1. Introduction
Sea surface temperature (SST) in the equatorial eastern Pacific Ocean shows marked interannual variation, which is referred to as El Niño and La Niña. Observational studies have found that the interannual variation of SST is closely related to thermocline depth (Wyrtki 1985; Meinen and McPhaden 2000; Hasegawa and Hanawa 2003). Their relationship can be observed by the Niño-3.4 SST anomaly and 20°C isothermal depth anomaly averaged over the equatorial Pacific Ocean (Fig. 1). Deepening of the thermocline in the equatorial band leads to high SST and the appearance of El Niño. In turn, the thermocline shoals during El Niño, which is followed by lower SST and the reemergence of La Niña conditions. La Niña causes deepening of the thermocline and continuation of the El Niño–La Niña cycle. The time lag between the SST and thermocline depth variations is about half a year (Meinen and McPhaden 2000). Traditional theories of El Niño–Southern Oscillation (ENSO) are based on this observational relationship (Battisti 1988; Schopf and Suarez 1988; Suarez and Schopf 1988; Jin 1997a, b; Weisberg and Wang 1997).
However, SST variability associated with ENSO cannot be explained only by thermocline depth variations. It is known that some La Niña events linger after the thermocline has deepened, and El Niño events do not immediately appear (Kessler 2002). For example, thermocline became deeper than normal in 2000, but the Niño-3.4 SST anomaly kept negative until 2002 and then turned to be positive (Fig. 1, marked with arrow numbered 4). Such delay of appearance of El Niño is also found in 1986 El Niño and 1991 El Niño (arrows 1 and 2 in Fig. 1), although the delay is not clear in 1997 El Niño (arrow 3). In contrast, the transition to La Niña is rapid: the Niño-3.4 SST anomaly became negative following the shoaling of the thermocline, as is seen in 1983, 1988, and 1998. Thus, transition to El Niño tends to be slow compared to that to La Niña, and the ENSO cycle seems to “pause” before the appearance of El Niño. This characteristic cannot be explained by the traditional ENSO theories.
This study examines SST variation during the slow transition from La Niña to El Niño. Hereafter, the slow transition is referred to as a “pausing” of the ENSO cycle. Investigation of such pausing of the ENSO cycle should contribute to our understanding of the onset timing of El Niño.
To examine SST variation, the heat balance in the surface mixed layer must be analyzed. According to Philander and Hurlin (1988), Brady (1994), Jin (2001), and Meinen and McPhaden (2001), equatorward heat transport is caused by geostrophic currents at the subsurface depth during the transition period from La Niña to El Niño. Resultant accumulation of heat accompanies the deepened thermocline in the equatorial band, which Jin (1997a, b) called the recharge process. The accumulated heat at the subsurface is vertically advected by equatorial upwelling and anomalously warms the surface mixed layer (Battisti 1988; Vialard et al. 2001; Zelle et al. 2004). Horizontal advection due to zonal current anomalies also contributes to anomalous warming in the mixed layer (Harrison and Schopf 1984; Picaut et al. 1996, 1997; Jin and An 1999). These warming anomalies cause the transition from La Niña to El Niño. On the other hand, surface heat flux works as a damping effect on the SST anomaly. During La Niña, the negative SST anomaly is anomalously warmed by the surface heat flux (Liu et al. 1994; Waliser et al. 1994; Wang and McPhaden 2000; Mestas-Nuñez et al. 2006). This damping effect contributes to the decay of La Niña.
The observed slow transition from La Niña to El Niño suggests that the above warming anomalies are reduced by a number of other processes. One candidate is eddy heat fluxes, caused by instability waves, which have not been explicitly considered in traditional ENSO theories. In the central and eastern tropical Pacific Ocean, energetic waves are prominent with 20- or 30-day periods, 1000-km wavelength and 0.5 m s−1 westward propagation speed (Legeckis 1977; Halpern et al. 1988; Flament et al. 1996; Chelton et al. 2000). Various studies have suggested that the main energy source of these waves is the instability of meridional shear between the South Equatorial Current (SEC) and North Equatorial Counter Current (NECC; Philander 1978; Cox 1980; Flament et al. 1996; Baturin and Niiler 1997). The waves are thus called tropical instability waves (TIWs). These waves are a dominant contributor to eddy heat flux in the equatorial region (Bryden and Brady 1989), which is equivalent to surface heat flux (Hansen and Paul 1984). This eddy heat flux causes equatorward heat transport to warm the mixed layer at the equator. It is known that activity of the TIWs temporally changes in association with variation in the strength of the SEC and NECC (Baturin and Niiler 1997; Johnson and Proehl 2004). Considering these studies, we focus on the possibility that interannual variations of the TIWs and their eddy heat fluxes affect the surface heat balance at the interannual time scale.
In this study, we analyzed the heat balance in the surface mixed layer using observational data in order to examine mechanisms of the pausing from La Niña to El Niño. Observational data collected over a long time period are necessary to evaluate interannual variation; high temporal resolution is also needed to resolve the instability waves. Both meteorological and oceanic observations are required for heat balance calculation. Therefore we used data from the Tropical Atmosphere Ocean/Triangle Trans-Ocean Buoy Network (TAO/TRITON; McPhaden et al. 1998), which satisfy the above requirements. Because of the limitation of available periods, our focus is on transition from the 1998 La Niña to 2002 El Niño, during which period there was a clear pausing of the ENSO cycle (Fig. 1).
Section 2 describes the method for calculating the heat balance and the datasets used. Results of the heat balance calculation are shown in section 3. Section 4 examines processes that determine the heat balance, and section 5 discusses remaining problems, followed by a summary of results in section 6.
2. Data and processing
a. Datasets
TAO/TRITON data are used to calculate the heat balance. Wind, air temperature, humidity over the sea surface, oceanic temperature, and current data are obtained for the analysis period of 1991 to the present. Both meteorological and oceanic observations are available for this period. In addition, weekly SST from National Oceanic and Atmospheric Administration (NOAA) Optimum Interpolation (OI) SST (Reynolds et al. 2002) data are used for the heat balance calculation. Although TAO observations provide daily data, we calculate the heat balance as weekly averaged values for consistency with the NOAA OI SST. The NOAA OI SST data are blended satellite–in situ SST data. Table 1 lists the horizontal–temporal resolution and available period. In sections 3 and 4, we show the heat balance at the center of the Niño-3.4 region (0°, 140°W). The results at the western edge and eastern side of the Niño-3.4 region (0°, 170°W and 0°, 110°W) are discussed in section 5a.
To examine large-scale variability, we analyze grid datasets of surface currents, 20°C isothermal depths, surface winds, and SST. Surface currents at a 15-m depth are provided by the Ocean Surface Current Analysis-Real time (OSCAR) project. OSCAR data are derived from Ocean Topography Experiment (TOPEX)/Poseidon sea surface height, Special Sensor Microwave Imager (SSM/I) winds and NOAA OI SST, which give surface currents consistent with TAO/TRITON observations (Bonjean and Lagerloef 2002). Isothermal depths of 20°C are obtained from temperature data from the Bureau of Meteorology Research Center (BMRC) ocean analysis. BMRC temperature data are compiled from TAO/TRITON and ship observations using optimal interpolation (Smith 1995). The 20°C isothermal depth is calculated by linear interpolation and used as a proxy for thermocline depth. Vertical resolutions of BMRC data are 25 m at depths of 50–150 m. We also analyze surface wind data from the National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis dataset (Kalnay et al. 1996). NOAA OI SST data are used for the analysis of SST. Table 1 summarizes the resolutions and available periods of these data.
b. Heat balance formalism
The first term of the right-hand side (rhs) in Eq. (1) represents the effect of surface heat flux, which is calculated as net surface heat flux over the sea surface (Qsurface) divided by the heat capacity of the mixed layer (ρcph). Here, ρ and cp are seawater density and specific heat, which are set as constant values of 1022.4 kg m−1 and 3940 J (kg °C)−1, respectively.
The net surface heat flux (Qsurface) consists of radiation and turbulent heat fluxes. Shortwave radiation is obtained from direct hourly measurements by mooring buoys, with missing data filled in using NOAA outgoing longwave radiation data (Liebmann and Smith 1996). Albedo and penetration of shortwave radiation across the bottom of the mixed layer are taken into account as they are by Wang and McPhaden (1999). Longwave radiation and turbulent heat fluxes are estimated using the bulk formulae in Clark et al. (1974) and Fairall et al. (1996).
The second term of the rhs of Eq. (1) represents horizontal heat advection. Horizontal current velocities (uH) are obtained from current meters at 10-m depths, with missing data filled by linear extrapolation of Acoustic Doppler Current Profiler (ADCP) observations at 30–45-m depths. Horizontal temperature gradients (∇HT) are estimated from NOAA OI SST using the 1° centered difference.
Vertical heat advection is represented by the third term. The method for calculating this term follows McPhaden’s (2002) method. The mass flux across the bottom of the mixed layer (w*) is estimated as the sum of the climatological upwelling velocity (
The last term in the rhs (R) is a residual of the heat balance. The method described in this section is effectively a closed heat budget at the interannual time scale, which is shown in the appendix. The error estimation of the heat balance terms and the residual are also presented therein.
c. Time filtering
All of the heat balance terms are calculated from TAO observations and NOAA OI SST as weekly averaged values. The interannual anomaly is extracted from weekly time series as follows; weekly time series are averaged to monthly means; the monthly climatology is calculated for 1991 to 2004 and subtracted from the monthly means; the 5-month running mean and the 3-month triangle filter are applied. Hereafter the monthly interannual anomaly and the monthly climatology of arbitrary variable X are denoted by X′ and
In addition to them, high-frequency variability can affect the heat balance because of its eddy heat flux. The high-frequency component denoted by Xhigh is calculated as a residual, namely Xhigh = X −
As a result,
The first term of the rhs of Eq. (3) (−u′H · 
3. Heat balance during the pausing period
First, we define the pausing period between the 1998 La Niña and 2002 El Niño. Figure 2 repeats the SST and thermocline depth index in Fig. 1, but for 1997 to 2003. The red and blue solid lines indicate the El Niño and La Niña events, respectively. Following the definition of Trenberth (1997), El Niño and La Niña are defined by the Niño-3.4 SST anomaly exceeding ±0.4°C for 6 months or more.
The Niño-3.4 SST anomaly shows El Niño in 1997 and subsequent La Niña. The thermocline is deep before the peak of the 1997 El Niño, but shoals up until the middle of 1998. La Niña persists until 2001 with gradual deepening of the thermocline probably associated with the recharge process (Meinen and McPhaden 2001). In March 2000, the thermocline depth becomes deeper than normal. However, the Niño-3.4 SST anomaly remains negative for almost 2 yr afterward until the El Niño onset in 2002. The light shading in Fig. 2 shows the period with the positive thermocline depth anomaly and negative Niño-3.4 SST anomaly. The period marked by the light shading is relatively long compared to the transition from the 1997 El Niño to the 1998 La Niña (hatched period). We refer to the light-shaded period as the “pausing” period hereafter.
Figure 3 shows the mixed layer temperature anomaly and heat balance in the pausing period for the center of the Niño-3.4 region (0°, 140°W). The mixed layer temperature anomaly at this site rises from 1999 to 2003 consistent with the Niño-3.4 SST anomaly (Fig. 3a, solid line). The rise of the temperature anomaly is caused by a warming anomaly due to the surface heat flux term in the first half of the pausing period (Fig. 3b, red line). In the second half of the pausing period, the vertical heat advection term anomalously warms the mixed layer (blue line). On the other hand, the horizontal heat advection term shows a cooling anomaly during the pausing period (green line). These characteristics exceed the error and are significant.
The dotted line in Fig. 3a confirms the quantitative contribution of the cooling anomaly, which is integrated mixed layer temperature without the horizontal advection term. The mixed layer temperature anomaly is overestimated by about 5°C at the end of the pausing period. Neither the length of the pausing period nor the amplitude of the mixed layer temperature anomaly is reproduced without the horizontal advection term. These results show that the slow rise of the temperature anomaly and the long pausing period are attributable to the cooling anomaly caused by the horizontal advection.
Warming anomalies due to surface heat flux and vertical heat advection are a well-known feature during the transition from La Niña to El Niño (Battisti 1988; Wang and McPhaden 2000; Vialard et al. 2001; Zelle et al. 2004). Here, we briefly examine the physics of these terms. The vertical heat advection is closely related to thermocline depth variability (Fig. 4a). Positive (negative) 20°C isothermal depth anomalies correspond to warming (cooling) anomalies of the vertical heat advection. Zelle et al. (2004) showed that this relationship is caused by advection of the subsurface temperature anomaly by equatorial upwelling. The anomalies of the surface heat flux and mixed layer temperature show an out-of-phase relationship (Fig. 4b). This relationship indicates a damping effect of the surface heat flux on the temperature anomaly, which has been pointed out by Liu et al. (1994), Waliser et al. (1994), Wang and McPhaden (2000), and Mestas-Nuñez et al. (2006).
These warming anomalies cause the rise of the temperature anomaly during the La Niña to El Niño transition. The results in this section show that these warming anomalies are reduced by the cooling anomaly of horizontal advection during the pausing period. In the next section, physical processes in the cooling anomaly are examined.
4. Cooling anomalies in the pausing period
To examine physical processes the horizontal advection anomaly is decomposed into three components using Eq. (3). These components are shown in Fig. 5 by solid lines. The dashed lines in all panels of Fig. 5 show the total horizontal advection anomaly [−(uH · ∇HT)′], which indicates the cooling anomaly in the pausing period. This cooling anomaly arises from advection of climatological temperature by current anomalies (−u′H ·
a. Eddy heat flux anomalies
Bryden and Brady (1989) pointed out that high-frequency variability associated with TIWs contributes greatly to eddy heat flux. The TIWs show temporal variation in their activity associated with the strength of equatorial currents (Baturin and Niiler 1997; Johnson and Proehl 2004), which possibly causes the interannual anomaly of eddy heat flux. Considering these studies, we compare the eddy heat flux anomaly (Q′eddy) with index to the TIWs activity, and meridional shear between the SEC and NECC (Fig. 6).
The interannual variation of the TIWs activity is estimated from the high-frequency component of the weekly meridional current (υhigh) and SST gradient (∂Thigh/∂y) at 0°, 140°W. The meridional current is usually used to estimate TIW activity (e.g., Halpern et al. 1988). In addition, we use the meridional gradient of SST, considering heat advection. To concentrate on the TIWs activity, the power in 14–50-day-period bands is extracted from the high-frequency component using the wavelet analysis (Torrence and Compo 1998). The same period band was used in McPhaden (1996) to study the TIWs. We had confirmed that the present results do not change essentially if we use the 14–30-day-period bands following Baturin and Niiler (1997) and Johnson and Proehl (2004). The interannual change of the TIWs activity is examined by the monthly interannual anomaly of the wavelet power in 14–50-day-period band. The interannual anomaly is extracted by the time filtering described in section 2c, namely the combination of the monthly average, subtraction of monthly climatology, 5-month running mean and 3-month triangle filter. The resultant time series is shown in Fig. 6b.
The meridional shear shown by solid lines in Fig. 6c are calculated as the difference of zonal current anomalies between 2.5° and 7.5°N along 140°W. Note that the latitudes 2.5° and 7.5°N are the locations of the cores of the SEC and NECC, respectively (Johnson et al. 2002). Positive (negative) values indicate strong (weak) shear. The dashed line in Fig. 6c shows the zonal current anomalies at the equator, which indicates the variation of the SEC.
The eddy heat flux anomalies indicate the cooling anomalies in mid-2000 and mid-2001 (Fig. 6a, open arrows). Simultaneously, the 14–50-day variability and meridional shear of the zonal currents weaken (Figs. 6b,c). A reversed situation is found at the end of 2000, that is, a warming anomaly of the eddy heat flux, strong variability in 14–50-day period, and strong shear (marked with filled arrows). This correspondence suggests that the shear strength controls TIW activity and resultant eddy heat flux. The small (large) variability in 14–50-day period is consistent with the cooling (warming) anomaly of the eddy heat flux at the equator, because it would reduce (increase) equatorward eddy heat flux. Table 2 presents a quantitative comparison of the four indices, showing the ratio to the climatology. The eddy heat flux, 14–50-day variability in the SST gradient, and meridional shear are reduced to 66%–83% of their climatology, which roughly shows quantitative coincidence. Note that 113% in the meridional current is due to the large amplitude at the end of 2000. In mid-2000 and mid-2001, negative anomalies are indicated (Fig. 6b).
Figure 6c demonstrates that shear strength in the pausing period negatively correlates with the zonal current anomalies at the equator. Eastward (westward) current anomalies, which indicate weak (strong) SEC, accompany the weak (strong) shear in the pausing period. As a result, the cooling anomaly by the eddy heat flux is attributable to the eastward current anomalies at the equator.
The eastward current anomalies can be explained by large-scale variability during the transition from La Niña to El Niño. Longitude–time diagrams of surface zonal current anomalies are shown in Fig. 7 along with SST, surface wind, and 20°C isothermal depth anomalies. During La Niña from 1999 to 2001, SST anomalies are negative in the central and eastern parts of the basin (Fig. 7a) with easterly wind anomalies in the west (Fig. 7b). Accompanied by the strengthened trade winds, 20°C isothermal depth anomalies have a deeper thermocline in the west and a shallower thermocline in the east; that is, the zonal tilt of the thermocline is stronger than normal (Fig. 7c). These anomalies reverse their signs around 2002 when El Niño appears. The deep thermocline in the west slowly extends to the east during the pausing period. This extension is associated with the deepening of the thermocline over the basin due to the recharge process.
In addition to the La Niña to El Niño transition, all anomalies show prominent annual vacillations. The SST, wind, and thermocline depth anomalies have large amplitudes in northern winter and small amplitudes in summer, from 1999 to 2001. Zonal current anomalies relate to these annual vacillations. In the first halves of the years, eastward current anomalies appear from the date line to 100°W, concurrent with relaxation of the easterly (Fig. 7d). At the interannual time scale, the zonal wind stress and zonal pressure gradient associated with thermocline tilt are nearly balanced in the upper ocean, and their slight imbalance forces currents (Philander and Pacanowski 1981). When the easterly relaxes, the eastward pressure gradient would be stronger than the westward wind stress; therefore, the current anomaly would be eastward, according to the pressure gradient. The reverse situation is found for the latter halves of the years, when the westward current anomalies appear in concurrence with intensification of the easterly (Fig. 7d). During the period from 1999 to 2003, the easterly wind anomalies gradually relax and turn to the westerly anomaly as the La Niña switches to El Niño. Thus, the eastward current anomalies prevail more than the westward current anomalies; that is, the SEC tends to be weak. This weak SEC, which is inevitable during transition from La Niña to El Niño, weakens the TIWs and causes the cooling anomaly in the eddy heat flux.
b. Advection of climatological temperature gradient by anomalous currents
Another component which contributes to horizontal heat advection is the advection of climatological temperature by current anomalies (−u′H ·
These results show that advection by current anomalies can cause a large cooling anomaly, but that the warming anomaly is small in the pausing period. Note that this feature is confined to the pausing period; for example, the eastward current anomalies induce an intense warming anomaly in the latter half of 2002.
5. Discussion
a. Longer time series and spatial distribution
In this study we analyze observational data at the center of the Niño-3.4 region (0°, 140°W) in order to show the effect of the TIW on slowing down the transition from the 1998 La Niña to the 2002 El Niño. This is a case study examining the one ENSO event. Longer data should be examined in order to assess whether the effect of the TIW on ENSO is universal, because individual ENSO events differ significantly (Wallace et al. 1998). With respect to the spatial distribution, the SST anomalies associated with ENSO are equatorially trapped and zonally wide (Rasmusson and Carpenter 1982). It is necessary to examine the longitudinal distribution of the TIW effect. In this section, we discuss these two issues.
1) Longer time series
In this study our analysis period is from 1998 to 2002. This choice is due to the limitation of the meteorological observations by TAO/TRITON buoys, which confines the heat balance analysis to 1991 to 2005. On the other hand, the amplitude of high-frequency variability can be examined for a longer period. Surface current data are available from 1984 at 0°, 140°W. Weekly NOAA OI SST data are also available from 1982. We discuss universality and limitation of the TIW effect using these data.
Figure 9 shows the Niño-3.4 SST anomaly and thermocline depth anomaly from 1981 to 2005, along with the interannual anomaly of the wavelet power in 14–50-day period. A long transition can be found from the 1988 La Niña to the 1991 El Niño. During these events, the thermocline depth anomaly becomes deep in 1989, but El Niño do not occur until 1991 (arrow 2 in Fig. 9a). The interannual anomaly of amplitude of the 14–50-day variability in the meridional gradient of SST shows weak variability from 1989 to 1991, except in the later half of 1990 (solid line in Fig. 9b). Note that the meridional current is lacking in this period (dotted line in Fig. 9b). This result suggests that the slow-down effect, due to the TIWs, could have worked during the slow transition from the 1988 La Niña to the 1991 El Niño.
There are some events during which the mechanism proposed in this study seems not to be effective. One example is the transition from the 1983 La Niña to the 1986/87 El Niño (arrows 1). The 14–50-day variability is stronger than normal in 1985, although La Niña begins to decay with the deepened thermocline in that year. In this study, the condition that determines the effectiveness of the slow-down effect of the TIWs is not clarified. Precise investigation is needed regarding the dynamics and thermodynamics of TIW interannual variation. Analysis of results from high-resolution numerical modeling is planned in order to examine this question.
The onset process of the 1997 El Niño should be mentioned here. The thermocline depth becomes deeper than normal in 1996 with stronger-than-normal high-frequency variability; onset of El Niño occurs in early 1997 (arrows 3). Many studies have suggested that extremely strong westerly wind bursts played a crucial role in the onset of the 1997 El Niño (McPhaden and Yu 1999; Boulanger et al. 2001; Zhang and Gottschalck 2002; Belamari et al. 2003). It is possible that the influence of the wind bursts is more important than the TIWs with regard to this event.
In contrast, wind bursts may not be crucial in determining the onset timings of 1991 El Niño and 2002 El Niño. Regarding the 2002 El Niño, westerly wind events are found from 1999 to 2002, but the onset of El Niño occurs in 2002 (McPhaden 2004). A similar situation is found for the 1991 El Niño. Prominent wind bursts had already appeared in 1989 and 1990, and the onset of El Niño is in 1991 (Bergman et al. 2001). The 1991 El Niño and 2002 El Niño are events in which the slow-down effect, due to the TIWs, is relatively strong. The effectiveness of the wind bursts and TIWs and their relative importance should be investigated in future study using numerical experiments.
2) Longitudinal distribution
The results of this study are obtained from the heat balance at the center of the Niño-3.4 region (0°, 140°W). In this section, we compare the result at 0°, 140°W with those at the western edge (0°, 170°W) and eastern side (0°, 110°W) of the Niño-3.4 region. The horizontal advection anomalies [−(uH · ∇HT)′] and the eddy heat flux anomalies [Q′eddy] are compared for these three locations in Fig. 10. The amplitudes of the 14–50-day variability, calculated from the meridional gradient of SST, are also shown in the figure.
At 110°W, the horizontal advection and eddy heat flux show a cooling anomaly in 2001; the amplitude of the 14–50-day variability is small in this year (dotted lines in Fig. 10). Both 110° and 140°W share this feature, which suggests that the cooling anomaly due to weak TIWs is present over extensive regions. In the latter half of 2000, a warming anomaly is found at 110°W despite the weak high-frequency variability. This feature does not fit the mechanisms proposed in this study and requires further investigation.
At 170°W, the interannual anomaly of the amplitude of the 14–50-day variability is slightly negative in the pausing period (dashed lines in Fig. 10c). The associated cooling anomaly is absent in the horizontal advection and eddy heat flux, which show warming anomalies (dashed line in Figs. 10a,b). Chelton et al. (2000) and Contreras (2002) indicated that energetic TIWs are confined to the central and eastern Pacific Ocean. The effect of TIWs would thus be lacking in the western region. This speculation accords with the zonal lag in the time evolution of SST anomalies (Fig. 7a). Warm SST anomalies appear earlier in the western regions (160°E–160°W) than in the eastern regions (160°–100°W). The lack of cooling anomalies in the west might contribute to the early appearance of the positive SST anomalies.
b. Wavelet analysis
In this study, we suggest the weakened TIWs in the pausing period using the wavelet analysis. An examination of satellite-based NOAA OI SST gives qualitative confirmation of this suggestion. Figure 11a shows the interannual anomaly of the wavelet power in 14– 50-day period in the meridional gradient of SST (∂T/∂y), which shows positive anomalies in 1998 and 1999, indicating above-normal active variability. A longitude–time diagram of weekly NOAA OI SST at 2.5°N shows clear westward-propagating signals in this year (Fig. 11b), representing the TIWs (Contreras 2002). In contrast, the amplitudes of the high-frequency variability are small in the pausing period (Fig. 11a). Correspondingly, westward-propagating signals are less clear in 2000 (Fig. 11c). This result confirms that the weak TIWs in the pausing period derived by the wavelet analysis is qualitatively realistic.
The wavelet analysis possibly has a caveat in quantitative estimation of the TIWs variability. This problem appears in the eddy heat flux. To evaluate contribution from 14–50-day variability to the eddy heat flux anomaly, the current and SST variability in this period band is extracted using the wavelet analysis; then the eddy heat flux is calculated from them. The result is shown by the solid line in Fig. 12a with the total eddy heat flux anomaly Q′eddy. The eddy heat flux calculated from the 14–50-day variability shows cooling anomalies in mid-2000 and mid-2001, which coincide with the total eddy heat flux anomaly (open arrows). This coincidence suggests the importance of the 14–50-day variability in the eddy heat flux anomaly. However, its amplitude is about half of the total eddy heat flux anomalies. This underestimation gradually vanishes as the cutoff period of the wavelet filter is lengthened from 50 days to 270 days (Figs. 12b–d). It is possible that the energy in the 14–50-days-period band leaks to the longer period band because of the imperfectness of the wavelet filter. However, at least it can be confirmed that the total eddy heat flux anomaly is dominated by the contribution from the variability with period shorter than 270 days (Fig. 12d), that is, the contribution from the high-frequency components, −( uhighH · ∇HThigh)′. The contribution from the interannual anomalies [−(u′H · ∇HT ′)′] is small.
c. Discussion about the transition from El Niño to La Niña
As is noted in the introduction, the transition from La Niña to El Niño is slower than that from El Niño to La Niña. This asymmetry implies that some asymmetry exists in the process of the two transition phases. This issue is not fully clarified in this study, but we try to specify the problem and present our future plan in this section. For this purpose the analysis is extended to the 1997 and 1998, when El Niño switches to La Niña.
Figure 13 shows the indices denoting the ENSO and TIWs variability during 1997 and 1998. The hatched region in the figure is the transition period from 1997 El Niño to 1998 La Niña, which is the same as that in Fig. 2. The positive Niño-3.4 SST anomaly turns to be negative with shoaling of the thermocline in the transition period (Fig. 13a), accompanied by the westward current anomaly on the equator and the strong meridional shear (Fig. 13b, filled arrows). If the mechanism proposed in this study were applied with the opposite sign, it is expected that the strengthened shear accompanies the active TIWs and warming anomaly of the horizontal heat advection. However, such tendency is not found in the transition period, but found after it (Figs. 13c,d, open arrows). As a result the warming anomaly seems to limit the amplitude of the negative SST anomaly, rather than to slow down the transition.
This disagreement between the shear strength and the TIW activity is an unusual feature because analysis for 10 yr shows the apparent correlation between them (Baturin and Niiler 1997; Johnson and Proehl 2004). The interannual variation of the TIWs might be controlled by other process in the transition period from El Niño to La Niña. Hansen and Paul (1984), McCreary and Yu (1992), and Baturin and Niiler (1997) suggest that shear between the Equatorial Undercurrent and SEC, or meridional temperature gradient in the upper ocean, is another possible energy source of the TIWs. To investigate this issue, we plan to examine dynamics of the TIWs using a high-resolution numerical model.
6. Summary
In this study we analyze the heat balance in the surface mixed layer at the eastern equatorial Pacific Ocean (0°, 140°W) from the 1998 La Niña to the 2002 El Niño, using TAO/TRITON moored buoy observations. The results show that the interannual variation of the tropical instability waves (TIWs) slows down the transition from La Niña to El Niño. Kessler (2002) described this slow transition as the pausing period of the ENSO cycle, that is, when La Niña lingers and El Niño does not immediately appear, despite a deepened thermocline.
The result of the heat balance analysis shows that the warming anomalies of the surface heat flux and vertical advection cause the rise of the mixed layer temperature anomaly from the 1998 La Niña to the 2002 El Niño. The warming anomaly of the vertical heat advection is related to the deepening of the thermocline due to the recharge process, which causes the La Niña to El Niño transition. The warming anomaly of surface heat flux represents a damping effect on the negative SST anomalies during La Niña. In contrast to these effects, horizontal advection shows a cooling anomaly in 2000 and 2001, which reduces the warming anomalies and suppresses the rise of the mixed layer temperature anomaly. This cooling effect substantially slows down the transition from La Niña to El Niño.
The energetic high-frequency variability in the equatorial eastern Pacific (i.e., the TIWs) significantly contributes to the cooling anomaly. Amplitudes of 14–50-day variability in the meridional current and SST gradient are small in 2000 and 2001, concurrent with a cooling anomaly of the eddy heat flux. These results suggest that weakened TIWs cause a reduction of the equatorward eddy heat flux, which results in the cooling anomaly at the equator. A similar feature is also found at 0°, 110°W. During the transition from La Niña to El Niño, eastward current anomalies (i.e., weakened SEC) prevail at the equator, associated with the relaxation of trade winds. The weakened SEC accompanies weak meridional shear between the SEC and NECC, which would suppress the TIW activity. These results suggest that synoptic-scale processes effectively work in the basin scale to slow down the transition from La Niña to El Niño.
Previous studies have emphasized that eastward current anomalies advect warm water eastward and cause a warming anomaly in the east (Harrison and Schopf 1984; Picaut et al. 1996, 1997). However, our results indicate that the warming anomaly due to this effect is small during the pausing period. Because of marked annual vacillations in wind anomalies, eastward current anomalies tend to appear in northern spring and summer. The zonal gradient of the climatological temperature is small in these seasons. This seasonality results in the smallness of heat advection by eastward current anomalies.
The above results would be able to apply to events other than the 2002 El Niño. The 1991 El Niño is a typical event, the onset timing of which does not coincide with deepening of the thermocline. Analysis shows the weak 14–50-day variability from 1989 to 1991. The slow-down effect, due to the TIWs, would have worked during the onset process of the 1991 El Niño, as well as the 2002 El Niño.
The results in this study contribute to clarify the process working in the slow transition from La Niña to El Niño. However, its counterpart, the rapid transition from El Niño to La Niña, is not yet fully examined. During the transition from 1997 El Niño to 1998 La Niña the slow-down effect of the TIWs seems not to work, but the reason for it is still an open question. Our next task is examinations of the asymmetry between El Niño and La Niña, and dynamics in the interannual variation of the TIWs.
Acknowledgments
We thank the TAO and TRITON project offices for providing observational data from the mooring buoys. We also appreciate use of the output from OFES for error evaluation in the appendix and thank Y. Sasai and H. Sasaki for their kind collaboration. Data analysis and numerical computations were conducted using the Interactive Data Language (IDL) Library.
REFERENCES
Battisti, D. S., 1988: Dynamics and thermodynamics of a warming event in a coupled tropical atmosphere–ocean model. J. Atmos. Sci., 45 , 2889–2919.
Baturin, N. G., and P. P. Niiler, 1997: Effects of instability waves in the mixed layer of the equatorial Pacific. J. Geophys. Res., 102 , C13. 27771–27794.
Belamari, S., J-L. Redelsperger, and M. Pontaud, 2003: Dynamic role of a westerly wind burst in triggering an equatorial Pacific warm event. J. Climate, 16 , 1869–1890.
Bergman, J. W., H. H. Hendon, and K. M. Weickmann, 2001: Intraseasonal air–sea interactions at the onset of El Niño. J. Climate, 14 , 1702–1719.
Bonjean, F., and G. S. E. Lagerloef, 2002: Diagnostic model and analysis of the surface currents in the tropical Pacific Ocean. J. Phys. Oceanogr., 32 , 2938–2954.
Boulanger, J-P., and Coauthors, 2001: Role of non-linear oceanic processes in the response to westerly wind events: New implications for the 1997 El Niño onset. Geophys. Res. Lett., 28 , 1603–1606.
Brady, E. C., 1994: Interannual variability of meridional heat transport in a numerical model of the upper equatorial Pacific Ocean. J. Phys. Oceanogr., 24 , 2675–2694.
Bryden, H. L., and E. C. Brady, 1989: Eddy momentum and heat fluxes and their effects on the circulation of the equatorial Pacific Ocean. J. Mar. Res., 47 , 55–79.
Chang, P., 1993: Seasonal cycle of sea surface temperature and mixed layer heat budget in the tropical Pacific Ocean. Geophys. Res. Lett., 20 , 2079–2082.
Chelton, D. B., F. J. Wentz, C. L. Gentemann, R. A. de Szoeke, and M. G. Schlax, 2000: Satellite microwave SST observations of transequatorial tropical instability waves. Geophys. Res. Lett., 27 , 1239–1242.
Clark, N. E., L. Eber, R. M. Laurs, J. A. Renner, and J. F. T. Saur, 1974: Heat exchange between ocean and atmosphere in the eastern North Pacific for 1961–71. NOAA Tech. Rep. NMFS SSRF-682, Washington, DC, 108 pp.
Contreras, R. F., 2002: Long-term observations of tropical instability waves. J. Phys. Oceanogr., 32 , 2715–2722.
Cox, M. D., 1980: Generation and propagation of 30-day waves in a numerical model of the Pacific. J. Phys. Oceanogr., 10 , 1168–1186.
Emery, W. J., and R. E. Thomson, 1998: Data Analysis Methods in Physical Oceanography. Pergamon, 634 pp.
Fairall, C. W., E. F. Bradley, D. P. Rogers, J. B. Edson, and G. S. Young, 1996: Bulk parameterization of air-sea fluxes for Tropical Ocean-Global Atmosphere Coupled-Ocean Atmosphere Response Experiment. J. Geophys. Res., 101 , C2. 3747–3764.
Flament, P. J., S. C. Kennan, R. A. Knox, P. P. Niiler, and R. L. Bernstein, 1996: The three-dimensional structure of an upper ocean vortex in the tropical Pacific Ocean. Nature, 383 , 610–613.
Halpern, D., R. A. Knox, and D. S. Luther, 1988: Observations of 20-day period meridional current oscillations in the upper ocean along the Pacific equator. J. Phys. Oceanogr., 18 , 1514–1534.
Hansen, D. V., and C. A. Paul, 1984: Genesis and effects of long waves in the equatorial Pacific. J. Geophys. Res., 89 , C6. 10431–10440.
Harrison, D. E., and P. S. Schopf, 1984: Kelvin-wave-induced anomalous advection and the onset of surface warming in El Niño events. Mon. Wea. Rev., 112 , 923–933.
Hasegawa, T., and K. Hanawa, 2003: Heat content variability related to ENSO events in the Pacific. J. Phys. Oceanogr., 33 , 407–421.
Hayes, S. P., P. Chang, and M. J. McPhaden, 1991: Variability of the sea surface temperature in the eastern equatorial Pacific during 1986–1988. J. Geophys. Res., 96 , C6. 10553–10566.
Jin, F-F., 1997a: An equatorial ocean recharge paradigm for ENSO. Part I: Conceptual model. J. Atmos. Sci., 54 , 811–829.
Jin, F-F., 1997b: An equatorial ocean recharge paradigm for ENSO. Part II: A stripped-down coupled model. J. Atmos. Sci., 54 , 830–847.
Jin, F-F., 2001: Low-frequency modes of tropical ocean dynamics. J. Climate, 14 , 3874–3881.
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 , 19. 2989–2992.
Johnson, E. S., and J. A. Proehl, 2004: Tropical instability wave variability in the Pacific and its relation to large-scale currents. J. Phys. Oceanogr., 34 , 2121–2147.
Johnson, G. C., M. J. McPhaden, and E. Firing, 2001: Equatorial Pacific Ocean horizontal velocity, divergence, and upwelling. J. Phys. Oceanogr., 31 , 839–849.
Johnson, G. C., B. M. Sloyan, W. S. Kessler, and K. E. McTaggart, 2002: Direct measurements of upper ocean currents and water properties across the tropical Pacific during the 1990s. Prog. Oceanogr., 52 , 31–61.
Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77 , 437–471.
Kessler, W. S., 2002: Is ENSO a cycle or a series of events? Geophys. Res. Lett., 29 .2125, doi:10.1029/2002GL015924.
Legeckis, R., 1977: Long waves in the eastern equatorial Pacific Ocean: A view from a geostationary satellite. Science, 197 , 1179–1181.
Liebmann, B., and C. A. Smith, 1996: Description of a complete (interpolated) outgoing longwave radiation dataset. Bull. Amer. Meteor. Soc., 77 , 1275–1277.
Liu, W. T., A. Zhang, and J. K. B. Bishop, 1994: Evaporation and solar irradiance as regulators of sea surface temperature in annual and interannual changes. J. Geophys. Res., 99 , C6. 12623–12638.
Masumoto, Y., and Coauthors, 2004: A fifty-year eddy-resolving simulation of the world ocean—–Preliminary outcomes of OFES (OGCM for the Earth Simulator). J. Earth Simul., 1 , 35–56.
McCreary, J. P., and Z. Yu, 1992: Equatorial dynamics in a 2 1/2-layer model. Prog. Oceanogr., 29 , 61–132.
McPhaden, M. J., 1996: Monthly period oscillations in the Pacific North Equatorial Countercurrent. J. Geophys. Res., 101 , C3. 6337–6360.
McPhaden, M. J., 2002: Mixed layer temperature balance on intraseasonal timescales in the equatorial Pacific Ocean. J. Climate, 15 , 2632–2647.
McPhaden, M. J., 2004: Evolution of the 2002/03 El Niño. Bull. Amer. Meteor. Soc., 85 , 677–695.
McPhaden, M. J., and X. Yu, 1999: Equatorial waves and the 1997–98 El Niño. Geophys. Res. Lett., 26 , 19. 2961–2964.
McPhaden, M. J., and Coauthors, 1998: The Tropical Ocean-Global Atmosphere observing system: A decade of progress. J. Geophys. Res., 103 , C7. 14169–14240.
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 , 3551–3559.
Meinen, C. S., and M. J. McPhaden, 2001: Interannual variability in warm water volume transports in the equatorial Pacific during 1993–99. J. Phys. Oceanogr., 31 , 1324–1345.
Mestas-Nuñez, A. M., A. Bentamy, and K. B. Katsaros, 2006: Seasonal and El Niño variability in weekly satellite evaporation over the global ocean during 1996–98. J. Climate, 19 , 2025–2035.
Philander, S. G. H., 1978: Instabilities of zonal equatorial currents, 2. J. Geophys. Res., 83 , C7. 3679–3682.
Philander, S. G. H., and R. C. Pacanowski, 1981: Response of equatorial oceans to periodic forcing. J. Geophys. Res., 86 , C3. 1903–1916.
Philander, S. G. H., and W. J. Hurlin, 1988: The heat budget of the tropical Pacific Ocean in a simulation of the 1982–83 El Niño. J. Phys. Oceanogr., 18 , 926–931.
Picaut, J., M. Ioualalen, C. Menkes, T. Delcroix, and M. J. McPhaden, 1996: Mechanism of the zonal displacements of the Pacific Warm Pool: Implications for ENSO. Science, 274 , 1486–1489.
Picaut, J., F. Masia, and Y. du Penhoat, 1997: An advective–reflective conceptual model for the oscillatory nature of the ENSO. Science, 277 , 663–666.
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 , 354–384.
Reynolds, R. W., N. A. Rayner, T. M. Smith, D. C. Stokes, and W. Wang, 2002: An improved in situ and satellite SST analysis for climate. J. Climate, 15 , 1609–1625.
Sasaki, H., M. Nonaka, Y. Masumoto, Y. Sasai, H. Uehara, and H. Sakuma, 2007: An eddy-resolving hindcast simulation of the quasi-global ocean from 1950 to 2003 on the Earth Simulator. High Resolution Numerical Modelling of the Atmosphere and Ocean, K. Hamilton and W. Ohfuchi, Eds., Springer, 157–186.
Schopf, P. S., and M. J. Suarez, 1988: Vacillations in a coupled ocean–atmosphere model. J. Atmos. Sci., 45 , 549–566.
Smith, N. R., 1995: An improved system for tropical ocean subsurface temperature analyses. J. Atmos. Oceanic Technol., 12 , 850–870.
Suarez, M. J., and P. S. Schopf, 1988: A delayed action oscillator for ENSO. J. Atmos. Sci., 45 , 3283–3287.
Torrence, C., and G. P. Compo, 1998: A practical guide to wavelet analysis. Bull. Amer. Meteor. Soc., 79 , 61–78.
Trenberth, K. E., 1997: The definition of El Niño. Bull. Amer. Meteor. Soc., 78 , 2771–2777.
Vialard, J., C. Menkes, J-P. Boulanger, P. Delecluse, E. Guilyardi, M. J. McPhaden, and G. Madec, 2001: A model study of oceanic mechanisms affecting equatorial Pacific sea surface temperature during the 1997–98 El Niño. J. Phys. Oceanogr., 31 , 1649–1675.
Waliser, D. E., B. Blanke, J. D. Neelin, and C. Gautier, 1994: Shortwave feedbacks and El Niño–Southern Oscillation: Forced ocean and coupled ocean-atmosphere experiments. J. Geophys. Res., 99 , C12. 25109–25126.
Wallace, J. M., E. M. Rasmusson, T. P. Mitchell, V. E. Kousky, E. S. Sarachik, and H. von Storch, 1998: On the structure and evolution of ENSO-related climate variability in the tropical Pacific: Lessons from TOGA. J. Geophys. Res., 103 , C7. 14241–14260.
Wang, W., and M. J. McPhaden, 1999: The surface-layer heat balance in the equatorial Pacific Ocean. Part I: Mean seasonal cycle. J. Phys. Oceanogr., 29 , 1812–1831.
Wang, W., and M. J. McPhaden, 2000: The surface-layer heat balance in the equatorial Pacific Ocean. Part II: Interannual variability. J. Phys. Oceanogr., 30 , 2989–3008.
Wang, W., and M. J. McPhaden, 2001: What is the mean seasonal cycle of surface heat flux in the equatorial Pacific? J. Geophys. Res., 106 , C1. 837–858.
Weisberg, R. H., and C. Wang, 1997: A western Pacific oscillator paradigm for the El Niño-Southern Oscillation. Geophys. Res. Lett., 24 , 7. 779–782.
Wyrtki, K., 1985: Water displacements in the Pacific and the genesis of El Niño cycles. J. Geophys. Res., 90 , C4. 7129–7132.
Zelle, H., G. Appeldoorn, G. Burgers, and G. J. van Oldenborgh, 2004: The relationship between sea surface temperature and thermocline depth in the eastern equatorial Pacific. J. Phys. Oceanogr., 34 , 643–655.
Zhang, C., and J. Gottschalck, 2002: SST anomalies of ENSO and the Madden–Julian oscillation in the equatorial Pacific. J. Climate, 15 , 2429–2445.
APPENDIX
Error in Heat Balance
In this appendix, the errors in the heat balance calculation are evaluated. It is considered that the error is caused by three sources: 1) measurement error of the sensors mounted on the mooring buoy, 2) sampling error caused by the sparse observation of the mooring buoy, and 3) uncertainty in the formulation of the heat balance equation. In the next two sections, we evaluate these errors. In the last section, we show the resultant errors and the closeness of the heat balance.
The errors 2) and 3) are estimated by examination of output of numerical model. The brief summary of the results is presented here, and detail will be shown in a separate paper which is now under preparation.
Measurement error
The measurement error was already estimated by Wang and McPhaden (2000, 2001). The errors for the mixed layer depth, mixed layer temperature, horizontal current, and surface heat flux are presented in their paper based on accuracy of the sensors on the buoy. The error for NOAA OI SST is also shown by them. In addition, we consider the uncertainty in the climatological vertical current (
Sampling and formulation error
The mooring buoy observes a point in the ocean, which can cause the sampling error, namely alias or bias. In addition, the NOAA OI SST with 1° resolution is used for the calculation of the horizontal heat advection, which is another source of sampling error. The formulation error would arise from the vertical heat advection term. Because the vertical velocity cannot be observed directly, it is estimated by the rough method. These errors are evaluated using outputs from high-resolution numerical model.
The used model is the Ocean general circulation model For Earth Simulator (OFES), with 0.1° × 0.1° resolution and 54 vertical levels (Masumoto et al. 2004). The model has the realistic topography and is forced by the daily averages of the NCEP reanalysis (Sasaki et al. 2007). First, the heat balance terms are estimated using Eqs. (1) and (2), with the model data at one grid point and the spatially averaged model SST over the 1° × 1° boxes. This mimics the one-point observation by the mooring buoy and the spatially averaged NOAA OI SST. This method is referred to as the “buoy” method in the rest of this section. Second, the heat budget terms are calculated by the exact method at each horizontal grid and averaged over the 2° × 2° boxes. Note that the vertical velocity is directly obtained from the model output in the exact method. The difference of the results between the two methods represents the sampling and formulation error.
The differences between the two methods are quite small for the time rate of change of temperature and surface heat flux terms. The root-mean-squares (rms) of their differences are only 0.05 and 0.08°C month−1, respectively. The horizontal heat advection is characterized by the underestimation by the buoy method, which amounts to 0.25°C month−1 in rms. This would be caused by the use of the spatially averaged SST, which reduces the small-scale variability and the eddy heat flux. This result suggests that the effect of the TIWs can be larger than that presented in Fig. 5 and Fig. 6. The rms difference of the vertical heat advection term is 0.39°C month−1.
Although these evaluations are based on the numerical model, the results give a measure to uncertainty in the observation. Note that we had confirmed that the heat budget terms in the OFES have the similar amplitudes to those in the observation, which supports the validity of the error evaluation using the OFES.
Closeness of the heat budget
Considering the errors presented in the appendix sections a and b, we examine the closeness of the heat balance. The measurement, sampling and formulation errors are combined using the sampling theory for propagation of error by Emery and Thomson (1998), except for the horizontal heat advection term. For the horizontal heat advection, the random measurement error and the underestimation error are simply summed.
The solid line in Fig. A1a shows the summation of the temperature tendency and the advection terms. The dashed line shows the surface heat flux term. The discrepancy between them is equal to the residual (R′). The vertical lines and light shading illustrate the evaluated errors. The result shows that the solid and dashed lines show the similar tendency of decrease with time, and their amplitudes correspond within the errors.
However, the residual is not negligible. Figure A1b repeats Fig. 3b with the residual shown by the heavy black line. The amplitude of the residual is not small compared to the other terms in the heat balance. As is noted in the appendix section b, the eddy heat flux anomalies tend to be underestimated by our method. The negative anomaly of the residual in the first halves of 2000 and 2001 (Fig. A1b, open arrows) might be partly explained by the underestimation of the cooling anomaly of the eddy heat flux around these times (Fig. 6, open arrows). In addition, the residual and the surface heat flux terms tend to be negatively correlated from 1999 to 2001. The residual might be associated with uncertainty in the bulk formulation of the surface heat flux or the definition of the mixed layer depth.
SST anomaly (°C) averaged over Niño-3.4 region (5°S–5°N, 170°–120°W) (solid line) and 20°C isothermal depth anomaly (m) averaged over the equatorial Pacific Ocean (5°S–5°N, 130°E–80°W) (dashed line). The SST and 20°C isothermal depth are obtained from NOAA OI SST (Reynolds et al. 2002) and Bureau of Meteorology Research Centre ocean analysis (Smith 1995), respectively. The anomaly is defined as a deviation from the monthly climatology from 1991 to 2004. The 5-month running mean and 3-month triangle filter are applied.
Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1765.1
Niño-3.4 SST anomaly (°C) (solid line) and 20°C isothermal depth (m) anomaly averaged over the equatorial Pacific Ocean (dashed line). The red (blue) line indicates El Niño (La Niña) event. The solid light shading shows the pausing period from 1998 La Niña to 2002 El Niño. The hatched shading is transition period from 1997 El Niño to 1998 La Niña, which is a period of negative 20°C isothermal depth anomaly and positive Niño-3.4 SST anomaly.
Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1765.1
(a) Mixed layer temperature anomaly (T ′, °C) at 0°, 140°W (solid line). The dotted line shows mixed layer temperature anomaly integrated from the beginning of the pausing period without horizontal heat advection term. (b) Heat balance (°C month−1) in the surface mixed layer at 0°, 140°W. The red line shows the surface heat flux term [(Qsurface/ρ0cph)′]. The green and blue lines show the horizontal heat advection term [−(uH · ∇HT)′] and vertical heat advection term [−(w*ΔT/h)′], respectively. Error bars are shown for each term. The detail of the error estimation is presented in the appendix. The light shading denotes the pausing period.
Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1765.1
(a) Vertical heat advection [(w*ΔT/h)′] in the heat balance (solid) and the 20°C isothermal depth anomalies (dotted) at 0°, 140°W. (b) Surface heat flux term [(Qsurface/ρ0cph)′] (solid) and the mixed layer temperature anomalies (dotted) at 0°, 140°W. Error bars are shown for the terms in the heat balance. The light shade shows the pausing period. Units are °C month−1 and m, respectively.
Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1765.1
(a) Advection of climatological temperature by current anomalies (−u′H ·
Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1765.1
(a) Eddy heat flux anomaly Q′eddy (°C month−1) at 0°, 140°W. (b) Interannual anomaly of wavelet power in 14–50-day-period bands; the solid line is the meridional gradient of SST [°C (100 km)−1] and the dotted line the meridional current (m s−1). (c) Difference of surface zonal current anomalies (m s−1) between 2.5° and 7.5°N along 140°W from OSCAR (solid line). The dashed line shows zonal current anomalies at 0°, 140°W from buoy. The light shading shows the pausing period. The open (filled) arrows point mid-2000 and mid-2001 (end of 2000) similarly to those in Fig. 7.
Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1765.1
Longitude–time diagrams of (a) SST anomalies (°C), (b) surface zonal wind anomalies (m s−1), (c) 20°C isothermal depth anomalies (m), and (d) surface zonal current anomalies (m s−1) averaged over 1°S–1°N. Gray shading between the panels show the pausing period.
Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1765.1
(a) Advection of climatological temperature by current anomalies (−u′H ·
Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1765.1
(a) Niño-3.4 SST anomaly (solid, °C) and 20°C isotherm depth anomaly (m) averaged over the equatorial Pacific Ocean (dashed). (b) Interannual anomaly of wavelet power in 14–50-day period at 0°, 140°W; the solid line is the meridional gradient of SST [°C (100 km)−1] and the dotted line the meridional current (m s−1). The numbered arrows are same as those in Fig. 1.
Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1765.1
(a) Horizontal heat advection [−(uH · ∇HT)′], (b) eddy heat flux anomalies (Q′eddy), and (c) interannual anomaly of the wavelet power in 14–50-day-period band in the meridional gradient of SST. The solid line shows the time series at 0°, 140°W. The dashed and dotted lines are for 0°, 170°W and 0°, 110°W, respectively. Units are °C month−1 and °C (100 km)−1, respectively.
Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1765.1
(a) Interannual anomaly of wavelet power in 14–50-day-period band for meridional gradient of SST [°C (100 km)−1]. The light shading shows the pausing period. (b) Longitude–time diagram of weekly SST (°C) at 2.5°N from April 1998 to March 1999. (c) Same as (b), but from April 2000 to March 2001.
Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1765.1
(a) Eddy heat flux anomaly calculated from the variability in the 14–50-day-period band (solid line). (b)–(d) The same as (a), but the cutoff period of the wavelet filter is (b) 100, (c) 200, and (d) 270 days. The dashed lines in all panels show the total eddy heat flux anomalies (Q′eddy). The light shading shows the pausing period. Units are °C month−1.
Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1765.1
Indices during the transition from 1997 El Niño to 1998 La Niña. (a) Niño-3.4 SST anomaly (solid line, °C) and 20°C isothermal depth anomaly averaged over the equatorial Pacific Ocean (dashed line, m). (b) Difference of surface zonal current anomalies (m s−1) between 2.5° and 7.5°N along 140°W from OSCAR (solid line) and the zonal current anomalies at 0°, 140°W from buoy (dashed line). (c) Interannual anomaly of wavelet power in 14–50-day period at 0°, 140°W; the solid line is the meridional gradient of SST [°C (100 km)−1] and the dotted line the meridional current (m s−1). (d) Horizontal advection anomalies [−(uH · ∇HT)′, °C month−1].
Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1765.1
Fig. A1. Closeness of the heat balance calculation at 0°, 140°W. (a) Summation of the temperature tendency, and the horizontal and vertical advection terms—∂T ′/∂t + (uH · ∇HT)′ + 〈H(w*)w*{[T − T(h + 20 m)]/h}〉′—(solid line), and surface heat flux term (Qsurface/ρ0cph)′ (dashed line). Their error bars are shown by the vertical lines and light shading. (b) The residual in the heat balance at 0°, 140°W (heavy black line). The red, green, and blue lines are same as those in Fig. 3b. Units are °C month−1.
Citation: Journal of Climate 21, 2; 10.1175/2007JCLI1765.1
Horizontal and temporal resolutions and available periods of the datasets used in this study.
Ratios to the climatological mean. The ratios are averaged value over the pausing period.
