2.2.4 (Giese and Ray 2011) for the period from January 1871 to December 2008. The ocean model in the data assimilation system is based on the Parallel Ocean Program (POP) version 2.0.1 (Smith et al. 1992) with a horizontal resolution of 0.25° latitude by 0.4° longitude. There are 40 vertical levels with a resolution of about 10 m in the upper 100 m. This global ocean model is forced by an extended atmospheric forcing field from a new NOAA reanalysis dataset [the Twentieth-Century reanalysis, version 2 (20CRv2)] (Whitaker et al. 2004; Compo et al. 2006) for the period 1871–2008.

In addition, the assimilation products generated from the Global Ocean Data Assimilation System (GODAS) (Behringer and Xue 2004) from the National Centers for Environmental Prediction (NCEP) are used for comparison with the SODA-data-based results. The GODAS analysis is available at a ⅓° × ⅓° horizontal resolution in the tropics for the period from January 1980 to February 2011.

Traditional Niño-3 and Niño-4 indices from January 1950 to February 2011 are directly obtained online (http://www.cpc.noaa.gov/data/indices/sstoi.indices) from the NOAA/Climate Prediction Center: the indices are defined as SSTA averages over the Niño-3 region (5°S–5°N, 150°–90°W) and Niño-4 region (5°S–5°N, 160°E–150°W). In this study, we remove the long-term linear trend from all indices before further analyses.

The two traditional Niño indices, Niño-3 and Niño-4, have been extensively used to quantify the ENSO phenomenon. A combination of these two, the Niño-3.4 index defined by the SSTA averaged in the Niño-3.4 region (5°S–5°N, 170°–120°W) (Trenberth 1997), is widely used in ENSO research and operational climate monitoring. Since the SSTA patterns of the ENSO types are highly correlated, neither of the two traditional indices alone can characterize the WP ENSO independently. All three indices capture the broad-scale nature of the ENSO-related SSTA and, thus, exhibit the main signals of the different ENSO types. Therefore, based on a transformation of Niño-3 and Niño-4 indices, RJ11 proposed WPI and CTI (refer to the appendix for details) to describe the two different ENSO types. These two indices can be used effectively to delineate time evolutions and extract characteristic patterns for both ENSO types.

We also utilize other indices proposed previously to represent the different ENSO flavors. One is the El Niño Modoki index (EMI) proposed by Ashok et al. (2007) to replicate the second empirical orthogonal function (EOF) of tropical Pacific SSTAs, which is defined as [SSTA]C − 0.5[SSTA]E − 0.5[SSTA]W, where brackets denote spatial averages in the central (C: 165°E–140°W, 10°S–10°N), eastern (E: 110°–70°W, 15°S–5°N), and western (W: 125°–145°E, 10°S–20°N) areas, respectively. This index is somewhat similar to the trans-Niño index proposed by Trenberth and Stepaniak (2001). Another is the central Pacific ENSO index (CPI) proposed by Kao and Yu (2009), who realized that the second SSTA EOF is not quite adequate to describe the main features of the different ENSO flavors independently and devised a somewhat more complicated index. CPI is also based on an EOF decomposition of data with SSTA related to the Niño-1+2 index, which is defined at (10°S–0°, 80°–90°W), linearly removed. Similarly, an eastern Pacific ENSO index (EPI) is defined by linearly removing the Niño-4-index-related SSTA. All of these somewhat related indices utilize SSTA beyond the Niño-3 and Niño-4 regions. Overall all these new indices, despite being derived from different definitions, are essentially describing the same phenomenon.

3. Removal of the decadal signal from the ENSO indices

Figure 2a shows time evolutions of the normalized WPI, EMI, and CPI. All of these indices represent the temporal characteristics of the WP ENSO type and exhibit more or less the same characteristics in the last six decades. The correlation between WPI and EMI amounts to 0.86 and to 0.80 between WPI and CPI. These indices exhibit strong interannual and decadal variability besides a weak long-term linear trend. Their spectral coherence seems to be partly due to the decadal variability. Another main common feature found among the indices is the clear interannual variability superimposed on the slower-varying decadal signal. Weng et al. (2007) have noted that EMI is dominated by decadal time-scale variability, whereas Kao and Yu (2009) reported that the CP ENSO shows a dominant period near the 2-yr band. The considerable divergence of these conclusions reflects that the SSTA in the tropical central Pacific exhibit multi-time-scale characteristics. Therefore, to solely focus on the interannual variability of the WP ENSO type, we filtered out the decadal signals from our indices. We first produce low-pass filtered (LPF) WPI and CTI with a Gaussian filter, where the spectrum cutoff is chosen at 6 yr for WPI and 8 yr for CTI, and then subtract the LPF indices from the original indices to obtain high-frequency (HF) WP and CT indices (Fig. 2c).

Fig. 2.
Fig. 2.

Normalized indices: (a) CPI, WPI, and EMI; (b) CTI and EPI; and (c) HF CTI and WPI. All indices are subject to 3-month running mean after the normalization. Yellow lines correspond to one standard deviation of the indices. Gray number pairs in (a) or (b) are correlation coefficients between the indices with and without the decadal, respectively, while the gray number in (c) is correlation between the detrended indices. All of correlations are obtained by using the indices before the running mean.

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00601.1

In Fig. 2c, both HF WPI and HF CTI clearly feature variability on interannual time scales without any background decadal signal. Their correlation amounts to 0.11, indicating that WPI and CTI are independent from each other on interannual time scales. Furthermore, the correlations between HF WPI, HF CPI, and HF EMI are calculated (the second elements of gray number pairs in Figs. 2a,b). Their differences relative to the first elements reflect the impact of the decadal signals on the correlations. That is, the good coherence between the nonfiltered indices (Fig. 2a) is partly attributed to the decadal signal. This implies that the classification of CT El Niño, WP El Niño, and La Niña events directly based on unfiltered indices are subject to large interference from the decadal signal. In contrast, the decadal signal exhibits little impact on the canonical ENSO. Hence, we will use HF WPI and HF CTI to examine features and dynamical processes for the two types of ENSO in the present study. Compared with the SSTA pattern regressed upon WPI, the regressed SSTA pattern upon HF WPI displays weakened anomalies in the central Pacific and subtropics where the SSTAs are also related to the decadal variability (not shown).

Figure 3 shows the life cycles of the two ENSO types, where the differences between the patterns with and without the decadal signal are compared by using the original and HF indices. Clearly, the WPI-regressed pattern features a longer duration than the CTI-regressed one, an eastward extension in the developing phase, and a westward retreat in the decaying phase. In contrast, with the decadal signal removed, the WP ENSO exhibits a much reduced duration and evolves in a similar way to the CT ENSO, where the SSTAs diminish in the western Pacific but show almost no change in the eastern Pacific. It is evident that the distinct phase transition characteristics of the WP ENSO type have become much clearer after filtering out the decadal variability. To examine atmospheric wind responses to the SSTA, the regression patterns of anomalous zonal wind stress are computed (Fig. 3). For the CT ENSO type, the surface westerly wind anomalies dominate over the central Pacific during the mature positive ENSO phase and only weak anomalous easterly winds are found in the far eastern Pacific. In contrast, for the WP ENSO type, the center of the westerly wind anomalies shifts to the west of the date line and significant easterly wind anomalies occur over the eastern Pacific east of 150°W during the peak ENSO phase. These features are highlighted in Fig. 3c. The westward displacements of SSTA and wind stress patterns are the primary features of the WP ENSO type. Later, we will show that the westward shift of the SSTA pattern center and the presence of large easterly wind anomalies in the eastern Pacific are crucial for the reduction of SSTA growth in the eastern Pacific during WP ENSO events.

Fig. 3.
Fig. 3.

Evolution of SSTA (K) (shading) and zonal wind stress anomalies (vectors, 0.01 N m−2) regressed (top) upon the detrended original (left) CTI and (right) WPI and (middle) by their HF indices. Both fields are averaged over 5°S–5°N. Ordinates are from lead 24 months to lag 24 months to event peak. Vectors smaller than 10% of the reference size are masked out. (bottom) SSTA (solid lines) and zonal wind stress anomalies (dashed lines) at lag 0 from the middle panels are shown. SODA data are used for wind stress.

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00601.1

4. Recharge oscillator mechanisms of the two ENSO types

For the canonical ENSO cycle, the recharge oscillator mechanism that depicts the recharge and discharge of the equatorial upper-oceanic heat content (HC) has been proposed to be responsible for the transition process between warm and cold ENSO phases (Jin 1997a,b). This recharge mechanism was further extended to include two key dynamical feedbacks (Jin and An 1999): that is, one is the so-called zonal advective feedback that represents the zonal advection process of climate-mean temperature by anomalous zonal geostrophic current (e.g., Picaut et al. 1997) and the other is the so-called thermocline feedback that represents the vertical advection process of subsurface temperature anomalies by climate-mean upwelling (Suarez and Schopf 1988; Battisti and Hirst 1989; Jin 1997a,b). The two feedbacks can be linked dynamically through the geostrophic balance between thermocline depth and ocean current and contribute positively to the growth and phase transition of ENSO (Jin and An 1999; An and Jin 2001). The recharge (discharge) of equatorial HC during La Niña (El Niño) phase, which corresponds to a convergence (divergence) of the meridional geostrophic current, forms a zonally uniform positive (negative) excursion of the equatorial thermocline leading the following El Niño (La Niña) by a phase of 90°, which is accompanied by an eastward (westward) equatorial zonal geostrophic current. This typical characteristic of the recharge oscillator mechanism involving the zonal advective and thermocline feedback processes was demonstrated to operate for both the QQ and QB modes by Bejarano and Jin (2008), as also seen in Fig. 1. Following Bejarano and Jin, here we examine the typical characteristics of the recharge oscillation and the roles of these two dynamical feedbacks in the WP ENSO evolution compared to the CT ENSO.

In the first-order approximation, the thermocline variations can be well represented by sea surface height anomalies (SSHAs) or the upper-ocean HC that is defined by vertically integrated ocean temperature through the upper 300 m. Here, SSHAs are used to represent the thermocline variation and, following the approach of Jin and An (1999), anomalous zonal and meridional geostrophic currents (Ug and Vg) are estimated from the meridional and zonal gradients of SSHAs, respectively. A positive (negative) zonal gradient of SSHA can lead to a poleward (equatorward) heat transport by the divergence (convergence) of the meridional geostrophic current (Meinen and McPhaden 2000). Figures 4 and 5 show time evolutions of anomalous SSH, Ug, and Vg regressed upon the CT and WP ENSO indices. The former is the time–longitude cross section of the equatorial-mean SSHA and Ug, thereby examining the zonal ocean heat (or warm water) exchange between the east and west Pacific and the latter is the time–latitude cross section of zonal-mean SSHA and Vg, thereby examining the meridional heat exchange between the equatorial and off-equatorial regions.

Fig. 4.
Fig. 4.

Regressions of SSHA (shading, 5 × 10−3 m) and Ug anomalies (vectors, 5 × 10−3 m s−1) upon the detrended (a) CTI; (b) WPI; and (c),(d) HF WPI, where (a)–(c) are made by using SODA data and (d) GODAS data. An average over 5°S–5°N is used. Vectors smaller than 10% of the reference size are masked out; ordinates are lag months.

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00601.1

Fig. 5.
Fig. 5.

As in Fig. 4 but for SSHA (shading, 5 × 10−3 m) and Vg anomalies (vectors, 5 × 10−3 m s−1), where a zonal-mean over 130°E–90°W is used. Abscissas are lag months.

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00601.1

The CT ENSO case (Figs. 4a and 5a) shows a typical recharge–discharge process of upper-ocean HC: that is, a basinwide meridional heat exchange between the equatorial and off-equatorial regions and an accompanied basin-scale zonal heat transport between the western and eastern equatorial Pacific. Notably, Vg is poleward (equatorward) with a positive (negative) zonal gradient of SSHA and Ug is eastward (westward) when the equatorial SSHA are greater (less) than the off-equatorial SSHA. In particular, Vg (Ug) reaches a maximum when the zonal (meridional) contrast of SSHA is the largest. We observe that Ug becomes uniformly eastward when the equatorial SSHA are larger than the off-equatorial SSHA with about a 12-month lead time to event peak. About 4 months later, Vg switches from an equatorward to a poleward current as the zonal gradient of the SSHA changes from negative to positive sign, indicating the beginning of the discharge process of equatorial HC. Then, Ug reaches its maximum when the zonal-mean equatorial SSHA reach a maximum at around a 4-month lead time to event peak and Vg reaches its peak poleward velocities at lag 0 month. Afterward, Ug reverses its sign at about lag 4 months to event peak. Then, for lag 10 months, Vg reverses its sign when the recharge process of equatorial HC begins, and Ug also reaches the peak westward velocity at this time. Almost the same regressed patterns are obtained using HF CTI (not shown).

The patterns for the WP ENSO evolution (Figs. 4b and 5b) based on the original nonfiltered WPI appear to be quite different from those of the CT ENSO type (Figs. 4a and 5a). The zonal contrast pattern of SSHA exhibits a long duration, which yields a weak phase transition. Interestingly, the discharge process in terms of the poleward zonal-mean Vg can still be clearly observed during the warm phase. It is overall difficult to see a clear phase transition signature and an accompanying recharge–discharge process because the ENSO signal is obscured by strong decadal variability. In contrast, the HF-WPI-regressed patterns (Figs. 4c and 5c) are very similar to the CT ENSO patterns (Figs. 4a and 5a), but quite different from the unfiltered WP ENSO patterns (Figs. 4b and 5b). It suggests that the influence of the decadal signal on the WP ENSO features may be greater than the distinctness between the two ENSO types. The largest feature in the distinctness is the westward shift of the positive SSHA center for the WP ENSO.

In Figs. 4c and 5c, the timings of the SSHA and current changes are also identified. The eastward Ug (poleward Vg) is set up when the meridional (zonal) contrast of the equatorial SSHA reverses at around lead 14 (10) months to event peak, then reaches its peak amplitude when the equatorial SSHA reach a maximum in the meridional (zonal) direction at lead 4 (0) months, and afterward reverses when the equatorial SSHA reverses again the sign of its meridional (zonal) contrast at about lag 2 (14) months. Then, Ug reaches its maximum westward amplitude at around lead of 16 month. Further, these results for the major features of the WP ENSO have been reconfirmed using the GODAS dataset (Figs. 4d and 5d). In addition, the meridional asymmetry of mass exchange represented by the zonal-mean SSHA and Vg, mentioned by Kug et al. (2003), are likewise apparent during the WP ENSO evolution.

Figure 6, similar to Fig. 1, highlights the major features of the recharge oscillation for the two ENSO types: that is, the clear recharge–discharge processes of HC as measured by RDI or Vg and the leading zonal-mean equatorial thermocline variations as expressed by the SSHA indices, which only are approximations of the real zonally uniform-distributed thermocline variations that lead strictly in quadrature (90° phase shift) to the SSTA indices (Jin 1997a; Meinen and McPhaden 2000). Overall, the intensity of the recharge–discharge process (measured by RDI or Vg) during the WP ENSO evolution is evidently smaller compared to the CT ENSO evolution. This weakening, on one hand, is because the climatological variance of SSHA is smaller in the central Pacific than in the eastern Pacific (not shown). On the other hand, since the center of the recharge–discharge process shifts to the west for the WP ENSO evolution, a weak additional recharge–discharge process of opposite sign occurs in the far eastern Pacific and acts to partly cancel the zonal-mean meridional exchange of HC, as also noted by Kug et al. (2010).

Fig. 6.
Fig. 6.

Time evolutions of the zonal-mean SSHA (solid lines, 2 × 10−3 m), anomalous Ug (solid dotted lines, 1 × 10−4 m s−1) and Vg (dot–dashed lines, 2 × 10−5 m s−1) indices, RDIs (dashed lines, 0.5 × 10−9). (a)–(d) correspond to those in Figs. 4 and 5, respectively. All indices are calculated from 5°S–5°N, 130°E–90°W, where Vg indices are obtained by subtracting the south from north of equator. Ordinates are lag months.

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00601.1

The above results suggest that the recharge oscillator mechanism operates for the WP ENSO type. To further contrast the WP and CT ENSO life cycles, we sample the regressed patterns of SSTA, SSHA, and anomalous geostrophic currents upon the HF WP and HF CT indices at the different lag months relative to event peak (Fig. 7), where an empirical constant factor is applied to amplify the magnitude of the patterns at large lag times. Based on the ENSO durations in Fig. 3, we approximate an e-folding time scale of 20 months and define the amplification factor as ei/M, where i is the lead/lag month and M = 20.

Fig. 7.
Fig. 7.

Phase evolutions of the regressed SSTA (contours, 0.1 K), SSHA (shading, 5 × 10−3 m) and geostrophic current anomalies (black vectors; Ug 5 × 10−3 m s−1, Vg 2.5 × 10−3 m s−1) at eight different lag months for (a) CT ENSO and (b) WP ENSO by using the detrended HF CTI and HF WPI, respectively. Amplification factors are used to recover amplitude of fields at different lags with e-folding time of 20 months for CT and WP ENSOs. SODA data are used.

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00601.1

Overall, the WP ENSO and CT ENSO types exhibit similar spatial patterns during their transition phases but different patterns during their peak phases as the former exhibits the westward-shifted SSTA center and the stronger easterlies in the far eastern Pacific compared to the latter. During the negative (positive) ENSO phases, the recharge (discharge) of basinwide equatorial ocean HC is clearly captured as represented by the convergence (divergence) of the near-equatorial Vg. As a result, the zonally uniform positive (negative) equatorial thermocline excursions and accompanied Ug on the equator are formed, which lead the observed SSTA peak. These are typical features of the recharge oscillator mechanism. For the WP ENSO type, a variation of this mechanism is clearly visible with a westward-shifted recharge–discharge process center relative to the CT ENSO type, which is probably a consequence of the strong equatorial wind stress anomalies over the far eastern Pacific during the peak ENSO phases. In other words, the westward-shifted SSHA center can be dynamically balanced by the strong surface wind anomalies in the equatorial eastern Pacific.

5. Heat budget analysis

Relative contributions of different physical processes to the SST thermodynamics associated with ENSO have been examined through ocean mixed layer heat budget analysis using model outputs and oceanic analysis datasets (e.g., An et al. 1999; Kang et al. 2001; Jin et al. 2006; Zhang et al. 2007; Kug et al. 2009, 2010). In this study, heat budget analyses, based on the two oceanic reanalysis datasets, are performed to investigate the relative importance of different dynamical feedbacks in the SSTA evolution for the two different types of ENSO.

The mixed layer averaged temperature tendency equation can be generally expressed as
e1
where an overbar denotes a climatological mean and a prime its departure from it (anomaly); T, u, υ, and w denote oceanic temperature, zonal current, meridional current, and vertical velocity, respectively. The last terms, Q and R, denote the thermal forcing and residual terms, which are not considered in this study. The mean upwelling advection in Eq. (1) can be further decomposed in an approximation following (Jin and Neelin (1993); An et al. 1999; Zhang et al. 2007):
e2
where H is the effective mean mixed layer depth for the vertical advection (constant 50 m in this study). Subscript “sub” denotes a subsurface-layer average between 50 and 100 m. The first term in Eq. (2) is often referred to as the thermocline feedback (Jin and An 1999). All dynamical terms in Eq. (1) are first estimated using the monthly reanalysis datasets and then their evolution patterns from 24-month lead to 24-month lag are obtained by regressing them upon the HF WP and CT indices.
Following Jin et al. (2006), we regroup these terms into six feedback terms as follows:
e3
where
e4
e5
e6
e7
e8
e9
Here, MC denotes the effect of mean circulation, ZA the zonal advective feedback, EK the Ekman pumping feedback, TH the thermocline feedback, NDH the nonlinear dynamical heating, and TD denotes the thermodynamical damping. The ZA, EK, and TH terms, as the three major dynamical feedbacks (cf., Jin and Neelin 1993; Jin and An 1999), all tend to make positive contributions to the growth of ENSO with TH being the largest term (Jin et al. 2006). The NDH term acts to generate an asymmetry (skewness) of SSTA amplitude between El Niño and La Niña (Jin et al. 2003; An and Jin 2004).

In this study, we focus on the first four linear terms and examine their contributions to the growth and phase transitions of the two types of ENSO. We note that the WPI-related tendency patterns are not confined to the western equatorial Pacific. To fully understand the WP ENSO evolution through heat budget analysis, we need to examine the SSTA tendency patterns across the entire equatorial Pacific.

In Fig. 8, the MC terms, which are overall negative during ENSO peak phase, generally serve as a negative feedback for both ENSO types. The EK terms are relatively small and appear less important, although they also show a weak positive feedback to the growth of the two types of ENSO. Overall, the features of the MC and EK terms are nearly the same in the two datasets, except that the mature-phase EK term patterns in the eastern Pacific are slightly different for the WP ENSO type. The ZA and TH are the two most robust terms, serving in the phase transitions of both ENSO types. The ZA term exhibits a clear 90° phase shift in both CT and WP ENSO cycles relative to the SSTA peak, indicating that it contributes to the phase transition. In the GODAS dataset, the ZA term seems to be somewhat stronger for the WP ENSO type than in the SODA dataset. This difference reflects the common problem that the ocean advection estimation in reanalysis datasets is still associated with large uncertainties. The contributions of the ZA term to the growth rate of both CT and WP ENSO types seem to be small in both datasets. The TH term is the dominant term and leads the peak time by less than 90° during both CT and WP ENSO evolution, indicating that it contributes largest to both ENSO growth and phase transition.

Fig. 8.
Fig. 8.

Evolution of temperature tendency (0.01 K month−1) of the four dynamical feedback terms for (a),(c) CT ENSO and (b),(d) WP ENSO by using (a),(b) SODA and (c),(d) GODAS datasets. Fields are averaged over 5°S–5°N. Ordinates denote lag months. A zonal 31 (5) point running mean is used for SODA (GODAS) data.

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00601.1

To compare directly the roles of the ZA with the TH terms in contributing to the growth of ENSO, we examine zonal distributions of the equatorial tendency averaged at peak phase. As shown in Fig. 9, the ZA terms are much less important for the growth of ENSO than the TH terms owing to their relatively small amplitude, independent of the sign difference between the two datasets. The TH term dominates not only in the growth of the CT ENSO but also in that of the WP ENSO in both datasets. The tendency pattern of the WP ENSO type, relative to the CT ENSO type, shifts to the west and its eastern part is largely suppressed over the eastern equatorial Pacific region because of the eastern Pacific easterly wind anomalies.

Fig. 9.
Fig. 9.

Zonal distributions of (left) ZA and (right) TH feedback terms (0.01 K month−1) at the peak phase averaged from lag −2 to 2 months in Fig. 8 for CT ENSO (dashed lines) and WP ENSO (solid lines) based on the (a) SODA and (b) GODAS datasets.

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00601.1

To more clearly depict the contributions of the ZA and TH terms to the growth and phase transition of the two ENSO types, we decompose the tendency patterns, which evolve with lag time in Fig. 8 into symmetric and asymmetric parts with respect to 0 lag. These two parts represent the contributions of the feedback terms to the growth and phase transition of ENSO, respectively. Figure 10 shows these two parts of the tendencies as a function of lag time to event peak. All of the asymmetric curves, averaged over either the Niño-3 or Niño-4 region, exhibit a positive peak before 0 lag and a negative peak after, indicating clearly that both ZA and TH terms contribute to the phase transition for both ENSO types. However, only the TH-related symmetric parts show a positive contribution to their growths. In particular, the amplitude of such a contribution is almost equal in the two Niño regions for the WP ENSO type but quite different for the CT ENSO type. This reflects that the TH term serves for the WP ENSO growth identically in both Niño-3 and Niño-4 regions. In contrast, the symmetric curves of the ZA terms exhibit less amplitude.

Fig. 10.
Fig. 10.

Time evolution of symmetric (red lines) and asymmetric (blue lines) tendency parts of ZA and TH feedback terms (0.01 K month−1) averaged over the Niño-3 (solid lines) and Niño-4 (dashed lines) regions in Fig. 8 for (left) CT ENSO and (right) WP ENSO based on the (a) SODA and (b) GODAS datasets. Abscissas are lag months.

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00601.1

Overall, the TH term makes dominant contributions to the growth and phase transitions for both ENSO types, while the ZA term plays a significant role in their phase transitions. In addition, the results of a heat budget analysis as shown in Figs. 810 are consistent with those in Figs. 37, such as the westward-shifted SSHA center and the westward tendency centers in the mature phase for the WP ENSO compared to the CT ENSO, as well as the coincident signs of the zonal geostrophic current anomalies and the ZA tendencies estimated from ocean currents.

6. Summary and discussion

Evidence exists to support the hypothesis that two different types of ENSO coexist in our climate regime: the cold-tongue (CT) ENSO with maximum SSTA variability located in the eastern equatorial Pacific and the warm-pool (WP) ENSO with major SSTA variability located in the central Pacific at the edge of the WP. A number of studies have revealed that this WP ENSO type exhibits some distinct spatiotemporal features and dynamics from the CT ENSO. In this study, motivated by the study of Bejarano and Jin (2008) in which they found two leading ENSO-like coupled modes coexisting under current climate condition and the striking similarity of the two modes to the observed ENSO types, we examined the observed features and physical processes associated with the recharge oscillator mechanisms for the two ENSO types.

To contrast the observed features, we compared the indices for the two ENSO types and found that all indices for the WP ENSO show strong decadal variability besides the interannual variability that ENSO dominates. To focus on the interannual variability, we first removed the decadal signal and then examined the spatiotemporal characteristics and dynamics for the two ENSO types. We showed a clear recharge–discharge process of ocean heat content throughout the life cycle for the WP ENSO. The mixed layer heat budget analyses further indicated that both thermocline feedback and zonal advective feedback play important roles in the phase transitions of both ENSO types, where the former is the dominant contributor to the growth rate for both ENSO types and the latter contributes little to the growth rate. Regrettably, the oceanic reanalysis is known for poor estimates of oceanic currents, which may add an uncertainty to our conclusions regarding the zonal advective feedback.

To test the sensitivity of our conclusions regarding the applicability of the recharge oscillator to the WP ENSO type, we conducted additional analyses using other indices (HF EMI and HF CPI) and a composite analysis sampling typical WP El Niño events. All of these results (not shown) are quite similar to those above, indicating well that our main conclusions are not dependent on the definitions of the indices used and the contributions of the La Niña phase. Also, a natural question is why the recharge oscillation for the WP ENSO has not been detected in the previous studies (e.g., Kug et al. 2009). Our current study shows that the strong background decadal signal has interfered with, apparently, the recharge–discharge process and hence biased their conclusion.

The results in this study indicate that the WP ENSO can be understood to a large extent by a variation of the recharge oscillator theory. Recent studies showed indirect evidence in support of this conclusion. For example, McPhaden (2012) recently pointed out that the zonal-mean warm water volume (WWV) index that is based on the recharge oscillator mechanism is becoming less leading the Niño-3.4 SSTA index since 2000 when the WP El Niño events occurred frequently. Dewitte et al. (2012) also showed that the total WWV index corresponding to the first two oceanic baroclinic waves of the equatorial Pacific significantly lead the modified Niño-4 index for the WP type of events. Based on the results in this study that the ZA and TH terms play similar roles between the two ENSO types, except for the zonally different center positions of the SSTA pattern and associated zonal wind patterns, a conceptual diagram for delineating the recharge oscillator mechanism of the WP ENSO along with the original recharge oscillator model is shown in Fig. 11. Figures 11a–d display the four phases for the CT ENSO evolution, based on the recharge oscillator diagram of Jin and An (1999). Noting that the ZA term exhibits an uncertain contribution to the growth of ENSO during its peak phases, we thus express the zonal (geostrophic) current anomalies by dashed arrows (Figs. 11a,c).

Fig. 11.
Fig. 11.

Schematic diagrams for (a)–(d) CT ENSO and (e)–(h) WP ENSO recharge oscillator mechanisms in the (a),(e) warm, (b),(f) warm-to-cold, (c),(h) cold, and (d),(g) cold-to-warm phases. The red (blue) shadings on the top planes representing the sea surface denote positive (negative) SST anomalies, and those on the inclined planes representing the climatic thermocline denote positive (negative) subsurface temperature anomalies. The dark gray arrows denote mean upwelling,; the black arrows represent the zonal and meridional upper-ocean geostrophic current anomalies, and the solid yellow arrows stand for wind stress anomalies: W, C, H, and L denote warm, cold, high, and low, respectively.

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00601.1

The recharge oscillator model for the WP ENSO type is similar to that for the CT ENSO. The schematic diagram (Figs. 11e–h) exhibits the westward shift of SSTA, surface wind anomalies, and thermocline variation patterns during the peak phases and the presence of strong surface wind anomalies over the far eastern equatorial Pacific. The shifted SSTA and wind patterns are crucial for the reduction of SSTA growth in the eastern Pacific and, hence, maintain the characteristic WP ENSO pattern. This is the main factor that makes the WP ENSO different from the CT ENSO.

Our results on the roles of the ZA term in contributing to the growth and phase transition during the entire ENSO life cycle are consistent with previous studies. For example, Wang and McPhaden (2001), in an observational analysis, showed that the anomalous zonal advection term is of particular importance in the onset and development phases of ENSO. Zhang et al. (2007) found that the ZA term prefers to act as a transition driver and that the TH term contributes to both the growth and phase transition of ENSO. Kug et al. (2009), as shown in their Fig. 10, also found that the ZA term plays an important role in the development phase of the WP El Niño. However, the role of the ZA term in driving the SSTA growing at the peak phase is still unclear for both the WP and CT ENSO, based on the results obtained from the model simulations and reanalysis datasets (e.g., Zhang et al. 2007; this study). Moreover, the importance of the TH term for the WP ENSO, as revealed in this study, appears to be different from that by Kug et al. (2009, 2010). The major reason is that the TH term here has a different definition from their studies in which it was directly defined as the vertical advection of mixed layer temperature anomalies by mean upwelling. Their definition actually underestimated the TH term because it involves the negative feedback by the mean upwelling damping ().

This study has suggested that the WP ENSO operates as a variation of the classical recharge oscillator mechanism in the sense that both ZA and TH terms contribute to the growth and phase transition as for the CT ENSO. Either of the two ENSO types will have a tendency to emerge under the particular initial values, preconditions, or background states, which needs to be studied in the future. The theoretical results of Bejarano and Jin (2008) suggest that the two independent but similar modes, with likeness of the CT and WP ENSO, can coexist under current climate conditions. Thus, we contend that both CT and WP ENSO may be named as a different type of interannual modes of variability.

Our current focus is only on understanding how the recharge oscillator mechanism (Jin 1997a,b; Jin and An 1999) operates for the WP ENSO. Indeed, there are other dynamical mechanisms in the literature depicting the different paradigms for ENSO: for example, the delayed oscillator (Suarez and Schopf 1988; Battisti and Hirst 1989), the coupled wave oscillator (Cane et al. 1990), the advective–reflective oscillator (Picaut et al. 1997), and the western Pacific oscillator (Weisberg and Wang 1997). For the WP ENSO, besides the apparent difference of wind anomalies over the eastern Pacific, differences also exist over the western Pacific as well as the eastern and western boundaries, indicating the possibility that the other dynamical oscillator mechanisms may play roles in WP ENSO, which will be addressed in future studies.

Acknowledgments

This work is jointly supported by National Science Foundation (NSF) Grants ATM 1034798, the 973 Program of China (2010CB950404), DOE Grant DE-SC0005110, NOAA Grant NA10OAR4310200, NSF of China (NSFC) Grant 40805028, and the China Meteorological Special Project (GYHY201206016 and GYHY200906015). The authors especially appreciate M. F. Stuecker, W. Zhang, A. Levine, and J. Boucharel for their careful reading and revising of the manuscript.

APPENDIX

Definitions of Two Niño Indices for CT and WP ENSO

By using a coordinate transform in the Niño-3–Niño-4 phase space, RJ11 defined the CT Niño index and WP Niño index (NCT and NWP) as follows:
ea1
Indices N3 and N4 denote Niño-3 and Niño-4 indices, respectively, and NCT and NWP are a piecewise linear combination of N3 and N4 conditioned by the ENSO phase. The transformation parameter α can be determined by minimizing a cost function defined by the total metrics of NCT over the WP El Niño time and NWP over the CT El Niño time. However, one key question is how to determine a priori the time of occurrence of the WP/CT El Niño. This is discussed here in terms of different schemes. RJ11 has directly used a method of cluster analysis designed initially by Kug et al. (2009) to identify the WP (CT) El Niño month once N4 (N3) of this month is greater than the other and both greater than 0.5°C. The cost function for this minimization scheme is written as
ea2
where the timing of WPEN and CTEN is always fixed with α changed. One can search α in a broad parameter domain to make
ea3
Here α* is the optimal. Figure A1 presents the J (RJ11) as a function of α, where α* ≈ 0.42. In this study, the training period for determining α is January 1951–February 2011.
Fig. A1.
Fig. A1.

Metrics as a function of α (abscissa) for defining the CT and WP indices using three kinds of unified cost functions (ordinate).

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00601.1

To test the impacts of different schemes on valuing α*, a dynamic scheme is designed in contrast to the static scheme of RJ11. That is, with α changed, we redetermine the timing of the WPEN and CTEN by taking one standard deviation of the newly generated indices from Eq. (A1) as the criterion for identification. The new J curve is also plotted in Fig. A1 as a contrast, where α* ≈ 0.41. So far, only the signatures of the determination of the two types of El Niño have been utilized for valuing α*. This is because a natural separation has been observed between the two clusters in N3N4 phase space that correspond to the two types of El Niño, as shown in Fig. 1a of RJ11, whereas no such separation exists for the La Niña case. This is also why La Niña is difficult to separate into two clear types even though the transformation in Eq. (A1) is used. A question is whether there is a possibility to separate both the El Niño and La Niña through varying α. Therefore, we simply take the correlation between NWP and NCT as the cost function of α. When these two indices are irrelevant (viz., zero correlation), α* = 0.45 is easily obtained by the J curve (JCORR) in Fig. A1. However, in this case, the separation makes it difficult to represent exactly some observed El Niño events despite the fact that La Niña is not well separated yet (not shown).

Based on this comparison, it appears that the dynamic scheme is much simpler and more effective. Figure A2 further tests the sensitivity of transformed indices to α. It is clear that the WP index is not sensitive to α, indicating that the transformed indices can be widely applied.

Fig. A2.
Fig. A2.

WP indices (gray lines) and their mean (black line) using different α with an internal of 0.01 from 0.3 to 0.5.

Citation: Journal of Climate 26, 17; 10.1175/JCLI-D-12-00601.1

REFERENCES

  • An, S.-I., , and F.-F. Jin, 2001: Collective role of thermocline and zonal advective feedbacks in the ENSO mode. J. Climate, 14, 34213432.

    • Search Google Scholar
    • Export Citation
  • An, S.-I., , and F.-F. Jin, 2004: Nonlinearity and asymmetry of ENSO. J. Climate, 17, 23992412.

  • An, S.-I., , F.-F. Jin, , and I.-S. Kang, 1999: The role of zonal advection feedback in phase transition and growth of ENSO in the Cane-Zebiak model. J. Meteor. Soc. Japan, 77, 11511160.

    • Search Google Scholar
    • Export Citation
  • Ashok, H., , and T. Yamagata, 2009: Climate change: The El Niño with a difference. Nature, 461, 481484, doi:10.1038/461481a.

  • 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
  • Battisti, D. S., , and A. C. Hirst, 1989: Interannual variability in a tropical atmosphere–ocean model: Influence of the basic state, ocean geometry, and nonlinearity. J. Atmos. Sci., 46, 16871712.

    • Search Google Scholar
    • Export Citation
  • Behringer, D. W., , and Y. Xue, 2004: Evaluation of the global ocean data assimilation system at NCEP: The Pacific Ocean. Preprints, Eighth Symp. on Integrated Observing and Assimilation Systems for Atmosphere, Oceans, and Land Surface, Seattle, WA, Amer. Meteor. Soc., 2.3. [Available online at https://ams.confex.com/ams/pdfpapers/70720.pdf.]

  • Bejarano, L., , and F.-F. Jin, 2008: Coexistence of equatorial coupled modes of ENSO. J. Climate, 21, 30513067.

  • Cane, M. A., , M. Münnich, , and S. E. Zebiak, 1990: A study of self-excited oscillations of the tropical ocean–atmosphere system. Part I: Linear analysis. J. Atmos. Sci., 47, 15621577.

    • Search Google Scholar
    • Export Citation
  • Compo, G. P., , J. S. Whitaker, , and P. D. Sardeshmukh, 2006: Feasibility of a 100-year reanalysis using only surface pressure data. Bull. Amer. Meteor. Soc., 87, 175190.

    • Search Google Scholar
    • Export Citation
  • Dewitte, B., , J. Choi, , S.-I. An, , and S. Thual, 2012: Vertical structure variability and equatorial waves during central Pacific and eastern Pacific El Niños in a coupled general circulation model. Climate Dyn., 38, 22752289, doi:10.1007/s00382-011-1215-x.

    • Search Google Scholar
    • Export Citation
  • Giese, B. S., , and S. Ray, 2011: El Niño variability in simple ocean data assimilation (SODA), 1871–2008. J. Geophys. Res., 116, C02024, doi:10.1029/2010JC006695.

    • Search Google Scholar
    • Export Citation
  • Ham, Y.-G., , and J.-S. Kug, 2012: How well do current climate models simulate two types of El Niño? Climate Dyn., 39, 383398, doi:10.1007/s00382-011-1157-3.

    • Search Google Scholar
    • Export Citation
  • Jin, F.-F., 1996: Tropical ocean-atmosphere interaction, the Pacific cold tongue, and the El Niño-Southern Oscillation. Science, 274, 7678.

    • Search Google Scholar
    • Export Citation
  • Jin, F.-F., 1997a: An equatorial ocean recharge paradigm for ENSO. Part I: Conceptual model. J. Atmos. Sci., 54, 811829.

  • Jin, F.-F., 1997b: An equatorial ocean recharge paradigm for ENSO. Part II: A stripped-down coupled model. J. Atmos. Sci., 54, 830847.

    • Search Google Scholar
    • Export Citation
  • Jin, F.-F., , and J. D. Neelin, 1993: Modes of interannual tropical ocean-atmosphere interaction—A unified view. Part I: Numerical results. J. Atmos. Sci., 50, 34773503.

    • Search Google Scholar
    • Export Citation
  • Jin, F.-F., , and S.-I. An, 1999: Thermocline and zonal advective feedbacks within the equatorial ocean recharge oscillator model for ENSO. Geophys. Res. Lett., 26, 29892992.

    • Search Google Scholar
    • Export Citation
  • Jin, F.-F., , S.-I. An, , A. Timmermann, , and J. Zhao, 2003: Strong El Niño events and nonlinear dynamical heating. Geophys. Res. Lett., 30, 1120, doi:10.1029/2002GL016356.

    • Search Google Scholar
    • Export Citation
  • Jin, F.-F., , S. T. Kim, , and L. Bejarano, 2006: A coupled-stability index for ENSO. Geophys. Res. Lett., 33, L23708, doi:10.1029/2006GL027221.

    • Search Google Scholar
    • Export Citation
  • Kang, I.-S., , S.-I. An, , and F.-F. Jin, 2001: The role of zonal advection feedback in phase transition and growth of ENSO in the Cane-Zebiak model. J. Meteor. Soc. Japan, 79, 110.

    • Search Google Scholar
    • Export Citation
  • Kao, H.-Y., , and J.-Y. Yu, 2009: Contrasting eastern-Pacific and central-Pacific types of El Niño. J. Climate, 22, 615632.

  • Kim, H.-M., , P. J. Webster, , and J. A. Curry, 2009: Impact of shifting patterns of Pacific Ocean warming on North Atlantic tropical cyclones. Science, 325, 7780, doi:10.1126/science.1174062.

    • Search Google Scholar
    • Export Citation
  • Kug, J.-S., , I.-S. Kang, , and S.-I. An, 2003: Symmetric and antisymmetric mass exchanges between the equatorial and off-equatorial Pacific associated with ENSO. J. Geophys. Res., 108, 3284, doi:10.1029/2002JC001671.

    • Search Google Scholar
    • Export Citation
  • Kug, J.-S., , F.-F. Jin, , and S.-I. An, 2009: Two types of El Niño events: Cold tongue El Niño and warm pool El Niño. J. Climate, 22, 14991515.

    • Search Google Scholar
    • Export Citation
  • Kug, J.-S., , J. Choi, , S.-I. An, , F.-F. Jin, , and A. T. Wittenberg, 2010: Warm pool and cold tongue El Niño events as simulated by the GFDL 2.1 coupled GCM. J. Climate, 23, 12261239.

    • Search Google Scholar
    • Export Citation
  • Larkin, N. K., , and D. E. Harrison, 2005a: On the definition of El Niño and associated seasonal average U.S. weather anomalies. Geophys. Res. Lett., 32, L13705, doi:10.1029/2005GL022738.

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

    • Search Google Scholar
    • Export Citation
  • Mason, S. J., , and L. Goddard, 2001: Probabilistic precipitation anomalies associated with ENSO. Bull. Amer. Meteor. Soc., 82, 619638.

    • Search Google Scholar
    • Export Citation
  • McPhaden, M. J., 2012: A 21st century shift in the relationship between ENSO SST and warm water volume anomalies. Geophys. Res. Lett., 39, L09706, doi:10.1029/2012GL051826.

    • Search Google Scholar
    • Export Citation
  • Meinen, C. S., , and M. J. McPhaden, 2000: Observations of warm water volume changes in the equatorial Pacific and their relationship to El Niño and La Niña. J. Climate, 13, 35513559.

    • Search Google Scholar
    • Export Citation
  • Picaut, J., , F. Masia, , and Y. du Penhoat, 1997: An advective-reflective conceptual model for the oscillatory nature of ENSO. Science, 277, 663666.

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

    • Search Google Scholar
    • Export Citation
  • Ren, H.-L., , and F.-F. Jin, 2011: Niño indices for two types of ENSO. Geophys. Res. Lett., 38, L04704, doi:10.1029/2010GL046031.

  • Ropelewski, C. F., , and M. S. Halpert, 1987: Global and regional scale precipitation associated with El Niño/Southern Oscillation. Mon. Wea. Rev., 115, 16061626.

    • Search Google Scholar
    • Export Citation
  • Smith, R. D., , J. K. Dukowicz, , and R. C. Malone, 1992: Parallel ocean general circulation modeling. Physica D, 60, 3861, doi:10.1016/0167-2789(92)90225-C.

    • Search Google Scholar
    • Export Citation
  • Smith, T. M., , R. W. Reynolds, , T. C. Peterson, , and J. Lawrimore, 2008: Improvements to NOAA's historical merged land–ocean surface temperature analysis (1880–2006). J. Climate, 21, 22832296.

    • Search Google Scholar
    • Export Citation
  • Suarez, M. J., , and P. S. Schopf, 1988: A delayed action oscillator for ENSO. J. Atmos. Sci., 45, 32833287.

  • Trenberth, K. E., 1997: The definition of El Niño. Bull. Amer. Meteor. Soc., 78, 27712777.

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

  • Wang, W., , and M. J. McPhaden, 2001: Surface layer temperature balance in the equatorial Pacific during the 1997–98 El Niño and 1998–99 La Niña. J. Climate, 14, 33933407.

    • Search Google Scholar
    • Export Citation
  • Weisberg, R. H., , and C. Wang, 1997: A western Pacific oscillator paradigm for the El Niño-Southern Oscillation. Geophys. Res. Lett., 24, 779782.

    • Search Google Scholar
    • Export Citation
  • Weng, H., , K. Ashok, , S. Behera, , A. S. 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
  • Weng, H., , S. K. Behera, , and T. Yamagata, 2009: Anomalous winter climate conditions in the Pacific Rim during recent El Niño Modoki and El Niño events. Climate Dyn., 32, 663674, doi:10.1007/s00382-008-0394-6.

    • Search Google Scholar
    • Export Citation
  • Whitaker, J. S., , G. P. Compo, , X. Wei, , and T. M. Hamill, 2004: Reanalysis without radiosondes using ensemble data assimilation. Mon. Wea. Rev., 132, 11901200.

    • Search Google Scholar
    • Export Citation
  • Yeh, S.-W., , J.-S. Kug, , B. Dewitte, , M.-H. Kwon, , B. P. Kirtman, , and F.-F. Jin, 2009: El Niño in a changing climate. Nature, 461, 511514, doi:10.1038/nature08316.

    • Search Google Scholar
    • Export Citation
  • Yu, J.-Y., , and S. T. Kim, 2010a: Three evolution patterns of central-Pacific El Niño. Geophys. Res. Lett., 37, L08706, doi:10.1029/2010GL042810.

    • Search Google Scholar
    • Export Citation
  • Yu, J.-Y., , and S. T. Kim, 2010b: Identification of central-Pacific and eastern-Pacific types of ENSO in CMIP3 models. Geophys. Res. Lett., 37, L15705, doi:10.1029/2010GL044082.

    • Search Google Scholar
    • Export Citation
  • Yu, J.-Y., , F. Sun, , and H.-Y. Kao, 2009: Contributions of Indian Ocean and monsoon biases to the excessive biennial ENSO in CCSM3. J. Climate, 22, 18501858.

    • Search Google Scholar
    • Export Citation
  • Yu, J.-Y., , H.-Y. Kao, , and T. Lee, 2010: Subtropics-related interannual sea surface temperature variability in the equatorial central Pacific. J. Climate, 23, 28692884.

    • Search Google Scholar
    • Export Citation
  • Zhang, Q., , A. Kumar, , Y. Xue, , W. Wang, , and F.-F. Jin, 2007: Analysis of the ENSO cycle in the NCEP coupled forecast model. J. Climate, 20, 12651284.

    • Search Google Scholar
    • Export Citation
  • Zhang, W., , F.-F. Jin, , J. Li, , and H.-L. Ren, 2011: Contrasting impacts of two-type El Niño over the western north Pacific during boreal autumn. J. Meteor. Soc. Japan, 89, 563569.

    • Search Google Scholar
    • Export Citation
  • Zhang, W., , F.-F. Jin, , H.-L. Ren, , J. Li, , and J.-X. Zhao, 2012: Differences in teleconnection over the North Pacific and rainfall shift over the USA associated with two types of El Niño during boreal autumn. J. Meteor. Soc. Japan, 90, 535552.

    • Search Google Scholar
    • Export Citation
Save