Interannual variations in low potential vorticity water and the subtropical countercurrent in an eddy-resolving OGCM
- First Online:
- Cite this article as:
- Nonaka, M., Xie, S. & Sasaki, H. J Oceanogr (2012) 68: 139. doi:10.1007/s10872-011-0042-3
Interannual-to-decadal variations in the subtropical countercurrent (STCC) and low potential vorticity (PV) water and their relations in the North Pacific Ocean are investigated on the basis of a 60-year-long hindcast integration of an eddy-resolving ocean general circulation model. Although vertically coherent variations are dominant for STCC interannual variability, a correlation analysis shows that an intensified STCC vertical shear accompanies lower PV than usual to the north on 25.5- to 26.1-σθ isopycnal surfaces, and intensified meridional density gradient in subsurface layers, consistent with Kubokawa’s theory (J Phys Oceanogr 29:1314–1333, 1999). The low-PV signals appear at least 2 years before peaks of STCC, propagating southwestward from the subduction region.
KeywordsSubtropical countercurrentLow potential vorticity waterNorth PacificInterannual variabilityEddy-resolving ocean general circulation model
In the late 1960s, a weak eastward current embedded in the broad westward current in the southern part of the North Pacific subtropical gyre was discovered (Uda and Hasunuma 1969). As its direction is opposite to that expected from the gyre circulation, it is named the subtropical countercurrent (STCC). Although its whole distribution had not been clarified for a long time because of high eddy activity in the region (Qiu 1999), the accumulation of observational data recently allowed its detailed distributions to be described (Kobashi and Kawamura 2001, 2002; Kobashi et al. 2006). Although STCC is not a strong current, the associated temperature front, the subtropical front (STF), can influence atmospheric fields aloft (Kobashi et al. 2008; Kobashi and Xie 2011). Coupled model results suggest positive coupled feedback associated with the atmospheric response (Xie et al. 2011). In coupled model projections, changes in STCC under global warming can significantly affect distributions of sea surface temperature (SST) changes in the North Pacific (Xie et al. 2010).
Formation mechanisms for STCC have been debated since its discovery. Yoshida and Kidokoro (1967) and Roden (1975) discuss the possible importance of detailed distributions of zonal wind stress, and meridional Ekman convergence, respectively. On the basis of ocean general circulation model (OGCM) experiments, Takeuchi (1984), however, shows that STCC can be formed without narrow scale structures in zonal wind and meridional Ekman convergence. In a theoretical study, Kubokawa (1999) proposes that STCC is associated with meridional density gradient caused by the accumulation of low potential vorticity (PV), thick water layers subducted to the north. A numerical simulation supports this hypothesis (Kubokawa and Inui 1999). Specifically, the mode water pushes the upper pycnocline upward. On the southern flank of the mode water, this creates a northward shoaling of the upper pycnocline, which sustains an eastward current shear by thermal wind. Observational data also suggest that relative distributions between STCC and low-PV water are consistent with Kubokawa’s (1999) hypothesis (Aoki et al. 2002). Furthermore, from detailed investigation of historical observed data, Kobashi et al. (2006) show that the northern and eastern branches of STCC locate at the southern edge of the low-PV regions corresponding to the North Pacific Subtropical Mode Water (NPSTMW, Masuzawa 1969; Hanawa and Talley 2001), and to the North Pacific Central Mode Water (NPCMW, Nakamura 1996; Suga et al. 1997; Oka and Suga 2005), respectively.
Although these observed climatological fields strongly support Kubokawa’s (1999) hypothesis, it is still not clear if interannual-to-decadal variations in STCC are caused by variations of low-PV waters. On decadal timescales, in a 300-year simulation of a coupled model, Xie et al. (2011) show that STCC variations are caused by changes in low-PV water subduction. Although the resolution of the coupled model is limited and low-PV water tends to be exaggerated in the model, on the basis on an eddy-resolving OGCM simulation Yamanaka et al. (2008) show that decadal differences in STCC are associated with those in low-PV waters, consistent with Kubokawa’s hypothesis. On interannual timescales, Qiu and Chen (2010) and Kobashi and Xie (2011) indicate the importance of Ekman convergence for STF and STCC, but the possibility of contribution of low-PV water variations has not been explored. The present study investigates if variations in low-PV water subduction can affect STCC on interannual timescales, based on an eddy-resolving OGCM simulation that is different from Yamanaka et al.’s.
In Sect. 2, we introduce the model and describe simulated low-PV waters. Simulated STCC is described in Sect. 3, and the relationship between interannual-to-decadal variations in STCC and low-PV waters is investigated in Sect. 4. Section 5 gives a summary and discussion.
We used the Modular Ocean Model 3 (MOM3) OGCM (Pacanowski and Griffies, 2000) with substantial modification for the vector-parallel hardware system of Japan’s Earth Simulator. Our ocean model for the Earth Simulator (OFES; Masumoto et al. 2004) covers a near-global domain of 75°N–75°S, with a horizontal resolution of 0.1°. The model has 54 vertical levels with resolutions of 5 m at the surface, and the maximum depth is 6065 m. For horizontal mixing of momentum and tracers, we adopted scale-selective damping with a bi-harmonic operator (Smith et al. 2000). The nonlocal K-profile parameterization (KPP) boundary layer mixing scheme (Large et al. 1994) was adopted for the vertical mixing.
The surface heat flux and evaporation were calculated by the bulk formula with atmospheric variables based on National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) reanalysis data (Kalnay et al. 1996) and the simulated SST field. The freshwater flux was evaluated from the evaporation field and daily precipitation rate data, under the constraint that sea surface salinity (SSS) is restored to the observed monthly climatology at a timescale of 6 days. Wind stress is also obtained from the NCEP/NCAR reanalysis. For details, see Masumoto et al. (2004) and Sasaki et al. (2008).
Following a 50-year integration with climatological monthly-mean forcing from the annual-mean temperature and salinity climatological fields without motion, we conducted a 60-year hindcast integration with daily mean atmospheric fields of the NCEP/NCAR reanalysis data from 1950 to 2009. This hindcast simulation successfully captures variability with intraseasonal-to-decadal timescales (Sasaki et al. 2008) and has been used to investigate interannual-to-decadal variability in the western North Pacific region (Nonaka et al. 2006, 2008; Taguchi et al. 2007, 2010). In the following analyses, the resolution of our model output is reduced to 0.5° for analytical convenience by selecting the data at every five grid points both in the zonal and meridional directions.
2.2 Simulated low-PV water distributions
3 Simulated STCC
3.1 Mean STCC
3.2 Relation between low-PV water and STCC in long-term mean and interdecadal variations
4 Interannual variations in STCC and low-PV water
4.1 Interannual variations in simulated STCC
These vertically coherent interannual variations are not related to variations in low-PV water. As the purpose of the present paper is to investigate the possible influence of low-PV variations on STCC, in the following analyses, we focus on interannual variations that deviate from the vertically uniform structure. To extract upper layer variations independent of deeper layer variations, we compute the linear regression of the time series of 54-m zonal velocity upon velocity at 404-m depth, and then subtract the regression from the original 54-m zonal velocity time series (Fig. 7b). In the following analyses, we use this new time series averaged at (26°–28°N, 160°–180°E) (Fig. 7b) to represent interannual variations in STCC that is trapped to the near-surface layer and independent of the coherent variations in the upper 1000 m. Selection of the depth for the regression, 404 m, is arbitrary, but the resultant time series does not strongly depend on this choice of depth; almost the same results are obtained if regression upon time series at a different depth, say 604 m, is removed. This STCC shear index is a measure of current shear and a good representation of the surface-trapped current in climatology.
4.2 Correlation between STCC and low-PV water
4.3 Causes of thickness variations
5 Summary and discussion
On the basis of a 60-year-long hindcast integration of an eddy-resolving OGCM, OFES, we have investigated interannual-to-decadal variations in STCC and low-PV water and their relations in the North Pacific Ocean. In the model climatology, STCC is found on the south edge of low-PV water as predicted by Kubokawa’s (1999) theory and consistent with observations. This close relationship holds for interdecadal variations in STCC and low-PV water in the model. The specific purpose of the present study is to investigate if the relationship between STCC and low-PV water also holds in interannual variations.
In the OFES hindcast, STCC appears from 150°E to around the international dateline, and is trapped to near the surface. As it peaks in summer near the dateline, we examine interannual variations in summertime STCC and low-PV water. The STCC variability on interannual timescales is dominated by a vertically deep structure (>1000 m) that seems to relate to narrow bands of zonal currents or striations. We then define an STCC shear by removing vertically coherent variations from variations in the near-surface layer zonal velocity, to represent the surface-trapped structure of STCC.
A correlation analysis shows that intensified STCC is accompanied by lower PV than usual to the north on 25.5- to 26.1-σθ isopycnal surfaces, and by intensified meridional density gradient in subsurface layers (Fig. 10). The low-PV signals appear at least 2 years before peaks of STCC, develop and shift southwestward (Fig. 11). These low-PV signals can be traced back to subduction processes. The results of the correlation analysis are further confirmed by latitude-time sections (Fig. 12).
These results strongly suggest that interannual-to-decadal variations in STCC are associated with variations of low-PV water ventilation. In other words, the relation between STCC and low-PV water suggested by Kubokawa (1999) and found in the observed climatology (Kobashi et al. 2006) holds in interannual-to-decadal variations at least in this particular model. This mechanism is at work for decadal changes in STCC (Yamanaka et al. 2008), and our study demonstrates that it works also on interannual timescales in addition to the influence of Ekman flow variations shown by Qiu and Chen (2010) and Kobashi and Xie (2011). Sasaki et al. (2011) suggest a similar influence of low-PV water ventilation on interannual variations of HLCC. The influence of Ekman flow on STCC is, however, not strong in this model. This may be due to the difference in the region of analyses: the studies of Qiu and Chen (2010) and Kobashi and Xie (2011) focus upon the western northern STCC, whereas the eastern STCC is investigated in this study. Also, the weaker meridional temperature gradient in this model than in the observations can weaken the impact of the Ekman flow.2
Low-PV water distributions in OFES are slightly different from observations as discussed in Sect. 2.2. Indeed, the low-PV waters in the layers around 25.6 σθ that correlate with STCC tend to develop and extend more eastward than in observations. This suggests that the relationships between variations in the low-PV water and STCC found in the model may be stronger than in the real ocean. Recently, subsurface observations by Argo profilers (Argo Science Team 2001; Hosoda et al. 2010) have accumulated rapidly, and it becomes possible to investigate subsurface interannual variations. Indeed, on the basis of Argo observations, Oka (2009) shows that interannual temperature anomalies in NPSTMW can be advected southward by at least 2° latitude with a half-year lag. The low-PV water related to STCC variability in the present study has higher density and subducts in higher latitudes than NPSTMW. Also, as shown in Fig. 12, thickness variability extends southward from the source region (32°–34°N) to STCC (26°–28°N) with a few years lag. Further accumulation of Argo observations will make it possible in the near future to investigate the propagation of PV anomalies in relation to STCC variability in the real ocean.
Around 160°–140°W, positive PV correlations are found to the north of positive zonal velocity anomalies. In contrast to that found in Fig. 10a, these are not accompanied by negative PV anomalies below. The positive PV anomalies are probably due to advection by westward current anomalies found there (not shown) on a background of strong zonal PV gradients.
Indeed, simulated STF in early spring also show weak correlation (r ~ 0.4 at the maximum) between the Ekman convergence and meridional temperature gradient to about 100-m depth. This is consistent with the result of Qiu and Chen (2010) but the correlation is weaker and spatially incoherent (not shown).
The OFES simulations were conducted on the Earth Simulator under the support of JAMSTEC. We thank the members of the OFES group, including Drs. H. Sakuma, Y. Masumoto, and T. Yamagata, for their efforts and support in the model development. Our thanks are extended to Drs. A. Kubokawa, F. Kobashi, and E. Oka for useful discussions. This study is partially supported by Grand-In-Aid for Scientific Research defrayed by the Ministry of Education, Culture, Sports, Science and Technology of Japan (22106006, 23340139). IPRC/SOEST publication #790/8196.