• An, S-I., , J-S. Kug, , Y-G. Ham, , and I-S. Kang, 2008: Successive modulation of ENSO to the future greenhouse warming. J. Climate, 21 , 321.

    • Search Google Scholar
    • Export Citation
  • Chang, P., 1994: A study of the seasonal cycle of sea surface temperature in the tropical Pacific Ocean using reduced gravity models. J. Geophys. Res., 99 , 77257741.

    • Search Google Scholar
    • Export Citation
  • Chavez, F. P., , P. G. Strutton, , G. E. Friederich, , R. A. Feely, , G. C. Feldman, , D. G. Foley, , and M. J. McPhaden, 1999: Biological and chemical response of the equatorial Pacific ocean to the 1997-98 El Niño. Science, 286 , 21262131.

    • Search Google Scholar
    • Export Citation
  • Chepurin, G. A., , J. A. Carton, , and D. Dee, 2005: Forecast model bias correction in ocean data assimilation. Mon. Wea. Rev., 133 , 13281342.

    • Search Google Scholar
    • Export Citation
  • Collins, M., 2000: The El Niño–Southern Oscillation in the second Hadley Centre coupled model and its response to greenhouse warming. J. Climate, 13 , 12991312.

    • Search Google Scholar
    • Export Citation
  • Durand, F., , and T. Delcroix, 2000: On the variability of the tropical Pacific thermal structure during the 1979–96 period, as deduced from XBT sections. J. Phys. Oceanogr., 30 , 32613269.

    • Search Google Scholar
    • Export Citation
  • Fedorov, A. V., , and S. G. Philander, 2001: A stability analysis of tropical ocean–atmosphere interaction: Bridging measurements and theory for El Niño. J. Climate, 14 , 30863101.

    • Search Google Scholar
    • Export Citation
  • Ji, M., , A. Leetmaa, , and J. Derber, 1995: An ocean analysis system for seasonal to interannual climate studies. Mon. Wea. Rev., 123 , 460481.

    • Search Google Scholar
    • Export Citation
  • Kessler, W. S., 1990: Observations of long Rossby waves in the northern tropical Pacific. J. Geophys. Res., 95 , 51835217.

  • Levitus, S., , J. Antonov, , and T. Boyer, 2005: Warming of the world ocean, 1955–2003. Geophys. Res. Lett., 32 , L02604. doi:10.1029/2004GL021592.

    • Search Google Scholar
    • Export Citation
  • Meehl, G. A., , P. R. Gent, , J. M. Arblaster, , B. L. Otto-Bliesner, , E. C. Brady, , and A. Craig, 2001: Factors that affect the amplitude of El Niño in global coupled climate models. Climate Dyn., 17 , 515526.

    • 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
  • Meyers, G., 1979a: On the annual Rossby wave in the tropical North Pacific Ocean. J. Phys. Oceanogr., 9 , 663674.

  • Meyers, G., 1979b: Annual variation in the slope of the 14° isotherm along the equator in the Pacific Ocean. J. Phys. Oceanogr., 9 , 885891.

    • Search Google Scholar
    • Export Citation
  • Pedlosky, J., 2006: A history of thermocline theory. Physical Oceanography: Developments since 1950, M. Jochum and R. Murtugudde, Eds., Springer, 139–152.

    • Search Google Scholar
    • Export Citation
  • Philip, S., , and G. J. van Oldenborgh, 2006: Shifts in ENSO coupling processes under global warming. Geophys. Res. Lett., 33 , L11704. doi:10.1029/2006GL026196.

    • Search Google Scholar
    • Export Citation
  • Sverdrup, H. U., , M. W. Johnson, , and R. H. Fleming, 1942: The Oceans: Their Physics, Chemistry, and General Biology. Prentice-Hall, 1087 pp.

    • Search Google Scholar
    • Export Citation
  • Swenson, M. S., , and D. V. Hansen, 1999: Tropical Pacific Ocean mixed layer heat budget: The Pacific cold tongue. J. Phys. Oceanogr., 29 , 6981.

    • Search Google Scholar
    • Export Citation
  • Timmermann, A., , M. Latif, , A. Bacher, , J. Oberhuber, , and E. Roeckner, 1999: Increased El Niño frequency in a climate model forced by future greenhouse warming. Nature, 398 , 694696.

    • Search Google Scholar
    • Export Citation
  • Vecchi, G. A., , and D. E. Harrison, 2000: Tropical Pacific sea surface temperature anomalies, El Niño, and equatorial westerly wind events. J. Climate, 13 , 18141830.

    • Search Google Scholar
    • Export Citation
  • Vecchi, G. A., , and B. J. Soden, 2007: Global warming and the weakening of the tropical circulation. J. Climate, 20 , 43164340.

  • Vecchi, G. A., , B. J. Soden, , A. T. Wittenberg, , I. M. Held, , A. Leetmaa, , and M. J. Harrison, 2006: Weakening of tropical Pacific atmospheric circulation due to anthropogenic forcing. Nature, 441 , 7376.

    • Search Google Scholar
    • Export Citation
  • Vialard, J., , and P. Delecluse, 1998: An OGCM study for the TOGA decade. Part II: Barrier-layer formation and variability. J. Phys. Oceanogr., 28 , 10891106.

    • Search Google Scholar
    • Export Citation
  • Wang, B., , R. Wu, , and R. Lukas, 2000: Annual adjustment of the thermocline in the tropical Pacific Ocean. J. Climate, 13 , 596616.

  • Yang, H., , and Q. Zhang, 2008: Anatomizing the ocean role in ENSO changes under global warming. J. Climate, 21 , 65396555.

  • Zebiak, S. E., , and M. A. Cane, 1987: A model El Niño–Southern Oscillation. Mon. Wea. Rev., 115 , 22622278.

  • View in gallery

    Climatological annual mean upper-ocean temperature (°C) averaged between 5°S and 5°N from Levitus data (1955–2003). The dashed line shows the location of the 20°C isotherm (Z20), and the solid line shows the location of the maximum vertical temperature gradient (Ztc).

  • View in gallery

    Time evolution (curves) of (a) Z20 and (b) Ztc averaged over the WEP (dashed; 5°S–5°N, 120°E–160°W) and the EEP (solid; 5°S–5°N, 130°–80°W) from the monthly JEDAC-XBT over the period 1955–2003 in meters. The climatological annual mean values of Z20 and Ztc are removed. The linear regressions of these curves are also plotted as straight lines. The 5-yr running mean is applied to the time series lines.

  • View in gallery

    The values of trends (triangles) and corresponding confidence intervals (short straight lines) at the 99% significant level for the EEP Z20 (dashed) and Ztc (solid) in meters per century. It is seen that all the trends are significant and the trends of Z20 and Ztc are opposite in the EEP for both the XBT data and coupled models (GFDL CM2.1 and NCAR CCSM3).

  • View in gallery

    (a) Linear trend of the upper-ocean temperature (°C century−1) averaged between 5°S and 5°N from the monthly JEDAC-XBT over the period 1955–2003. The thick black and gray lines represent the Z20 and Ztc, respectively. The solid (dashed) lines shows the locations of Z20 and Ztc averaged over 1960–75 (1985–2000). (b) The vertical temperature profile at 100°W. The solid line is for 1960–75 and the dashed line is for 1985–2000. The black arrows denote the changes in direction of Ztc and Z20. The small inset shows the vertical temperature profile in the upper 200 m, in which the dashed box marks the region of interest.

  • View in gallery

    Linear trends of (a) Z20 and (b) Ztc (m century−1) in the tropical Pacific from the monthly JEDAC-XBT over the period 1955–2003.

  • View in gallery

    Linear trends of the upper-ocean temperature (°C century−1) averaged between 5°S and 5°N from the 1pctto2x experiments of (a) GFDL CM2.1 and (b) NCAR CCSM3 over years 1–70. The thick solid and dashed lines in (a) and (b) represent the time averaged Z20 and Ztc over years 1–70, respectively.

  • View in gallery

    Linear trends of (a),(c) Z20 and (b),(d) Ztc (m century−1) in the tropical Pacific from the 1pctto2x experiments of (left) GFDL CM2.1 and (right) NCAR CCSM3 over years 1–70.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 3 3 3
PDF Downloads 2 2 2

Revisiting the Thermocline Depth in the Equatorial Pacific

View More View Less
  • 1 Department of Atmospheric Science, School of Physics, Peking University, Beijing, China
© Get Permissions
Full access

Abstract

The thermocline depth is defined as the depth of the maximum vertical temperature gradient. In the equatorial Pacific, the depth of 20°C isotherm is widely used to represent the thermocline depth. This work proposes that under the circumstance of a significant mean climate shift, it is better to use the original definition of the thermocline depth in studying the long-term changes in mean climate and tropical coupled climate variabilities. For instance, during the transient period of global warming, the tropical thermocline is usually enhanced because the surface layer warms more and faster than the lower layers. The depth of maximum vertical temperature gradient shoals, which is consistent with the enhanced thermocline. However, the 20°C isotherm depth deepens, which suggests a weakened thermocline. This discrepancy exists in both the observations and the future climate simulations of coupled models.

Corresponding author address: Haijun Yang, Department of Atmospheric Science, School of Physics, Peking University, 209 Chengfu Road, Beijing 100871, China. Email: hjyang@pku.edu.cn

Abstract

The thermocline depth is defined as the depth of the maximum vertical temperature gradient. In the equatorial Pacific, the depth of 20°C isotherm is widely used to represent the thermocline depth. This work proposes that under the circumstance of a significant mean climate shift, it is better to use the original definition of the thermocline depth in studying the long-term changes in mean climate and tropical coupled climate variabilities. For instance, during the transient period of global warming, the tropical thermocline is usually enhanced because the surface layer warms more and faster than the lower layers. The depth of maximum vertical temperature gradient shoals, which is consistent with the enhanced thermocline. However, the 20°C isotherm depth deepens, which suggests a weakened thermocline. This discrepancy exists in both the observations and the future climate simulations of coupled models.

Corresponding author address: Haijun Yang, Department of Atmospheric Science, School of Physics, Peking University, 209 Chengfu Road, Beijing 100871, China. Email: hjyang@pku.edu.cn

1. Introduction

The thermocline is a layer within a body of water where the temperature changes rapidly with depth. The word “thermocline” first appears in the limnology literature of the late nineteenth century (Pedlosky 2006) and is then more frequently referred as a discontinuity layer or transition layer (Sverdrup et al. 1942). It was very difficult to detect the deep ocean thermocline until the bathythermograph (BT) was invented in 1940s and the expendable bathythermograph (XBT) was developed during the 1970s. Based on the definition, theoretically, the depth of thermocline (DTC) should be taken as the depth of the maximum vertical temperature gradient (Ztc). However, in practice, the vertical resolution of observations is usually insufficient to resolve the maximum temperature gradient. Thus, the DTC is often determined by the depth of a “representative” isotherm [e.g., the depth of the 14° or 20°C isotherm (Z14 and Z20, respectively)] within the thermocline layer (Meyers 1979a,b; Kessler 1990).

For the equatorial Pacific, Z20 is broadly used as the DTC for studies (Kessler 1990; Ji et al. 1995; Vialard and Delecluse 1998; Swenson and Hansen 1999; Chavez et al. 1999; Durand and Delcrolx 2000; Vecchi and Harrison 2000; Meinen and McPhaden 2000; Fedorov and Philander 2001; Chepurin et al. 2005) because the 20°C isotherm is located near the center of the main thermocline. In earlier studies, the Z14 was chosen to represent the permanent DTC because it indicates strength of the North and South Equatorial Currents better than shallower isotherms (Meyers 1979a), although it is below the level of the Ztc. Both the Z20 and Z14 have some problems in representing the DTC (Wang et al. 2000). The Z20 is suitable for the warm pool, but it may fail for the cold tongue and coastal upwelling regions because the 20°C isotherm may outcrop to the surface during the cold season. On the other hand, the Z14 is too deep to represent the warm pool thermocline temperature. Therefore, Wang et al. (2000) proposed to use a location-dependent representative temperature Tc to characterize the thermocline temperature. The DTC is defined by averaging 12°C (representing the temperature at the base of the thermocline layer) and the local long-term mean sea surface temperature (SST; representing the temperature at the top of the thermocline). Such defined representative temperature Tc varies from 20.5°C in the warmest places to about 16°C near the Peru coast.

The essence of wind-driven circulation is the thermocline dynamics. For the equatorial Pacific, the DTC variations play a key role in both the mean SST and the coupled climate variabilities, such as El Niño–Southern Oscillation (ENSO). The basin-wide adjustment of the equatorial thermocline, including the vertical displacement and zonal tilting, is closely related to the phase, amplitude, and time scale of the tropical variabilities (Zebiak and Cane 1987; Chang 1994; Wang et al. 2000). Because of different physical processes involved, the thermocline adjustment may exhibit distinctive features and impact differently on the SST and ENSO (Chang 1994). An optimal proxy of the DTC is therefore critical to properly understand the thermocline dynamics.

Usually, in the equatorial Pacific, the location of the maximum vertical temperature gradient Ztc is best represented by Z20, as suggested in most studies. The Ztc and Z20 are well overlapped with each other except in the warm pool region where the Ztc tends to be located at the upper portion of the thermocline layer (Wang et al. 2000). One of advantages of using Z20 is that it implies the studies of movements along an isopycnal level. The tropical Pacific may be well approximated as a two-layer system divided by the 20°C isotherm (Kessler 1990).

It is worth discussing if the Z20 is still the best proxy of the Ztc under the circumstance of a significant mean climate shift. For instance, in the global warming, the 20°C isotherm will be intuitionally pushed downward because of the general warming in the upper layer of oceans. The Ztc, instead, may move upward because the upper portion of the thermocline layer warms more and faster than the lower ocean. The opposite changes in the Z20 and Ztc may bring discrepancy in understanding the impact of thermocline change on the surface coupled climate variabilities such as ENSO. If during the transient period of global warming a stronger ENSO is observed, then this could result from either an intensified or weakened equatorial thermocline (referring to Ztc and Z20, respectively). This discrepancy arises unnecessarily from the different treatment of the DTC. The relationship of a stronger ENSO to an intensified thermocline is widely accepted (e.g., An et al. 2008; Meehl et al. 2001; Collins 2000; Timmermann et al. 1999).

In this note, we propose that it is necessary to use the original definition of the thermocline depth to avoid this discrepancy. The vertical resolution issue is not a practical obstacle to resolve the Ztc anymore, considering the great improvement in both the observations and coupled models in the past decades. Some recent works have already adopted Ztc in the global warming studies (e.g., Vecchi et al. 2006). In the following context, we will show that using Ztc is better for the consistencies between the thermocline dynamics and the coupled variabilities under the background of mean climate shift.

2. Data

Observational datasets examined here include 1) the monthly upper-ocean temperature dataset from the Scripps Institution of Oceanography Joint Environmental Data Analysis Center (JEDAC-XBT) and 2) the yearly ocean temperature dataset from the National Oceanographic Data Center (NODC; Levitus et al. 2005). The JEDAC-XBT data have a horizontal resolution of 5° × 2° in longitude and latitude and 11 vertical levels (0, 20, 40, 60, 80, 120, 160, 200, 240, 300, and 400; available online at http://jedac.ucsd.edu/DATA_IMAGES/index.html). The Levitus data have a horizontal resolution of 1° × 1° in longitude and latitude and 16 vertical levels (0, 10, 20, 30, 50, 75, 100, 125, 150, 200, 250, 300, 400, 500, 600, and 700). Both datasets span the period 1955–2003.

The model datasets examined here are from phase 3 of the Coupled Model Intercomparison Project (CMIP 3) evaluated for the Fourth Assessment Report (AR4) of the Intergovernmental Panel on Climate Change (IPCC): the Geophysical Fluid Dynamics Laboratory Climate Model version 2.1 (GFDL CM2.1) and the National Center for Atmospheric Research Community Climate System Model, version 3 (NCAR CCSM3). The “1% increase per year to doubling” (1pctto2x) experiments from these two models are examined. In this paper, only the first 70 yr of outputs of 1pctto2x experiments are analyzed to identify the linear change in the DTC in a future warming climate; that is, we only focus on the transient period of the global warming during which the atmospheric CO2 is increased by 1% yearly.

In this note, only the long-term linear trend of the DTC is of interest. The least squares method is used to determine the linear trends. For the time evolution of the DTC, a 5-yr running mean is applied to both the observational and model data before the linear trend calculation. Furthermore, a Student’s t test is used to test the significance of the linear trends (see appendix for details).

3. Mean thermocline depth in the equatorial Pacific

The Ztc and Z20 in the equatorial Pacific generally coincide with each other, except in the warm pool region where the Ztc is shallower than the Z20 by about 20–40 m (Fig. 1). Here, the climatological annual mean Levitus data are analyzed because of their relatively higher vertical resolution than XBT data. In the eastern Pacific, where the thermocline dynamics is vital to the SST and coupled variabilities, the Z20 can fully represent the Ztc. In the western Pacific, where the thermocline is deep, the discrepancy between the Z20 and Ztc does not matter because the thermocline hardly affects the surface locally. Therefore, representing the Ztc with Z20 will not bring dynamical inconsistency for a climate with a steady time mean state, as shown in many previous studies.

4. Different changes in Z20 and Ztc

The Ztc and Z20 in the equatorial Pacific can change in different directions if the mean climate has a significant shift. Figure 2 shows the linear trends of the equatorial thermocline depth during 1955–2003. For the western equatorial Pacific (WEP; 5°S–5°N, 120°E–160°W), the Z20 and Ztc exhibit the similar shoaling trend during the past half century (black lines in Figs. 2a,b). However, for the eastern equatorial Pacific (EEP; 5°S–5°N, 130°–80°W), the Z20 shows a deepening trend whereas the Ztc shows a shoaling trend (black lines in Figs. 2a,b). Considering negative correlation between the thermocline strength and its depth in the EEP, the deepening of Z20 means a weakening thermocline, whereas the shoaling of Ztc means a strengthening thermocline. An obvious inconsistency in understanding the thermocline dynamics occurs here. Figure 2 is plotted using monthly JEDAC-XBT. The same situation also occurs in the yearly Levitus data (figure not shown).

These trends in the equatorial DTC are significant at the 99% level (Table 1; Fig. 3). The Student’s t test is used to test the significance. The trends in the WEP are remarkable, as shown in Fig. 2. For simplicity and clarity only, the values in the EEP are listed in Table 1 and plotted in Fig. 3. It is seen clearly that the lower and upper limits of the confidence intervals have the same sign, suggesting the significance of the linear trends. Moreover, the trends of Z20 and Ztc are clearly opposite of each other.

There is evidence showing that the global ocean has been warming for a long time (Levitus et al. 2005; Vecchi et al. 2006). The warming rate of tropical SST in the past 50 yr is generally over 1°C century−1; it reaches 1.5°C century−1 in the EEP (Fig. 4a). This big surface warming, however, has not penetrated sufficiently deep into the thermocline, where a cooling trend occurs instead. The cooling rate in the western thermocline reaches as much as −2°C century−1. Although the cooling rate in the eastern thermocline is not as significant as that in the west, the overall vertical temperature gradient is enhanced during the past 50 yr. This clearly shows a strengthening trend in the equatorial Pacific thermocline.

The different trends in the Z20 and Ztc can be also seen in Fig. 4. Because of the overall surface warming, the 20°C isotherm level has to be pushed downward. Because of the much faster warming in the surface layer than in the subsurface, the location of the maximum vertical temperature gradient has to move upward so that the Ztc has a shoaling trend in the eastern Pacific (Fig. 4b). The real situation that happened in the EEP is that the thermocline is strengthening and shoaling as disclosed by Ztc instead of weakening and deepening as suggested by Z20.

Figure 4a also shows an overall reduction in the Ztc and Z20 in the equatorial Pacific. The shoaling trend of the DTC is more notable in the western equatorial Pacific than in the east, suggesting a flattening in the east–west tilt of the equatorial Pacific thermocline (Fig. 4a). The horizontal pattern of Z20 trend clearly shows a zonally flattening thermocline with a shoaling in the west and a deepening in the east (Fig. 5a), whereas the horizontal pattern of Ztc shows a zonal shoaling and flattening in the whole equatorial Pacific (Fig. 5b). The flattening of the equatorial thermocline can be attributed to the weakened trade wind and, therefore, the weakening of tropical Walker circulation (Vecchi et al. 2006; Vecchi and Soden 2007). It is worth noting that the magnitude of the Ztc trend is much larger than that of the Z20 trend: about 5 times more in the west and about 10 times more in the east. This implies that the significant change in the thermocline depth could be incorrectly ignored if the focus is only on the Z20.

The shoaling of Ztc (i.e., the strengthened thermocline in the EEP) is consistent with the enhancing ENSO amplitude in the past 50 yr (Yang and Zhang 2008). The shoaling of Ztc increases the SST sensitivity to the thermocline in the Niño-3 region (5°S–5°N, 150°–90°W; Philip and van Oldenborgh 2006). A shallower Ztc means the thermocline water would be easier to affect the SST through the anomalous upwelling. A stronger thermocline means a bigger impact on the SST. The shallower and stronger thermocline can result in an enhanced ENSO variability because of the positive feedback among the wind variability, anomalous upwelling, and SST variability in the central-east Pacific (Yang and Zhang 2008). If the Z20 is used to represent the DTC in this situation, a false dynamic connection between the DTC and ENSO changes would be concluded.

Here, we further check the DTC change in 1pctto2x experiments of two coupled models: GFDL CM2.1 and NCAR CCSM3. Because coupled models are the most important tools in climate studies, the proper treatment of the equatorial thermocline in coupled models is crucial to correctly understand the mechanisms of the tropical climate variabilities. Figure 6 shows that, during the transient stage of global warming, the equatorial thermocline is enhanced in both of the models and the mean Ztc is usually shallower than the mean Z20. It is seen that, in the EEP, the Z20 has a deepening trend (Figs. 7a,c), whereas the Ztc has a shoaling trend (Figs. 7b,d). The Ztc differs from the Z20 in both the sign and size. The deepening of Z20 should suggest a weakening thermocline, particularly in the eastern Pacific; however, the trend in the upper-ocean temperature (Fig. 6) suggests that the thermocline is enhanced during the transient stage of global warming, which is consistent with the shoaling of Ztc, as shown in Figs. 7b,d. Moreover, conformed to the enhancing thermocline, ENSO amplitude is also increased in these two coupled models (figure not shown). This has been examined carefully in Yang and Zhang (2008). The change in ENSO amplitude is well complied with the change in Ztc in dynamics. Therefore, the Ztc is better to reflect the real changes in the equatorial thermocline and should be paid more attention in the study of the tropical coupled climate variabilities, particularly under the circumstance of mean climate shift.

The significance of these trends in the two models is also checked. All the trends are significant at the 99% confidence level (Table 1). It is noticed that the shoaling trend of the Ztc in the EEP is much smaller than that in the observation (Fig. 3). This is consistent with the relatively weaker changes in the equatorial thermocline strength suggested in Fig. 6 than those in the XBT (Fig. 4a). In the coupled models, the warming trends in the EEP penetrate much deeper than those in the XBT, whereas the cooling trends mostly occur to the west of 130°W (Figs. 6a,b). This is responsible for the smaller Ztc changes in the coupled model (Fig. 3). As for what causes the vertical temperature structures in the coupled models as shown in Fig. 6, we have to turn to a thorough examination of the coupled models simulations, which is outside of the discussions of this work.

5. Summary

The analyses above remind us that we should be cautious when investigating the long-term change in the equatorial thermocline and its impact on the surface dynamics. The different treatment of the DTC may result in controversies in understanding the mechanisms of the coupled climate variabilities in the tropics. Although using Z20 to represent the DTC has many advantages (e.g., it is the most convenient and simplest method to determine the DTC in the tropical Pacific and it simplifies the three-dimensional variability into a two-dimensional field, which can be mapped and studied conveniently; Kessler 1990), Z20 may fail to reflect the realistic long-term trend of the thermocline strength. Identifying the change in the depth of the maximum vertical temperature gradient is of great necessity in understanding the long-term climate change properly.

To avoid some unnecessary discrepancies, it is particularly important to figure out if Z20 or Ztc is better to represent the DTC in the tropical Pacific when studying the past or future climate change by coupled models, because the changes in Z20 and Ztc can be out of phase along the equator in the situation of global warming or cooling. Some of recent modeling studies have already used Ztc (e.g., Vecchi et al. 2006). Considering the great improvement of the vertical resolution of coupled models, reasonably identifying Ztc no longer has any practical obstacles.

Acknowledgments

This work is jointly supported by the NSF of China (No. 40576004) and the National Basic Research Program of China (2007CB411801). The comments from two anonymous reviewers are greatly appreciated. We thank the international groups for providing observational data and coupled model outputs.

REFERENCES

  • An, S-I., , J-S. Kug, , Y-G. Ham, , and I-S. Kang, 2008: Successive modulation of ENSO to the future greenhouse warming. J. Climate, 21 , 321.

    • Search Google Scholar
    • Export Citation
  • Chang, P., 1994: A study of the seasonal cycle of sea surface temperature in the tropical Pacific Ocean using reduced gravity models. J. Geophys. Res., 99 , 77257741.

    • Search Google Scholar
    • Export Citation
  • Chavez, F. P., , P. G. Strutton, , G. E. Friederich, , R. A. Feely, , G. C. Feldman, , D. G. Foley, , and M. J. McPhaden, 1999: Biological and chemical response of the equatorial Pacific ocean to the 1997-98 El Niño. Science, 286 , 21262131.

    • Search Google Scholar
    • Export Citation
  • Chepurin, G. A., , J. A. Carton, , and D. Dee, 2005: Forecast model bias correction in ocean data assimilation. Mon. Wea. Rev., 133 , 13281342.

    • Search Google Scholar
    • Export Citation
  • Collins, M., 2000: The El Niño–Southern Oscillation in the second Hadley Centre coupled model and its response to greenhouse warming. J. Climate, 13 , 12991312.

    • Search Google Scholar
    • Export Citation
  • Durand, F., , and T. Delcroix, 2000: On the variability of the tropical Pacific thermal structure during the 1979–96 period, as deduced from XBT sections. J. Phys. Oceanogr., 30 , 32613269.

    • Search Google Scholar
    • Export Citation
  • Fedorov, A. V., , and S. G. Philander, 2001: A stability analysis of tropical ocean–atmosphere interaction: Bridging measurements and theory for El Niño. J. Climate, 14 , 30863101.

    • Search Google Scholar
    • Export Citation
  • Ji, M., , A. Leetmaa, , and J. Derber, 1995: An ocean analysis system for seasonal to interannual climate studies. Mon. Wea. Rev., 123 , 460481.

    • Search Google Scholar
    • Export Citation
  • Kessler, W. S., 1990: Observations of long Rossby waves in the northern tropical Pacific. J. Geophys. Res., 95 , 51835217.

  • Levitus, S., , J. Antonov, , and T. Boyer, 2005: Warming of the world ocean, 1955–2003. Geophys. Res. Lett., 32 , L02604. doi:10.1029/2004GL021592.

    • Search Google Scholar
    • Export Citation
  • Meehl, G. A., , P. R. Gent, , J. M. Arblaster, , B. L. Otto-Bliesner, , E. C. Brady, , and A. Craig, 2001: Factors that affect the amplitude of El Niño in global coupled climate models. Climate Dyn., 17 , 515526.

    • 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
  • Meyers, G., 1979a: On the annual Rossby wave in the tropical North Pacific Ocean. J. Phys. Oceanogr., 9 , 663674.

  • Meyers, G., 1979b: Annual variation in the slope of the 14° isotherm along the equator in the Pacific Ocean. J. Phys. Oceanogr., 9 , 885891.

    • Search Google Scholar
    • Export Citation
  • Pedlosky, J., 2006: A history of thermocline theory. Physical Oceanography: Developments since 1950, M. Jochum and R. Murtugudde, Eds., Springer, 139–152.

    • Search Google Scholar
    • Export Citation
  • Philip, S., , and G. J. van Oldenborgh, 2006: Shifts in ENSO coupling processes under global warming. Geophys. Res. Lett., 33 , L11704. doi:10.1029/2006GL026196.

    • Search Google Scholar
    • Export Citation
  • Sverdrup, H. U., , M. W. Johnson, , and R. H. Fleming, 1942: The Oceans: Their Physics, Chemistry, and General Biology. Prentice-Hall, 1087 pp.

    • Search Google Scholar
    • Export Citation
  • Swenson, M. S., , and D. V. Hansen, 1999: Tropical Pacific Ocean mixed layer heat budget: The Pacific cold tongue. J. Phys. Oceanogr., 29 , 6981.

    • Search Google Scholar
    • Export Citation
  • Timmermann, A., , M. Latif, , A. Bacher, , J. Oberhuber, , and E. Roeckner, 1999: Increased El Niño frequency in a climate model forced by future greenhouse warming. Nature, 398 , 694696.

    • Search Google Scholar
    • Export Citation
  • Vecchi, G. A., , and D. E. Harrison, 2000: Tropical Pacific sea surface temperature anomalies, El Niño, and equatorial westerly wind events. J. Climate, 13 , 18141830.

    • Search Google Scholar
    • Export Citation
  • Vecchi, G. A., , and B. J. Soden, 2007: Global warming and the weakening of the tropical circulation. J. Climate, 20 , 43164340.

  • Vecchi, G. A., , B. J. Soden, , A. T. Wittenberg, , I. M. Held, , A. Leetmaa, , and M. J. Harrison, 2006: Weakening of tropical Pacific atmospheric circulation due to anthropogenic forcing. Nature, 441 , 7376.

    • Search Google Scholar
    • Export Citation
  • Vialard, J., , and P. Delecluse, 1998: An OGCM study for the TOGA decade. Part II: Barrier-layer formation and variability. J. Phys. Oceanogr., 28 , 10891106.

    • Search Google Scholar
    • Export Citation
  • Wang, B., , R. Wu, , and R. Lukas, 2000: Annual adjustment of the thermocline in the tropical Pacific Ocean. J. Climate, 13 , 596616.

  • Yang, H., , and Q. Zhang, 2008: Anatomizing the ocean role in ENSO changes under global warming. J. Climate, 21 , 65396555.

  • Zebiak, S. E., , and M. A. Cane, 1987: A model El Niño–Southern Oscillation. Mon. Wea. Rev., 115 , 22622278.

APPENDIX

Linear Trend and Student’s t Test

A linear regression line for a random sample y(x) can be represented as y = bx + a. Here, x is the independent variable and y is the dependent variable. For convenience, we introduce Sxx and Syy as the variances of variables x and y, respectively, and Sxy as the covariance of x and y as follows:
i1520-0442-22-13-3856-ea1

The slope (i.e., the linear trend) can then be calculated by the formula = Sxy/Sxx, according to the least squares method.

The significance of the linear trend can be tested as follows: First, the sum of square errors is calculated as Qe = Syyxy. If the variance of the error term is denoted by σ2, then an unbiased estimate of σ2 can be given by σ̂2 = Qe/(n − 2).

Second, the test statistic is defined as the following:
i1520-0442-22-13-3856-ea2
where b0 is given. For the two-tailed t test H0:b0 = 0 and H1:b0 ≠ 0. Under the significant level α, the rejection region for H0 is the following:
i1520-0442-22-13-3856-ea3
Therefore, the (1 −α) percent confidence interval for the slope (linear trend) is the following:
i1520-0442-22-13-3856-ea4
Fig. 1.
Fig. 1.

Climatological annual mean upper-ocean temperature (°C) averaged between 5°S and 5°N from Levitus data (1955–2003). The dashed line shows the location of the 20°C isotherm (Z20), and the solid line shows the location of the maximum vertical temperature gradient (Ztc).

Citation: Journal of Climate 22, 13; 10.1175/2009JCLI2836.1

Fig. 2.
Fig. 2.

Time evolution (curves) of (a) Z20 and (b) Ztc averaged over the WEP (dashed; 5°S–5°N, 120°E–160°W) and the EEP (solid; 5°S–5°N, 130°–80°W) from the monthly JEDAC-XBT over the period 1955–2003 in meters. The climatological annual mean values of Z20 and Ztc are removed. The linear regressions of these curves are also plotted as straight lines. The 5-yr running mean is applied to the time series lines.

Citation: Journal of Climate 22, 13; 10.1175/2009JCLI2836.1

Fig. 3.
Fig. 3.

The values of trends (triangles) and corresponding confidence intervals (short straight lines) at the 99% significant level for the EEP Z20 (dashed) and Ztc (solid) in meters per century. It is seen that all the trends are significant and the trends of Z20 and Ztc are opposite in the EEP for both the XBT data and coupled models (GFDL CM2.1 and NCAR CCSM3).

Citation: Journal of Climate 22, 13; 10.1175/2009JCLI2836.1

Fig. 4.
Fig. 4.

(a) Linear trend of the upper-ocean temperature (°C century−1) averaged between 5°S and 5°N from the monthly JEDAC-XBT over the period 1955–2003. The thick black and gray lines represent the Z20 and Ztc, respectively. The solid (dashed) lines shows the locations of Z20 and Ztc averaged over 1960–75 (1985–2000). (b) The vertical temperature profile at 100°W. The solid line is for 1960–75 and the dashed line is for 1985–2000. The black arrows denote the changes in direction of Ztc and Z20. The small inset shows the vertical temperature profile in the upper 200 m, in which the dashed box marks the region of interest.

Citation: Journal of Climate 22, 13; 10.1175/2009JCLI2836.1

Fig. 5.
Fig. 5.

Linear trends of (a) Z20 and (b) Ztc (m century−1) in the tropical Pacific from the monthly JEDAC-XBT over the period 1955–2003.

Citation: Journal of Climate 22, 13; 10.1175/2009JCLI2836.1

Fig. 6.
Fig. 6.

Linear trends of the upper-ocean temperature (°C century−1) averaged between 5°S and 5°N from the 1pctto2x experiments of (a) GFDL CM2.1 and (b) NCAR CCSM3 over years 1–70. The thick solid and dashed lines in (a) and (b) represent the time averaged Z20 and Ztc over years 1–70, respectively.

Citation: Journal of Climate 22, 13; 10.1175/2009JCLI2836.1

Fig. 7.
Fig. 7.

Linear trends of (a),(c) Z20 and (b),(d) Ztc (m century−1) in the tropical Pacific from the 1pctto2x experiments of (left) GFDL CM2.1 and (right) NCAR CCSM3 over years 1–70.

Citation: Journal of Climate 22, 13; 10.1175/2009JCLI2836.1

Table 1.

The values of trends and corresponding confidence intervals at the 99% significant level for the EEP Z20 and Ztc (in m century−1). The confidence interval is obtained by Student’s t test. All the trends are significant at the 99% significance level. The trends are calculated from the monthly XBT data over the period of 1955–2003 and GFDL CM2.1 and NCAR CCSM3 data over years 1–70. These values are also visualized in Fig. 3.

Table 1.

* Department of Atmospheric Science Contribution Number 008.

Save