1. Introduction
It has been known that strong decadal variations can be found in the Kuroshio–Oyashio Extension (KOE) region, characterized by extremely cool sea surface temperature (SST) anomalies in the 1980s (e.g., White 1995; Deser et al. 1996; Schneider and Miller 2001). Using a coupled general circulation model, Latif and Barnett (1994, 1996) suggested that a slow dynamic adjustment in upper ocean circulation by the first baroclinic mode of the Rossby wave controls the phase reversal of the decadal variation. A number of numerical and observational studies have detected decadal SST anomalies in the KOE region, due to the delayed ocean response over a few years to wind stress curl anomalies (e.g., Miller et al. 1998; Deser et al. 1999; Venzke et al. 2000; Xie et al. 2000; Seager et al. 2001). It was also demonstrated that a fast thermal adjustment in the ocean mixed layer by winter heat flux anomalies at the sea surface enhanced the decadal SST anomalies in the midlatitudes of the North Pacific (Barsugli and Battisti 1998; Saravanan 1998). On the other hand, using a simple ocean model, Gu and Philander (1997) demonstrated that water temperature anomalies, enhanced by air–sea feedback in the extratropics, subduct into the Tropics (Zhang and Levitus 1997; Miller et al. 1998; Weaver 1999; Zhang and Liu 1999). It was also suggested that the advection of decadal temperature anomalies by a climatological gyre circulation was of great importance in the phase reversal of the decadal variations, although several subsequent studies have cast doubts on this as a major causal mechanism of the Pacific decadal variability (Schneider et al. 1999; Nonaka and Xie 2000; Nonaka et al. 2002).
In the KOE region, however, it has been pointed out that the horizontal temperature transport is so strong that the surface heat flux may damp the SST anomalies, rather than enhance the SST anomalies (e.g., Pierce et al. 2001; Seager et al. 2001; Schneider et al. 2002). In addition, Peng et al. (1995, 1997) pointed out that an atmospheric response to a typical SST anomaly around the KOE region has considerable dependence on the background atmospheric conditions, namely the seasonal and monthly differences.
Therefore, to understand the atmospheric behavior forced by a decadal SST anomaly in the KOE region, it is of great importance to examine the seasonality of the decadal SST anomalies in greater detail and to clarify the related physical mechanisms beyond the simple descriptions given in the previous studies. Namely, previous studies simply reported that decadal SST anomalies were observed to be much stronger in winter and weaker in summer (e.g., Nakamura and Yamagata 1999; Schneider et al. 2002).
The purpose of the present study is to clarify the physical processes controlling the seasonal variation of the decadal SST anomalies, defined as the 5-yr averaged SST anomalies deviating from the climatological monthly variations in the KOE region, in greater detail. To achieve these goals, a three-dimensional (3D) bulk mixed layer model with atmospheric properties at the sea surface as boundary conditions is integrated, and the results of the numerical solutions are then investigated.
The model design used in the present paper is described in section 2. In section 3, the temporal and spatial features of the fundamental physical variables of the model, such as the simulated SST and geostrophic velocity, are compared with the observation. The seasonality of the simulated decadal SST anomalies in the KOE region is briefly documented in section 4. The monthly heat budgets of the ocean mixed layer are examined and the related physical processes are clarified in section 5. Particularly, the contribution from the net sea surface heat flux is discussed in section 6. Concluding remarks are presented in section 7.
2. Model description and experiment design
a. Model description related to decadal variation
b. Integration of the model
The model coversd the region of 10°–70°N in the North Pacific with a 1° × 1° (latitude × longitude) grid. Daily atmospheric data (wind velocity, air temperature, and specific humidity) for near the sea surface from the NCEP–NCAR reanalysis dataset were utilized as boundary conditions. The model with dynamics governed by Eq. (4) was integrated from January 1950 to December 1999 with the anomalous geostrophic streamfunction set uniformly to zero (ψ′g = 0) as the initial condition. The bulk mixed layer model governed by Eq. (1) and the vertical diffusive model governed by Eq. (6) were integrated, independently from the dynamics, from January 1960 to December 1999. Data on SST anomalies for January 1960 were taken from the Simple Ocean Data Assimilation (SODA) dataset (Carton et al. 2000) and set as the initial conditions. Compiled monthly data from 1960 to 1999 were used in the following investigations. Considering the time scale of the Rossby wave adjustment of the first baroclinic mode in the extratropics, the first decade (1950–59) was excluded from the following analyses.
3. Fundamental features of the model results and observations
a. Sea surface temperature
Before detailed analyses were conducted, the temporal and spatial features of the fundamental physical variables simulated by the model were examined. The numerical solution of the standard run displayed a realistic decadal SST variation in the North Pacific Ocean. The 5-yr running means for each month (averaging the values for the same month over 5 years) were then applied to all of the simulated data to define the decadal anomalies as deviations from the climatological seasonal variations. The differences between 5-yr averages in the simulated SST (Fig. 1) were compared with those of observations (Fig. 2) to examine whether the simulated SST displayed typical changes of observed SST shown in the previous studies such as the late-1970s and late-1980s SST changes. The phases of the decadal SST anomalies around the central North Pacific reversed in the mid-1970s (Fig. 1a). As shown by Miller and Schneider (2000) and Seager et al. (2001) several years later, cold decadal SST anomalies appeared around the subtropical frontal (STF) region (Nakamura et al. 1997), defined here as 25°–35°N, 175°E–145°W (Fig. 1b). These anomalies correspond to well-known global climate changes in the late 1970s. After 5–7 yr (Fig. 1c), strong cold SST anomalies appeared over the KOE region, defined here as 35°–45°N, 150°E–180° in the present paper (Fig. 1d), as reported by White (1995) and Schneider and Miller (2001). These anomalies are thought to have been caused by slow changes in the strength of ocean gyre circulation corresponding to anomalous wind stress curl in the late 1970s (see section 1). The strong SST change in the KOE region in the late 1980s was also depicted (Fig. 1d). These temporal and spatial features of the simulated SST anomalies corresponded to those of the observed SST data acquired from the Japan Meteorological Agency (JMA: Japan Meteorological Agency 1990).
Figure 3 shows the time series of wintertime decadal variations in simulated and observed SST and those of the simulated mixed layer depth averaged for the KOE region. The simulated SST in the KOE region had significantly realistic decadal-scale time evolution. Furthermore, strong negative correlation relationships between the mixed layer depth anomaly and the SST anomaly were also found (correlation coefficient = −0.70), similar to what Deser et al. (1996) described in the central North Pacific. The mixed layer deepened in the cool SST phase (1975, 1985, and 1995) and shoaled in the warm SST phase (1970, 1980, and 1990).
b. Geostrophic advection
Since geostrophic advection makes crucial contributions to the upper-ocean heat budget in general, as well as in the region of the Kuroshio and its extension, geostrophic velocity calculated from the Ocean Topography Experiment (TOPEX)/Poseidon altimetry data was examined (see appendix A). The climatological geostrophic velocity field (Fig. 4) for the period 1993–2001 was considered acceptable, as it compared well with data that Vivier et al. (2002) estimated using the mean sea surface height (SSH) based on TOPEX/Poseidon altimetry data derived by Qiu (1995). The location of the strongest geostrophic current and intensity of the Kuroshio south of Japan were also consistent with Qiu (2000) and Vivier et al. (2002). Although the geostrophic velocity fields in Qiu (2000) and Vivier et al. (2002) were also calculated from the TOPEX/Poseidon datasets, the conversion methods to SSH data seemed to be different.
Figure 5 shows the time series of the decadal SST anomaly and the anomalous temperature transport caused by anomalous geostrophic advection to which a 60-month running mean was applied, averaged over the KOE region. The SST anomaly lagged the anomalous temperature transport by about 29 months (the lagged correlation coefficient was 0.87), consistent with findings from previous studies (e.g., Miller et al. 1998; Deser et al. 1999; Xie et al. 2000; Yasuda and Kitamura 2003). Corresponding to the positive (A, C) and negative (B, D) peaks of the temperature transport into the mixed layer, positive (A*, C*) and negative (B*, D*) peaks of the decadal SST anomalies were found a couple of years later. The amplitude of the temperature transport rate variation by the anomalous geostrophic advection was comparable, if integrated for one year, to that of the decadal SST variation, although the anomalous geostrophic velocity may have been slightly underestimated due to the coarse model resolution and nonlinear processes that were ignored. These concerns will be discussed in more detail in appendix B.
The equatorial region south of 10°N was excluded from the model domain because tropical ocean conditions are believed to affect decadal variations in the extratropics, generally via atmospheric transmissions. Impacts of the tropical ocean via the atmosphere were thus considered in the present study through the NCEP/NCAR reanalysis dataset that was used as a boundary condition for each grid point. The only exception to this pattern may be, as suggested by Jacobs et al. (1994), that the oceanic wave induced in the Tropics could change the strength of the subtropical ocean gyre circulation. The simulated anomalous heat transport by the geostrophic advection in this model, however, displayed a reasonable time series (Fig. 5) without considering the impacts of the tropical ocean condition through the ocean subsurface.
4. Seasonality of decadal SST anomalies
The coefficients displayed clear seasonal cycles, both by the model simulation and observations. When comparing the KOE region (solid lines) and the STF region (dashed lines), however, the seasonal variation phase of the decadal SST anomaly differed by a few months. The decadal SST anomaly in the STF region was strongest from February to April since decadal SST anomalies in the STF region were likely to be enhanced by wintertime air–sea interactions (see section 5: Nakamura and Yamagata 1999; Mochizuki and Kida 2003). On the other hand, in the KOE region the decadal SST anomaly was strongest from December to February. This result suggests that autumnal vertical entrainment, as well as local wintertime atmospheric forcing, is very important to the seasonal heat budget related to decadal SST anomalies in the KOE region. In addition, the amplitude of the seasonal modulation of the decadal SST anomaly was smaller in the KOE region than in the STF region. This would be due to differences in the strength of the local thermal damping rate (Watanabe and Kimoto 2000), resulting from differences in climatological mixed layer depth (water mass) and climatological surface wind speed. Details of these heat budget processes in the KOE region are discussed in the following sections.
5. Seasonal heat budget of the ocean mixed layer
To clarify the physical processes that control seasonal variation in decadal SST anomalies in the KOE region, the heat budget of the ocean mixed layer was examined for each month. Figure 7 shows the lagged regression coefficients of the decadal variations for each term in Eq. (1) of each month to the annual averaged decadal SST anomaly in the KOE region. Each of the five terms was normalized by the climatological mixed layer depth, because this study focused on the seasonal variations of the SST anomaly rather than those of the heat transport anomaly into the mixed layer, as the heat exchange rate at the sea surface is controlled by SST anomalies as well as the atmospheric conditions at the sea surface. Seasonal-scale enhancement and decay of the decadal SST anomaly were found from October to January and from May to August, respectively (Fig. 7a). The tendency of the SST anomaly in early winter was balanced by anomalous horizontal Ekman temperature transport (Fig. 7b) and the tendency in autumn was balanced by the anomalous vertical entrainment (Fig. 7d). However, the surface heat flux made a large contribution to the local summertime and autumnal thermal damping (Fig. 7e). Horizontal temperature transport by the geostrophic current was found to have no effect on seasonal variation in the decadal SST anomaly (Fig. 7c). On the other hand, in the STF region seasonal-scale enhancement of the decadal SST anomaly was found from December to February (Fig. 8a). The Ekman heat transport was the dominant contribution to the tendency of the SST anomaly in winter (Fig. 8b), and the vertical entrainment made only a slight contribution (Fig. 8d). The summertime damping effect of the surface heat flux was stronger than that of the KOE region (Fig. 8e).
To estimate the magnitude of the decadal SST anomaly formed by each term in Eq. (1) for autumn and winter, the numerical solutions of four control runs (experiments ML1–4) with differing temperature equations (Table 2) were compared. The SST anomalies of ML1 were controlled only by the sea surface heat flux. In ML2, variations in the vertical entrainment and the mixed layer depth were also considered. The effect of temperature transport by the geostrophic current was considered in ML3, and effects resulting from the addition of the Ekman current were included in ML4 (standard run). While the model was run using data starting in 1960, most of the following analyses were based on averages of years after the mid-1970s because both the model and observations showed strong decadal SST anomalies in the KOE region after the 1970s (Figs. 1, 2).
Figures 9a,b; 9c,d; and 9e,f show the decadal SST anomaly differences between experiments ML2 and ML1, ML3 and ML2, and ML4 and ML3, respectively, in October of the previous year (Figs. 9a,c,e) and in February (Figs. 9b,d,f). These differences in the simulated decadal SST anomalies correspond to the contributions of vertical entrainment (we), geostrophic advection (ug), and Ekman transport (ue), respectively. Note that the time scales of the contributions of each term to the decadal SST anomalies shown in Fig. 9 indicate not only seasonal modulation (see Fig. 7) but also decadal variation (see Fig. 5). The simulated decadal SST anomalies in the above months in ML4 are also shown in Fig. 10. The decadal SST anomalies in February for ML4 were very similar to those in October, despite being zonally stretched. Figures 9 and 10 were compared to examine whether each term in Eq. (1) enhanced or reduced the decadal SST anomalies in particular regions.
The SST anomaly north of the Kuroshio Extension was enhanced in winter and reduced in autumn by the Ekman current (Figs. 9e and 9f). This result is consistent with the fact that the Ekman temperature transport anomaly reaches a positive peak in early winter (Fig. 7b). In autumn, anomalous vertical entrainment formed the decadal SST anomaly (Fig. 9a), consistent with results shown in Fig. 7d. The SST anomaly reduced by the simultaneous mixed layer depth anomaly was several times smaller than that formed by the anomalous vertical entrainment (not shown). On the other hand, the enhancement of the decadal SST anomaly in winter was suppressed by the mixed layer depth anomaly (Fig. 9b), namely, the water mass caused by the autumnal vertical entrainment anomaly. The variation of depth was equivalent to that of the water mass and, hence, the heat capacity of the mixed layer. For example, in the case of a positive mixed layer depth anomaly (deep), the temperature tendency would be suppressed by the enhanced heat capacity of the mixed layer, even if the anomalous heat transport into the mixed layer was exactly zero. In the KOE region, in the cool SST phase, a deepened mixed layer acted to reduce heat release to the atmosphere at the sea surface and therefore reduced the cool temperature anomaly (Fig. 9b).
The SST anomalies in both seasons were enhanced because of temperature advection by the geostrophic current (Figs. 9c and 9d). This pattern was especially remarkable in winter when the enhancement was almost of the same magnitude as when influenced by the Ekman current in early winter or vertical entrainment in autumn. It should be noted that the decadal SST anomalies formed by the Ekman current in winter (Fig. 9f) and those caused by vertical entrainment in autumn (Fig. 9a) were enhanced within a few months by local atmospheric forcing (fast oceanic response), while those created by the geostrophic advection (Figs. 9c and 9d) required several years of basin-scale atmospheric forcing, namely, wind stress curl (slow oceanic response). Figures 9c and 9d showed the contribution of geostrophic temperature transport to the decadal SST anomaly on various time scales because the differences in SST between ML3 and ML2 were caused by differences in the governing equations, while Fig. 7c showed the contribution on seasonal time scales owing to the definition of the regression coefficient in Eq. (7). Thus, geostrophic advection contributed considerably to decadal-scale SST variation (Figs. 5, 9c–d), but only slightly to the seasonal-scale modulation of decadal SST variations (Fig. 7c). In other words, the time scales of decadal variation were controlled by geostrophic temperature transport, while seasonal modulation significantly depended on Ekman temperature transport and vertical entrainment.
As documented in appendix A, the net shortwave radiation flux at the sea surface had no deviation from the daily climatological values in the present study. An additional experiment using time series of net shortwave radiation flux taken from the NCEP–NCAR reanalysis data as one of the boundary conditions revealed that the net shortwave radiation flux contributed slightly to the seasonal heat budget only in summer when decadal SST anomalies decayed in the KOE region (not shown). This finding, however, did not change the major conclusions of the present study. When a positive decadal SST anomaly was found in the KOE region, for example, downward shortwave radiation flux was reduced at the sea surface. Changes in the rate of summer cloud cover associated with shortwave radiation flux variations may be induced by local forcing from SST anomalies or by alternation of the large-scale atmospheric circulation field or both. However, it must be noted that the shortwave radiation flux taken from the NCEP–NCAR reanalysis data was strongly dependent on the cloud field of that particular model.
6. Contribution of sea surface heat flux
The above lagged regression coefficients revealed no definite contribution from sea surface heat flux to the seasonal enhancement of the decadal SST anomaly, while sea surface heat flux acted to decay the decadal SST anomaly in summer and autumn (Fig. 7e). In contrast, a number of previous studies have suggested that sea surface heat flux is thought to enhance the decadal SST anomaly in the midlatitudes of the North Pacific (Blade 1997; Barsugli and Battisti 1998; Saravanan 1998). Focusing on the KOE region, however, several studies have recently pointed out that the sea surface heat flux is thought to damp the decadal SST anomaly (Barnett et al.1999; Seager et al. 2001; Schneider et al. 2002; Tanimoto et al. 2003).
As shown in Fig. 11, the local atmospheric thermal forcing anomaly Q′a(0) contributed more to the seasonal modulation of the decadal SST anomaly than did the sum of the vertical entrainment and Ekman heat transport effects, even in autumn and winter. This result suggested that local atmospheric thermal forcing Q′a(0) should be the most effective process acting to enhance decadal SST anomalies on seasonal time scales. On the other hand, the surface heat flux damping anomaly Q′o(0) also made a strong negative contribution to the seasonal heat budget (Fig. 11) through SST changes, mainly because of Q′a(0). While Q′a(0) was given as the boundary condition, the strength of Q′o(0) was also changed through ocean heat transport since the magnitude of Q′o(0) was almost proportional to that of the SST anomaly. That is, oceanic processes such as vertical entrainment and Ekman transport contributed to the seasonal enhancement of the decadal SST anomalies and also enhanced the surface heat flux damping anomaly Q′o(0). Thus, the surface heat flux anomaly made no contribution to the seasonal enhancement of the decadal SST anomalies since the reinforced local thermal damping rate resulting from the anomalous ocean heat transport canceled out the strong local atmospheric thermal forcing through surface heat flux.
Figure 12 shows the strength of the change of Q′o(0) caused by oceanic processes. The anomalous surface heat flux in ML4 acted to damp the decadal SST anomalies in the downstream KOE region (Fig. 12b), while that in ML1 enhanced the SST anomalies in the entire KOE region (Fig. 12a) as well as the upstream KOE region. Small, compared to Q′a(0), but significant differences in the heat flux were found in the KOE region (Fig. 12c). The spatial features of the anomalous surface heat flux in ML4 were consistent with those of the NCEP–NCAR reanalysis data (Fig. 12d), though slightly weaker in magnitude. The decadal anomaly in the downstream Kuroshio Extension region was reduced by heat flux at the sea surface (Fig. 12), while the heat flux anomaly enhanced the decadal SST anomaly in the Kuroshio region and the upstream Kuroshio Extension region. The difference in magnitude between the ML4 results and the NCEP–NCAR reanalysis data likely reflects the adjustment for upper-oceanic thermal conditions already present in the NCEP–NCAR reanalysis data used for the boundary condition. Tanimoto et al. (2003) indicated a similar damping effect on decadal SST anomalies caused by surface heat flux induced by oceanic processes. To focus on decadal time-scale air–sea coupling, they examined observational datasets from December to March, which was the mature season of decadal SST anomalies (see Fig. 6). The present study, in contrast, focused mainly on October–January (Figs. 9 –12) to examine seasonal modulation of the decadal SST anomalies. Detailed oceanic processes contributing to the damping components of surface heat flux were also clarified both in late autumn (vertical entrainment) and early winter (Ekman heat transport) in the present study (Fig. 9).
7. Concluding remarks
The present study clarified the physical process controlling the seasonal variation of decadal SST anomalies in the northwestern Pacific Ocean. A 3D bulk mixed layer model was integrated using atmospheric properties at the sea surface as boundary conditions. The decadal SST anomaly in the KOE region was strongest from December to February, while that in the central North Pacific was strongest from February to April. Seasonal-scale enhancement of the decadal SST anomaly in the KOE region was controlled by horizontal Ekman temperature transport in early winter and by vertical entrainment in autumn. Temperature transport by the geostrophic current contributed only slightly, despite controlling upper-ocean thermal conditions on decadal time scales through the slow Rossby wave adjustment to wind stress curl.
As documented in section 1, seasonality of the decadal SST anomalies is important to understanding the air–sea coupling process since the atmospheric response to the midlatitude SST anomaly depends considerably on the background atmospheric conditions. In the present paper, the contribution of sea surface heat flux was further examined using the anomalous local thermal damping rate Q′o and the anomalous local atmospheric forcing rate Q′a in both autumn and winter. Analyses revealed that local atmospheric thermal forcing through the sea surface heat flux anomaly Q′a was the strongest factor acting to enhance the decadal SST anomaly, as suggested by a number of previous studies. However, results showed that the sea surface heat flux anomaly made no contribution to seasonal enhancement of the decadal SST anomaly because of the strong local thermal damping anomaly Q′o, which resulted from strong oceanic heat transport anomalies. When focusing on the eastern KOE region, moreover, the sea surface heat flux anomaly dampened the decadal SST anomaly. These results imply a possible thermal forcing of the decadal SST anomaly in the KOE region to the local atmosphere in autumn, as well as in winter.
In the present paper, the model was integrated using atmospheric one-way forcing at the sea surface, and the input datasets used as boundary conditions were derived from different instrument systems and modeling, which may not be consistent. Further investigations are required using a general circulation model that couples the atmosphere and ocean to more faithfully and accurately examine atmospheric responses without employing uncharacteristic atmospheric thermal adjustments since the information of oceanic thermal conditions at the sea surface may have already been included in the atmospheric data from the NCEP–NCAR reanalysis data used as boundary conditions.
Acknowledgments
The authors wish to thank Professor A. J. Weaver and anonymous reviewers for their valuable comments and useful suggestions for the improvement of the manuscript. The authors also thank Dr. T. Satomura and Mr. N. Nishi for their thoughtful suggestions. This research was financially supported by Category 7 of the MEXT RR2002 Project for Sustainable Coexistence of Human, Nature and the Earth. All of the figures were produced by the GFD-DENNOU Library.
REFERENCES
Alexander, M. A., and C. Deser, 1995: A mechanism for the recurrence of wintertime midlatitude SST anomalies. J. Phys. Oceanogr., 25 , 122–137.
Alexander, M. A., C. Deser, and M. S. Timlin, 1999: The reemergence of SST anomalies in the North Pacific Ocean. J. Climate, 12 , 2419–2433.
Alexander, M. A., M. S. Timlin, and J. D. Scott, 2001: Winter-to-winter recurrence of sea surface temperature, salinity and mixed layer depth anomalies. Progress in Oceanography, Vol. 49, Pergamon, 41–61.
Barnett, T. P., D. W. Pierce, R. Saravanan, N. Schneider, D. Dommenget, and M. Latif, 1999: Origins of the midlatitude Pacific decadal variability. Geophys. Res. Lett., 26 , 1453–1456.
Barsugli, J. J., and D. S. Battisti, 1998: The basic effects of atmosphere–ocean thermal coupling on midlatitude variability. J. Atmos. Sci., 55 , 477–493.
Blade, I., 1997: The influence of midlatitude ocean–atmosphere coupling on the low-frequency variability of a GCM. Part I: No tropical SST forcing. J. Climate, 10 , 2087–2106.
Carton, J. A., G. Chepurin, X. Cao, and B. Giese, 2000: A simple ocean data assimilation analysis of the global upper ocean 1950–95. Part I: Methodology. J. Phys. Oceanogr., 30 , 294–309.
Deser, C., M. A. Alexander, and M. S. Timlin, 1996: Upper-ocean thermal variations in the North Pacific during 1970–1991. J. Climate, 9 , 1840–1855.
Deser, C., M. A. Alexander, and M. S. Timlin, 1999: Evidence for a wind-driven intensification of the Kuroshio Current Extension from the 1970s to the 1980s. J. Climate, 12 , 1697–1706.
Frankignoul, C., 1985: Sea surface temperature anomalies, planetary waves, and air–sea feedback in the middle latitudes. Rev. Geophys., 23 , 357–390.
Gill, A. E., 1982: Atmosphere–Ocean Dynamics. Academic Press, 662 pp.
Gu, D., and S. G. H. Philander, 1997: Interdecadal climate fluctuations that depend on exchanges between the tropics and extratropics. Science, 275 , 805–807.
Jacobs, G. A., H. E. Hurlburt, J. C. Kindle, E. J. Metzger, J. L. Mitchell, W. J. Teague, and A. J. Wallcraft, 1994: Decade-scale trans-Pacific propagation and warming effects of an El Ninõ anomaly. Nature, 370 , 360–363.
Japan Meteorological Agency, 1990: An objective analysis of 10-day mean sea surface temperature (in Japanese). Weather Service Bulletin, No. 57, Japan Meteorological Agency, Tokyo, Japan, 283–291.
Kawai, Y., H. Kawamura, S. Takahashi, K. Hosoda, H. Murakami, M. Kachi, and L. Guan, 2006: Satellite-based global daily 0.1°-grid optimum interpolation sea surface temperature data. J. Geophys. Res., in press.
Kim, J-W., 1976: A generalized bulk model of the oceanic mixed layer. J. Phys. Oceanogr., 6 , 686–695.
Kubota, M., N. Iwasaka, S. Kizu, M. Konda, and K. Kutsuwada, 2002: Japanese ocean flux data sets with use of remote sensing observations (J-OFURO). J. Oceanogr., 58 , 213–225.
Latif, M., and T. P. Barnett, 1994: Causes of decadal climate variability over the North Pacific and North America. Science, 266 , 634–637.
Latif, M., and T. P. Barnett, 1996: Decadal climate variability over the Northern Pacific and North America: Dynamics and predictability. J. Climate, 9 , 2407–2423.
Miller, A. J., and N. Schneider, 2000: Interdecadal climate regime dynamics in the North Pacific Ocean: Theories, observations and ecosystem impacts. Progress in Oceanography, Vol. 27, Pergamon, 257–260.
Miller, A. J., D. R. Cayan, and W. B. White, 1998: A westward-intensified decadal change in the North Pacific thermocline and gyre-scale circulation. J. Climate, 11 , 3112–3127.
Mochizuki, T., and H. Kida, 2003: Maintenance of decadal SST anomalies in the midlatitude North Pacific. J. Meteor. Soc. Japan, 81 , 477–491.
Nakamura, H., and T. Yamagata, 1999: Recent decadal SST variability in the northwestern Pacific and associated atmospheric anomalies. Beyond El Niño: Decadal and Interdecadal Climate Variability, A. Navarra, Ed., Springer Verlag, 49–72.
Nakamura, H., G. Lin, and T. Yamagata, 1997: Decadal climate variability in the North Pacific during the recent decades. Bull. Amer. Meteor. Soc., 78 , 2215–2225.
Niiler, P. P., and E. B. Kraus, 1977: One-dimensional models of the upper ocean. Modelling and Prediction of the Upper Layers of the Ocean, E. B. Kraus, Ed., Pergamon, 143–172.
Nonaka, M., and S-P. Xie, 2000: Propagation of North Pacific interdecadal subsurface temperature anomalies in an ocean GCM. Geophys. Res. Lett., 27 , 3747–3750.
Nonaka, M., S-P. Xie, and J. P. McCreary, 2002: Decadal variations in the subtropical cells and equatorial Pacific SST. Geophys. Res. Lett., 29 .1116, doi:10.1029/2001GL013717.
Paulson, C. A., and J. J. Simpson, 1977: Irradiative measurements in the upper ocean. J. Phys. Oceanogr., 7 , 952–956.
Peng, S., L. A. Mysak, H. Ritchie, J. Derome, and B. Dugas, 1995: The differences between early and midwinter atmospheric responses to sea surface temperature anomalies in the northwest Atlantic. J. Climate, 8 , 137–157.
Peng, S., W. A. Robinson, and M. P. Hoerling, 1997: The modeled atmospheric response to midlatitude SST anomalies and its dependence on background circulation states. J. Climate, 10 , 971–987.
Pierce, D. W., T. P. Barnett, N. Schneider, R. Saravanan, D. Dommenget, and M. Latif, 2001: The role of ocean dynamics in producing decadal climate variability in the North Pacific. Climate Dyn., 18 , 51–70.
Qiu, B., 1995: Variability and energetics of the Kuroshio Extension and its recirculation gyre from the first 2-year TOPEX data. J. Phys. Oceanogr., 25 , 1827–1843.
Qiu, B., 2000: Interannual variability of the Kuroshio Extension system and its impact on the wintertime SST field. J. Phys. Oceanogr., 30 , 1486–1502.
Qiu, B., and K. A. Kelly, 1993: Upper-ocean heat balance in the Kuroshio Extension region. J. Phys. Oceanogr., 23 , 2027–2041.
Qiu, B., and R. X. Huang, 1995: Ventilation of the North Atlantic and North Pacific: Subduction versus obduction. J. Phys. Oceanogr., 25 , 2374–2390.
Saravanan, R., 1998: Atmospheric low-frequency variability and its relationship to midlatitude SST variability: Studies using the NCAR climate system model. J. Climate, 11 , 1386–1404.
Schneider, N., and A. J. Miller, 2001: Predicting western North Pacific Ocean climate. J. Climate, 14 , 3997–4002.
Schneider, N., A. J. Miller, M. A. Alexander, and C. Deser, 1999: Subduction of decadal North Pacific temperature anomalies: Observations and dynamics. J. Phys. Oceanogr., 29 , 1056–1070.
Schneider, N., A. J. Miller, and D. W. Pierce, 2002: Anatomy of North Pacific decadal variability. J. Climate, 15 , 586–605.
Seager, R., Y. Kushnir, N. H. Naik, M. A. Cane, and A. J. Miller, 2001: Wind-driven shifts in the latitude of the Kuroshio–Oyashio Extension and generation of SST anomalies on decadal timescales. J. Climate, 14 , 4249–4265.
Tanimoto, Y., H. Nakamura, T. Kagimoto, and S. Yamane, 2003: An active role of extratropical sea surface temperature anomalies in determining anomalous turbulent heat flux. J. Geophys. Res., 108 .3304, doi:10.1029/2002JC001750.
Venzke, S., M. Munnich, and M. Latif, 2000: On the predictability of decadal changes in the North Pacific. Climate Dyn., 16 , 379–392.
Vivier, F., K. A. Kelly, and L. Thompson, 2002: Heat budget in the Kuroshio Extension region: 1993–99. J. Phys. Oceanogr., 32 , 3436–3454.
Watanabe, M., and M. Kimoto, 2000: Behavior of midlatitude decadal oscillations in a simple atmosphere–ocean system. J. Meteor. Soc. Japan, 78 , 441–460.
Weaver, A. J., 1999: Extratropical subduction and decadal modulation of El Ninõ. Geophys. Res. Lett., 26 , 743–746.
White, W. B., 1995: Design of a global observing system for gyre scale upper ocean temperature variability. Progress in Oceanography, Vol. 36, Pergamon, 169–217.
Xie, S-P., T. Kunitani, A. Kubokawa, M. Nonaka, and S. Hosoda, 2000: Inter-decadal thermocline variability in the North Pacific for 1958–1997: A GCM simulation. J. Phys. Oceanogr., 30 , 2798–2813.
Yasuda, I., T. Tozuka, M. Noto, and S. Kouketsu, 2000: Heat balance and regime shifts of the mixed layer in the Kuroshio Extension. Progress in Oceanography, Vol. 47, Pergamon, 257–278.
Yasuda, T., and K. Hanawa, 1997: Decadal changes in the mode waters in the midlatitude North Pacific. J. Phys. Oceanogr., 27 , 858–870.
Yasuda, T., and Y. Kitamura, 2003: Long-term variability of North Pacific subtropical mode water in response to spin-up of the subtropical gyre. J. Oceanogr., 59 , 279–290.
Zhang, R-H., and S. Levitus, 1997: Structure and cycle of decadal variability of upper-ocean temperature in the North Pacific. J. Climate, 10 , 710–727.
Zhang, R-H., and Z. Liu, 1999: Decadal thermocline variability in the North Pacific Ocean: Two pathways around the subtropical gyre. J. Climate, 12 , 3273–3296.
APPENDIX A
The 3D Bulk Mixed Layer Model
The linearized governing equation for the mixed layer temperature anomaly T ′ was written as Eq. (2). Daily climatological values of SST,
As documented in section 2, the anomalous geostrophic current velocity u′g was obtained from the first baroclinic Rossby wave adjustment [see Eq. (4)]. The climatological geostrophic current velocity 
The term [τ′ × k · ∇
The linearized governing equation for H′m was written as Eq. (3). The product of the mixed layer depth Hm and the Ekman velocity divergence ∇ · uE was equal to the divergence of the Ekman transport flux since the Ekman velocity was defined as the Ekman volume transport divided by the mixed layer depth in the model [see Eq. (2)]. The climatological mixed layer depth
If the vertical process was entrainment (we =
APPENDIX B
Ignored Components of Geostrophic Temperature Transport
To explore some concerns resulting from the representation of the anomalous geostrophic velocity in the model, two additional control runs were conducted (ML5 and ML6) using different methods of estimating the horizontal temperature transport terms.
Another concern was underestimation of shorter time-scale deviations of the geostrophic velocity arising from the calculation of the anomalous geostrophic velocity documented in section 2a. In the standard run, ML4, the anomalous geostrophic velocity was calculated as the first baroclinic Rossby mode, and hence the variations of the geostrophic velocity on shorter time scales were ignored. In the ML6 experiment, the effects of higher frequency variations of the geostrophic velocity were considered using the anomalous surface velocity taken from the SODA data as the anomalous geostrophic velocity. The lagged regression coefficients of the geostrophic temperature transport of experiment ML6 are shown as a dashed line in Fig. B2. In autumn and winter, when the strong decadal SST anomalies appeared, the effect of geostrophic heat transport in ML6 was greater than that in ML4, but still contributed very little to the seasonal decay of the decadal SST anomalies. It should be noted that strong positive peaks were found in winter in ML6, while those of the Ekman heat transport and the vertical entrainment in ML4 were found in autumn (Figs. 7b and 7d). These results imply that geostrophic heat transport acted to maintain, rather than enhance, the decadal SST anomalies on seasonal time scales. However, note that even in winter, the geostrophic heat transport was not more effective than the Ekman heat transport and the vertical entrainment.
Horizontal distributions of model-simulated wintertime (January–March) SST anomaly differences for (a) 1969–73 minus 1975–79, (b) 1972–76 minus 1978–82, (c) 1978–82 minus 1984–88, and (d) 1984–88 minus 1990–94. Five-year running means were applied to each month. Contour intervals are 0.2 K. The rectangles in (b) and (d), respectively, represent the STF (25°–35°N, 175°E–145°W) and the KOE region (35°–45°N, 150°E–180°) as defined in the present study.
Citation: Journal of Climate 19, 12; 10.1175/JCLI3807.1
As in Fig. 1 except that SST data were acquired from a JMA observation dataset.
Citation: Journal of Climate 19, 12; 10.1175/JCLI3807.1
Temporal distribution of winter (January–March) mean SST anomalies in the KOE region (left ordinate). The solid and broken lines represent the JMA observed data and the SST simulated by the model, respectively. Time series of the mixed layer depth anomalies (dashed line, right ordinate): positive (negative) values correspond to deeper (shallower) than normal mixed layers.
Citation: Journal of Climate 19, 12; 10.1175/JCLI3807.1
Magnitudes of annual-averaged geostrophic velocity, |
Citation: Journal of Climate 19, 12; 10.1175/JCLI3807.1
Time series of the wintertime (January–March) decadal SST anomaly, T ′ (solid line, left ordinate), and the monthly heat transport rate anomaly caused by anomalous geostrophic advection, u′g · ∇
Citation: Journal of Climate 19, 12; 10.1175/JCLI3807.1
Lagged regression coefficients of the decadal SST anomaly time series for each month relative to the annual-averaged decadal SST anomaly time series, using data (a) simulated by the model and (b) from the JMA, for the KOE (solid lines) and STF region (dashed lines). The vertical bars represent significant ranges at a 95% confidence level.
Citation: Journal of Climate 19, 12; 10.1175/JCLI3807.1
Lagged regression coefficients of the decadal time series for each term in Eq. (1) for each month relative to the annual averaged decadal SST anomaly time series in the KOE region: term (a) tendency [Hm∂T/∂t]′, (b) Ekman temperature transport −[HmuE · ∇T]′, (c) geostrophic temperature transport −[Hmug · ∇T]′, (d) vertical entrainment −[weΔT]′, and (e) sea surface heat flux −Q′(0)/coρpo. Values of each term plotted have been divided by the climatological mixed layer depth
Citation: Journal of Climate 19, 12; 10.1175/JCLI3807.1
Average of the difference in SST anomalies for 1978–82 minus 1984–88 (see Fig. 1c) and 1990–94 minus 1984–88 (see Fig. 1d) between the control runs (a), (b) ML2–ML1; (c), (d) ML3–ML2; and (e), (f) ML4–ML3 in (a), (c), (e) October of the previous year and (b), (d), (f) February. Here, (a), (b) We, (c), (d) Ug, and (e), (f) UE represent the entrainment and mixed layer depth anomaly effects, the geostrophic current effect, and the Ekman transport effect, respectively. Contour intervals are 0.1 K. Solid lines represent positive and dashed represent negative values. Shaded regions denote areas where the composited values are significant at a 95% confidence level.
Citation: Journal of Climate 19, 12; 10.1175/JCLI3807.1
Average of the SST anomalies for 1978–82 minus those in 1984–88 and those in 1990–94 minus those in 1984–88 in (a) October of the previous year and (b) February of ML4. Contour intervals are 0.2 K. Solid (dashed) lines represent positive (negative) values. Shaded regions denote areas where the composited values are significant at a 95% confidence level.
Citation: Journal of Climate 19, 12; 10.1175/JCLI3807.1
As in Fig. 7e except for the atmospheric thermal forcing anomalies Q′a(0) (solid line), the sea surface heat flux damping anomalies Q′o(0) (dashed line), and the net sea surface heat flux anomalies Q′(0) [=Q′a(0) + Q′o(0); broken line], respectively. The dashed–dotted line represents the sum of the effects of the horizontal Ekman transport (Fig. 7b) and the vertical entrainment (Fig. 7d). Plotted values of each term have been divided by the climatological mixed layer depth
Citation: Journal of Climate 19, 12; 10.1175/JCLI3807.1
Average of the sea surface heat flux anomalies in 1978–82 minus the anomalies in 1984–88 and those in 1990–94 minus 1984–88, averaged over October–January for (a) ML1, (b) ML4, (c) the differences between ML4 and ML1, and (d) the NCEP–NCAR reanalysis data. Contour intervals are 10 W m−2. Solid (dashed) lines represent positive (negative) values. Shaded regions denote areas where the composited values are significant at a 95% confidence level.
Citation: Journal of Climate 19, 12; 10.1175/JCLI3807.1
Fig. B1. Relationship between the values of the geostrophic temperature transport term ug · ∇T calculated for aa horizontal resolution of 1° × 1° and for 0.25° × 0.25° at each grid point along 35°N, 150°E–180°. The geostrophic velocity and the SST are 10-day averaged values from January 1995 to December 1999 from TOPEX/Poseidon altimetry data and Merged SST data.
Citation: Journal of Climate 19, 12; 10.1175/JCLI3807.1
Fig. B2. Lagged regression coefficients of the decadal time series of the geostrophic temperature transport term −[Hmug · ∇T]′ in Eq. (1) for each month, relative to the annual averaged decadal SST anomaly time series in the KOE region. Plotted values have been divided by the climatological mixed layer depth
Citation: Journal of Climate 19, 12; 10.1175/JCLI3807.1
Physical constants adopted in the mixed layer model.
List of experiments with various water temperature equations.
