1. Introduction
Observational studies have shown that decadal-scale sea surface temperature (SST) variations in the North Pacific are particularly strong along the oceanic subarctic frontal zone (SAFZ; e.g., Nakamura et al. 1997; Nakamura and Kazmin 2003). The SST variations along the SAFZ are primarily a result of its latitudinal displacement (Nakamura and Kazmin 2003), which has been shown to be a manifestation of oceanic dynamical adjustments to basin-scale wind changes in modeling studies (e.g., Seager et al. 2001; Nonaka et al. 2006, 2008). A question then arises whether and how the atmospheric circulation can respond to the persistent SST anomalies induced by oceanic frontal variability. The particular question is relevant to understanding of mechanisms for decadal-scale climate variability in the Pacific, including the Pacific decadal oscillation (PDO; Mantua et al. 1997) and North Pacific Gyre Oscillation (NPGO; Di Lorenzo et al. 2008).
Evidence has been emerging that SST gradients associated with midlatitude oceanic frontal zones, including the North Pacific SAFZ, can influence the time-mean state of the basin-scale atmospheric circulation beyond local responses that is confined to the planetary boundary layer (PBL). Nakamura et al. (2004) hypothesized that differential air–sea heat exchanges across an oceanic (SST) frontal zone maintain near-surface atmospheric baroclinicity, which leads to the organization of a large-scale storm track aloft. The hypothesis is supported by a series of studies using high-resolution numerical models of various complexity: an atmospheric general circulation model (AGCM) with an aquaplanet-type setting (Nakamura et al. 2008; Sampe et al. 2010; cf. Brayshaw et al. 2008), a coupled ocean–atmosphere GCM (CGCM; Nonaka et al. 2009) for the subpolar front in the southern Indian Ocean, a regional atmospheric model for the Kuroshio–Oyashio Extension region (KOE; Taguchi et al. 2009), and a linear planetary wave model (Hotta and Nakamura 2011). Furthermore, it is found in high-resolution operational analysis data (Minobe et al. 2008, 2010) and in a high-resolution AGCM simulation (Kuwano-Yoshida et al. 2010b) that enhanced evaporation along the Gulf Stream axis induces surface wind convergence, leading to the organization and anchoring of a narrow band of deep convections, especially in summer. The organization of convective clouds is found also along the Kuroshio Extension (KE; Tokinaga et al. 2009). It remains to be fully understood, however, whether the oceanic frontal variability as observed in the Gulf Stream and KE regions can affect the time-varying atmospheric state in any significant manner and, if so, how.
A number of studies have investigated large-scale atmospheric circulation responses to extratropical SST anomalies using statistical analysis of observational data and/or AGCM experiments with prescribed SST anomalies, or coupled with mixed layer or dynamical ocean models with anomalous subsurface temperature/heat transport convergence (e.g., Kushnir et al. 2002; Kwon et al. 2010, 2011 for reviews). However, both oceanic indices used in the previous observational studies and oceanic anomalies imposed on the AGCM studies can only represent large-scale features, and neither of them can fully resolve meridionally confined frontal or jet structures.
The objective of this study is to examine influences of low-frequency SST variability confined meridionally to the SAFZ on the seasonally varying atmospheric state over the North Pacific through our analyses of observational data and a 120-yr integration of a medium-resolution CGCM with its “front resolving” ocean component. Advantages of analyzing the CGCM integration in addition to the observations are threefold. First, the air–sea coupling allows two-way forcing between the atmosphere and ocean, and anomalies in SST and surface heat fluxes are thus mutually consistent within the particular model. Atmospheric responses, if any, to extratropical SST anomalies in CGCM experiments would therefore be more likely realistic than those in AGCM simulations with prescribed SSTs (Liu and Wu 2004). Second, the CGCM with relatively high resolution can resolve some, if not all, crucial processes that occur near oceanic frontal zones to exert thermal forcing on the atmosphere, including the differential air–sea heat/moisture exchanges across the oceanic front and enhanced latent heat release on the warmer flank of the front as mentioned above. Third, the longer integration period for our CGCM than the observational record can augment statistical significance of the signature of the atmospheric responses to SST anomalies, if extracted, as long as their evolution and mechanisms are found consistent between the observations and model.
Our analysis focuses on the cold season from October to March, in which decadal-scale signals in the North Pacific SST are prominent. SST anomalies can be rather persistent as a result of the reemergence mechanisms of subsurface temperature anomalies (Alexander and Deser 1995) and/or because of axial displacement of the SAFZ associated with gyre adjustment to basin-scale anomalous wind forcing (Seager et al. 2001; Nonaka et al. 2006, 2008). Compared to the persistency of the SST anomalies, a typical time scale with which the atmosphere responds to them should be much shorter. For example, an AGCM study has suggested that it takes only 2–2.5 months for the extratropical atmosphere over the North Atlantic to reach a large-scale, equivalent barotropic equilibrium state from the initial stage of local, baroclinic response to thermal forcing from the underlying ocean (Deser et al. 2007). Another AGCM study has suggested that atmospheric responses to extratropical SST anomalies can be sensitive to subtle differences in the climatological mean state of the atmosphere from one calendar month to another (Peng et al. 1997; Peng and Whitaker 1999). Given the short time scale of the atmospheric response and its sensitivity to the background climatological state, the present study investigates atmospheric anomalies on a monthly basis, unlike in most of the previous studies that focused on seasonal anomalies. Our analyses of observed data and a long-term CGCM integration reveal a distinct (sub)seasonality in anomalous atmospheric circulation accompanied by pronounced SST anomalies in the SAFZ. This particular (sub)seasonality may be a manifestation of (sub)seasonal dependence found in decadal modulations of storm-track activity and their feedback on the background flow.
The rest of this paper is organized as follows. Section 2 describes observational data and a CGCM used in this study. In section 3, we define an index that represents observed low-frequency SST anomalies in the North Pacific SAFZ induced primarily by oceanic frontal variability, to extract the associated anomalies in SST and surface heat fluxes. In section 4 we show the distinct seasonality of atmospheric anomalies associated with the SST anomalies in the SAFZ, which is compared with its counterpart simulated in the CGCM integration in section 5. Section 6 presents the conclusions and a discussion.
2. Observational data and a CGCM integration
a. Observational data
To extract decadal variability of variables at the ocean surface, we use monthly mean SST and turbulent (sensible and latent) heat fluxes derived from the International Comprehensive Ocean–Atmosphere Data Set (ICOADS) release 2.4, in which surface marine meteorological measurements from ships, buoys, and other marine platforms are complied. The data are originally mapped onto a 2° × 2° grid from the International Maritime Meteorological Archive of ICOADS for the 48-yr period from January 1959 to December 2006. Since the heat flux analysis requires not only SST but also meteorological variables that are sometimes not available in ship observations, data used for constructing an analysis value at a particular grid point in the SAFZ are 10%–20% fewer for the heat fluxes than for SST. The reader is referred to Tokinaga et al. (2009) for detailed gridding procedures. The ICOADS dataset is chosen rather than satellite-measured SST fields despite their much higher spatial resolution, because the longer analysis period of the former is suited for our purpose to extract decadal SST variations with statistical significance as high as possible. Its resolution (2° × 2°) enables midlatitude oceanic frontal zones including the North Pacific SAFZ to be resolved, while maximizing sampling for individual grid points. We also use a monthly SST dataset from the Hadley Centre Sea Ice and SST dataset, version 1.1 (HadISST) provided by the Met Office (Rayner et al. 2003), to examine SST anomalies in the tropics, where the number of data included in the ICOADS dataset is limited.
Atmospheric variables analyzed in the present study include monthly mean datasets of sea level pressure (SLP) and 250-hPa geopotential height (hereafter Z250), obtained from the National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis (Kalnay et al. 1996) for the period 1959–2006. Evaluation of synoptic-scale transient eddy vorticity and heat fluxes is based on daily-mean fields of wind and temperature.
b. CGCM
The present study utilizes a 120-yr CGCM integration, which is much longer than the observational data as mentioned above and thus augments statistical significance of our results. Despite uncertainties in physical parameterizations of subgrid processes, including cloud physics and PBL processes, the CGCM can nevertheless provide data of physically consistent ocean–atmosphere coupled variability. Allowing more complete sampling, especially for SST and surface heat fluxes than those based on shipboard observations, the CGCM output is thus complementary to the aforementioned observational data.
The model used in this study is the CGCM for the Earth Simulator (CFES), which consists of the AGCM for the ES (AFES; Ohfuchi et al. 2004, 2007; Enomoto et al. 2008; Kuwano-Yoshida et al. 2010a) and the Coupled Ocean–Sea Ice Model for the ES (OIFES; Komori et al. 2005). AFES is based on the Center for Climate System Research/National Institute for Environmental Studies (CCSR/NIES) AGCM version 5.4.02 (Numaguti et al. 1997), while OIFES is based on the Modular Ocean Model version 3 (MOM3; Pacanowski and Griffies 2000). Computational codes of AFES and OIFES have been substantially rewritten from their prototypes, in order to attain their high computational efficiency on the particular architecture of ES and to implement improved parameterizations of physical processes. Further details and applications of CFES are given by Komori et al. (2008a,b), Nonaka et al. (2009), and Richter et al. (2010).
To obtain a dataset that is sufficiently long for the study of the decadal variability, CFES has been integrated for 120 yr with medium horizontal resolution (CFES-M): T119 spectral truncation (equivalently 125-km grid intervals) with 48σ levels for its atmospheric component (AFES), and 0.5° grid intervals with 54 vertical levels for its oceanic component (OIFES). With these resolutions, synoptic-scale disturbances in the atmosphere are well represented, but mesoscale eddies in the ocean cannot be resolved. To parameterize eddy effects on diffusive processes in the ocean, isopycnal mixing (Redi 1982) and parameterized eddy-induced motions (Gent and McWilliams 1990) are incorporated in OIFES. Though unable to resolve oceanic eddies, CFES-M can nevertheless simulate a prominent oceanic frontal zone that corresponds to the SAFZ in the KOE region in the real ocean (Fig. 1b, contours), as discussed in section 5.
Wintertime (January–February–March) mean (contoured) and standard deviation (gray shaded) of SST based on (a) ICOADS data for the period 1959–2006 and (b) CFES-M integration for 120 yr. Areas used for defining indices for the observed and simulated SST anomalies along the SAFZ are indicated with black rectangular boxes.
Citation: Journal of Climate 25, 1; 10.1175/JCLI-D-11-00046.1
3. Observed low-frequency anomalies at the ocean surface in the North Pacific subarctic frontal zone
a. SAFZ-SST index
To extract SST anomalies that tend to be pronounced in the North Pacific SAFZ, we define an index (hereafter referred to as the “SAFZ-SST index”; thick black lines in Fig. 2a) as a monthly anomaly of the ICOADS SST averaged over a rectangular domain within the climatological-mean SAFZ (37.5°–42.5°N, 147.5°–160.5°E; black rectangular box in Fig. 1a), where the meridional gradient and interannual variance of SST are both particularly strong. The domain for defining the index does not include the coastal region west of 147.5°E, where the SST variance, though largest, is presumably induced by variations of the Oyashio coastal intrusions that are not necessarily synchronous with the SAFZ variability (e.g., Yasuda 2003). The index has been constructed after removing the linear trend and monthly climatology. During the cold season, low-frequency variability of the SAFZ-SST index is not necessarily coherent from one calendar month to another, but rather it displays a distinct (sub)seasonal dependence. Specifically, year-to-year (high frequency) variability is dominant in fall (October and November), while (quasi) decadal (low frequency) variability becomes more dominant in midwinter through early spring (January–March). This seasonality may be attributable to the seasonal deepening of the oceanic mixed layer that results in its increasing thermal inertia and tighter linkage with dynamically induced thermocline variability (e.g., Xie et al. 2000; Kwon et al. 2010).
(a) Time series of SST anomalies averaged over the SAFZ (thick black line; referred to as SAFZ-SST index) and the frontal axis (thin gray line) defined as the latitude of maximum meridional SST gradient, based on the ICOADS data for the period 1959–2005 (October–December) or 1960–2006 (January–March). The time series are constructed separately for the individual calendar months from which the linear trend and climatology had been removed before the 3-yr running mean was applied. Correlation between the area-averaged SST and the frontal position, and the climatological-mean frontal latitude are indicated at the bottom and on the right-hand side on the ordinate, respectively. (b) As in (a), but for SST based on the 120-yr CFES-M integration. Note that the range for the SST anomalies (ordinate on the left) is enhanced by 50% than in (a).
Citation: Journal of Climate 25, 1; 10.1175/JCLI-D-11-00046.1
As evident in Fig. 2a, low-frequency variations in the SAFZ-SST index, extracted through the 3-yr running mean, tend to be synchronized with the corresponding variations in the SAFZ axis defined as the latitude of maximum meridional SST gradient (thin gray lines in Fig. 2a). In fact, the correlation coefficient between the two time series exceeds 0.65 for all the calendar months of interest and it is even higher than 0.7 for January through March. The axial variations of the SAFZ can be attributed to oceanic adjustments to (semi) basin-scale wind forcing via barotropic and baroclinic Rossby waves as discussed by Nonaka et al. (2008) and Frankignoul et al. (2011). Consistent with their arguments, the atmospheric forcing for both positive SST anomalies in the SAFZ (lower-left panel of Fig. 5; see also section 4a) and its poleward displacement (not shown) corresponds to the weakened Aleutian low in December and January. The SAFZ-SST index thus represents SST anomalies induced primarily by oceanic changes, particularly the meridional displacement of the SAFZ (Seager et al. 2001; Nonaka et al. 2006, 2008; Frankignoul et al. 2011), although other process may also contribute, including horizontal thermal advection by geostrophic and Ekman motions and atmospheric thermal forcing (Nakamura and Kazmin 2003; Kelly and Dong 2004; Kwon et al. 2010). Consistent with previous analysis of ship-measured SST data (Nakamura and Kazmin 2003), over the last several decades the SAFZ in winter (January–March) underwent a pair of major warm and cold periods associated with the northward and southward displacement of the SAFZ, respectively. Specifically, the warm period around 1970 was followed by the cold period in the mid-1980s, with a marked SST decline over the North Pacific in the late 1970s, known as the climate regime shift (e.g., Yasuda and Hanawa 1997).
In the following, the 3-yr running mean SST-SAFZ index is used as a reference time series (thick black lines in Fig. 2a) for identifying the associated anomalies in the surface heat fluxes (section 3b) and atmospheric circulation1 (section 4). The 3-yr running mean can eliminate most of the year-to-year fluctuations observed, for example, as remote influences of El Niño–Southern Oscillation (ENSO), while retaining signals associated with (quasi-) decadal SST anomalies in the SAFZ. It should be emphasized that the running mean was applied to the year-to-year time series of the index separately for the individual calendar months so as to retain (sub)seasonality in the low-frequency variability.2
b. Seasonal evolution of SST and heat flux anomalies
The typical (sub)seasonal evolution of SST anomalies in the SAFZ is shown in Fig. 3a as local coefficients of its correlation rSST (l, January, M) (colored) and linear regression aSST (l, January, M) (contoured) with the January SAFZ-SST index. As a typical warm anomaly for a unit anomaly (1 K) of the January index, SST anomalies aSST as strong as 1 K K−1 are confined meridionally to the SAFZ, where the correlation exceeds 0.6. The confinement of the SST anomalies to the SAFZ is particularly pronounced in midwinter through spring (January–March), in which decadal-scale fluctuations are dominant in the SAFZ-SST index, whereas the meridional confinement is somewhat less pronounced if SST anomalies are correlated with the index for fall or early winter (October–December; not shown).
Maps of correlation (color shaded) and linear regression (contoured) coefficients of the SAFZ-SST index for January with monthly (a) SST and (b) surface turbulent heat flux during the cold season (October–March) based on the ICOADS data. Coloring convention for the correlation coefficients is shown at the bottom of (b). The contour interval for the regression coefficients is shown at the right-bottom corner of (a),(b).
Citation: Journal of Climate 25, 1; 10.1175/JCLI-D-11-00046.1
Figure 3a also indicates that SST anomalies around the SAFZ tend to persist throughout the cold season, which can be confirmed in the autocorrelation of the SAFZ-SST index evaluated for individual calendar months (Fig. 4a). The anomalies correlated with the January index, for example, emerge along the SAFZ as early as in October, remaining until as late as March (Fig. 3a).
(a) Autocorrelation of the monthly SAFZ-SST index (contoured) based on the ICOADS data (SAFZ-SST) and its cross correlation with the NPI index (color shaded; see Fig. 5 for the domain) evaluated from SLP based on the NCEP–NCAR reanalysis. For the cross correlation, the ordinate indicates the calendar month taken for the SAFZ-SST index whose interannual anomaly is correlated with the NPI index for a particular month indicated on the abscissa. The dashed line in the middle corresponds to simultaneous correlations, and the upper-right (lower left) domain of the line indicates correlations for the SAFZ-SST index leading (lagging) the NPI index. Other dashed lines correspond to the correlations with ±12-month lag. (b) As in (a), but for the correlations based on the CFES-M integration.
Citation: Journal of Climate 25, 1; 10.1175/JCLI-D-11-00046.1
The KOE region, including the SAFZ, is known as a region of intense air–sea heat exchanges, especially in winter, through particularly strong turbulent heat fluxes in the climatological-mean state with their large interannual variations (Qiu et al. 2007; Bond and Cronin 2008; Joyce et al. 2009; Konda et al. 2010; Kelly et al. 2010; Kwon et al. 2010; Jensen et al. 2011). As such, the relationship between anomalies in SST and surface heat fluxes in the region has been extensively studied (e.g., see Frankignoul 1985; Kwon et al. 2010, for reviews). Here we examine the seasonal evolution of turbulent latent and sensible heat fluxes at the surface in correlation with the SAFZ-SST index. The heat flux anomalies (Fig. 3b) are spatially less organized than the associated SST anomalies, partly because fewer data are available for evaluating heat fluxes than for SST as mentioned in section 2a. Nevertheless, our analysis of the heat flux anomalies in correlation with the SAFZ-SST index is overall in agreement with previous analyses. In fact, a spatially coherent band of enhanced upward heat fluxes over warm SST anomalies along the SAFZ is apparent particularly in November, February, and March, acting to damp the decadal-scale SST anomalies while exerting thermodynamic forcing on the overlying atmosphere (Frankignoul and Kestenare 2002; Tanimoto et al. 2003).
The positive correlation in anomalies between SST and local upward heat flux is observed to be more significant in late winter through early spring (February–March) than in early winter (December–January), suggesting a more (less) dominant role of heat flux in damping (forcing) on SST anomalies (Frankignoul et al. 1998) toward the late winter through early spring. This is also consistent with Tanimoto et al. (2003), who argued that the weakened surface westerlies in the central Pacific associated with the weakened Aleutian low (AL) contribute to the maintenance and eastward development of the warm SST anomalies by counteracting the damping effect in early winter through anomalous surface heat fluxes. They showed that the anticyclonic anomalies then weaken gradually into late winter, reducing the atmospheric contribution, and thereby increasing a contribution from the SST damping, to the heat flux anomalies, which is also the case in our analysis (section 4b) Our analysis further reveals an even more distinct subseasonal transition. Specifically, the warm SST anomalies in January along the SAFZ tend to be concomitant with virtually no net anomalous turbulent heat fluxes in the core region of the SAFZ (140°–160°E) and negative heat flux anomalies to the east. This indicates a significant contribution from heat flux forcing to the SST anomalies (Frankignoul et al. 1998; Frankignoul and Kestenare 2002), presumably due to interannual variability and decadal modulations in the East Asian winter monsoon and the associated air–sea heat exchange over the KOE region in early winter (Nakamura and Yamagata 1999; Nakamura et al. 2002; Yoshiike and Kawamura 2009; Kwon et al. 2010).
4. Observed atmospheric anomalies
a. Seasonal evolution
In this section, we examine atmospheric anomalies correlated with the SAFZ-SST indices. To obtain an overview of the typical seasonal evolution of the atmospheric anomaly fields, a matrix of maps of the regression aSLP (l, m, M) and correlation rSLP (l, m, M) of the NCEP–NCAR SLP fields with the SAFZ-SST index, has been constructed by scanning their lead–lag relationships between the reference and targeted calendar months (m and M, respectively). Each of the diagonal panels in Fig. 5 shows simultaneous correlation/regression for a particular calendar month (m = M). Panels in the upper-right sector of Fig. 5 display a possible atmospheric (SLP) response (for M) to SST anomalies in the SAFZ (for m) as the latter leads the former. Panels in the lower-left sector, on the contrary, represent atmosphere forcing on SST anomalies in the SAFZ.
Maps of lead-lag correlation (color shaded) rSLP (l, m, M) and linear regression (contoured) coefficient aSLP (l, m, M) between monthly SLP anomalies based on the NCEP–NCAR reanalysis from October to March (M, columns) and the monthly SAFZ-SST index based on the ICOADS data from October to March (m, rows). The coloring convention for the correlation is shown at the bottom. Contour interval for the regression coefficients is 1 hPa K−1. The purple rectangular box in the top middle indicates the domain for defining the NPI-index shown in Fig. 4.
Citation: Journal of Climate 25, 1; 10.1175/JCLI-D-11-00046.1
The sequence of maps in Fig. 5 reveals distinct (sub)seasonal evolution in SLP anomalies. Regardless of a calendar month m chosen for the reference time series, spatially coherent SLP anomalies first emerge in December and become matured in January, although the incipient anomalies are hinted in November. Positive (negative) SST anomalies in the SAFZ are accompanied by basin-scale anticyclonic (cyclonic) anomalies in January over the central North Pacific (centered at around 55°N, 170°W), corresponding to the weakening (strengthening) of the AL (see Fig. A4a for the climatological SLP in January). The particular SLP anomalies are statistically significant. Their simultaneous correlation with the SST-SAFZ index rSLP (l, January, January) exceeds 0.6 in the vicinity of the climatological AL center, and the corresponding linear regression coefficient aSLP (l, January, January) reaches as much as 8 hPa (1 K)−1 change in the SAFZ-SST index. This well-organized SLP signal in January is found to collapse suddenly in February. Compared to the January situation, the SLP anomalies in February and March over the North Pacific are markedly weaker and less coherent spatially. Specifically, the magnitude of the simultaneous correlation is below 0.3, except off California, and the simultaneous regression coefficients are less than 1 hPa K−1 over the entire North Pacific basin.
The seasonal locking of SLP anomalies in correlation with SST anomalies along the SAFZ as revealed in Fig. 5 can be confirmed in the cross correlation between the SAFZ-SST index and the North Pacific Index (NPI; colored in Fig. 4a). The NPI, defined as a weighted area average of SLP anomalies over the central North Pacific region (30°–65°N, 160°E–140°W; purple rectangular in Fig. 5) represents the AL variability (Trenberth and Hurrell 1994). Figure 4a highlights the strongest positive correlation of NPI for December and January, indicating that SST anomalies in the SAFZ persistent throughout the cold season are anticorrelated significantly with the AL intensity only in these two calendar months. Also evident in Fig. 4 is a sharp (sub)seasonal decline in the particular AL signal from January to February, in good agreement with Fig. 5. It is noteworthy that, despite the collapse of the AL signal in February, it recurs in the next January (Yr2), whose implication is discussed in section 6. Figure 4a also displays weak positive correlations of the April NPI with the SAFZ-SST index for April–June and strong negative correlation of August–September NPI with the SAFZ-SST index for the preceding winter through early summer. These correlations involve SLP anomalies residing over the eastern North Pacific and the west coast of North America (not shown), which differ from the summertime atmospheric anomalies over the KOE region pointed out previously (Frankignoul and Sennéchael 2007; Nakamura and Yamane 2010). Detailed investigations on these spring-to-summer signals are left for future studies.
b. Subseasonal contrast in the atmospheric response
We now perform a closer investigation of the contrasting atmospheric anomalies associated with the SAFZ-SST index between January and February. A particular focus is placed on the atmospheric anomalies correlated with the SAFZ-SST index for preceding calendar months. Those anomalies can be considered as an atmospheric response to SST anomalies in the SAFZ unless any other concomitant remote forcing exists (see section 4c). Figure 6 compares hemispheric distributions (20°S–80°N) of anomalies in SST, SLP, and Z250 in those two calendar months correlated with the SAFZ-SST index in November. If the atmospheric anomalies are regarded as the response to the November SST anomalies, the leading time is thus 2 or 3 months, which are longer than the atmospheric persistence time scale (~2–10 days; Frankignoul and Kestenare 2002). The maps represent their typical anomalies that tend to be observed when SST in the SAFZ is warmer than its climatology, whereas the signs of the anomalies have to be reversed for the situation in which SST in the SAFZ is cooler than its climatology.
Lag-correlation (color shaded) and regression coefficients (contoured) of January anomalies of (a) SST, (b) SLP, and (c) Z250 with the November SAFZ-SST index. The index is based on the ICOADS data, while SST anomalies in (a) are based on HadISST data. SLP and Z250 are based on the NCEP–NCAR reanalysis. (d)–(f) As in (a)–(c), but for February anomalies of (d) SST, (e) SLP, and (f) Z250. The contour intervals for the regression coefficients are shown at the lower-right corner of each with units of K K−1, hPa K−1, and m K−1 for SST, SLP, and Z250, respectively. The purple line in (a)–(c) shows a section that transects the PNA-like anomalies in January (see Fig. 7).
Citation: Journal of Climate 25, 1; 10.1175/JCLI-D-11-00046.1
The January SLP anomalies (Fig. 6b) are characterized by anticyclonic signal over the AL region as shown in Fig. 5. The corresponding Z250 anomalies (Fig. 6c) resemble the Pacific–North American (PNA) teleconnection pattern (Wallace and Gutzler 1981), as they consist of an anticyclonic anomaly over the Aleutian region and cyclonic anomalies to its south and over the west coast of North America. As shown in the cross section in Fig. 7a across the AL region and the SAFZ (along the red line in Fig. 6a), this anticyclonic signal is vertically coherent, exhibiting equivalent barotropic structure. A close inspection of Fig. 7a reveals, however, the anticyclonic height anomaly exhibits an apparent, westward phase tilt with height. From a viewpoint of Rossby wave dynamics, this phase tilt is equivalent to an upward and (north) eastward wave-activity flux and thus supportive of the interpretation that this PNA-like anomaly is a response to the heat source induced by SST anomalies over the SAFZ in the background westerlies. In fact, a shallow, warm, low-level cyclonic anomaly evident around the SAFZ with above-normal SST and the deeper, more barotropic anticyclonic anomaly observed downstream are consistent with theoretical argument (Hoskins and Karoly 1981; Held 1983). In addition to the anomalous surface sensible heat flux (Fig. 3b), enhanced precipitation around the SAFZ associated with anomalous storm-track activity also acts as diabatic heating (not shown). From a viewpoint of energetics, the anomaly pattern with baroclinic structure can maintain itself against dissipative processes by extracting available potential energy from the vertically sheared background westerlies. Figure 6c also indicates that the upper-tropospheric anomalies consist of a circum-global wave train with cyclonic anomalies over the central North Atlantic and central Eurasia and an anticyclonic anomaly over western Europe. Atmospheric anomaly patterns similar to the one identified for January in this study have been detected in earlier observational and modeling studies: a late fall to early winter signal resembling the PNA pattern correlated with a quadripolar SST anomaly during the preceding summer as extracted through lagged maximum covariance analysis from the NCEP–NCAR reanalysis for the 1977–2004 period (Frankignoul and Sennéchael 2007) and an early-winter atmospheric response to subsurface temperature anomalies in the KOE region simulated in a CGCM experiment (Liu et al. 2007). The SST anomaly patterns extracted in those studies are, however, much broader meridionally than those confined to the SAFZ in the present study due to the variance maximizing nature of their statistical analysis method (Frankignoul and Sennéchael 2007) or the horizontal model resolution insufficient for resolving the SAFZ (Liu et al. 2007).
Vertical sections of lag-regression coefficients of January anomalies of temperature (K, color shaded) and geopotential height (contoured for every 5 m) with the normalized November SAFZ-SST index. The sections transect the PNA-like anomalies shown in Figs. 6a–c (purple lines). (a) The index is based on the ICOADS data, while the atmospheric fields are based on the NCEP–NCAR reanalysis data. (b) Both the index and atmospheric fields are based on the 120-yr CFES-M integration. The lag-regression coefficients in (a) and (b) represent typical atmospheric anomalies corresponding to a unit standard deviation of the November SAFZ-SST index (with warm SST anomalies).
Citation: Journal of Climate 25, 1; 10.1175/JCLI-D-11-00046.1
Compared to its January counterpart, the atmospheric anomaly pattern observed in February correlated with the November SAFZ-SST index is much weaker (Figs. 6e,f). As opposed to the January signal, the February atmospheric anomaly pattern is characterized by a north–south dipole pattern with a cyclonic anomaly at the northern center of action over the AL region in association with warm SST anomalies in the SAFZ. The differences in the atmospheric circulation response between January and February are investigated further in relation to transient eddy feedback in section 4d.
c. Tropical influence
We have shown that the most prominent atmospheric signal in January correlated with the preceding SAFZ-SST index is nearly in equivalent barotropic structure and similar to the PNA teleconnection pattern. One may wonder whether this PNA-like signal is an atmospheric response to extratropical SST anomalies in the SAFZ or a response to other remote forcing covarying with SAFZ-SST. Earlier studies emphasized an aspect of PNA-like anomalies as a remote response to tropical convective heating (e.g., Horel and Wallace 1981; Nitta and Yamada 1989; Trenberth 1990; Trenberth et al. 1998), called the “atmospheric bridge” (Lau 1997; Alexander et al. 2002). Meanwhile, through a rotated principal component analysis of midtropospheric height anomalies observed in winter, Nigam (2003) differentiated the PNA pattern from another anomaly pattern associated with ENSO. Indeed, several studies argue that the PNA pattern can be regarded as an internal dynamical mode over the wintertime North Pacific that can develop without remote influence from the tropics (e.g., Simmons et al. 1983; Wallace and Lau 1985; Nakamura et al. 1987, 2010; Nakamura 1996).
Given the multiple factors that can contribute to the development of a PNA-like anomaly pattern, we first examine the global distribution of SST anomalies based on HadISST for January and February correlated with the November SAFZ-SST index (Figs. 6a and 6d, respectively). For each of the calendar months, the HadISST data represent the confinement of SST anomalies to the SAFZ, manifested as a zonally elongated band of high correlation (regression) coefficients that exceed 0.6 (0.5 K K−1; contours). Flanking the positive SST anomalies in the SAFZ, regions of negative SST anomalies along the west coast of North America all the way south to the South American coast. Unlike in the well-known horseshoe pattern associated with the PDO (Mantua et al. 1997), however, spatially coherent negative SST anomalies are observed only to the north of 20°N.
Though rather small in magnitude (with the regression coefficients less than 0.5 K K−1), La Niña–like SST anomalies over the off-equatorial tropical Pacific and SST anomalies similar to the Atlantic Multidecadal Oscillation (AMO; Schlesinger and Ramankutty 1994) are found to be correlated significantly with the SAFZ-SST index. Figure 8 shows lag-correlation/regression maps of January anomalies in SST, SLP, and Z250 with November ENSO and AMO indices, defined as the first principal component (PC-1) of the 3-yr running-mean tropical Pacific SST anomalies (5°S–5°N, 120°E–90°W) and standardized SST anomalies averaged over the North Atlantic basin (0°–70°N, 60°W–0°), respectively. Correlated with the negative PC-1 (La Niña phase), there are significant anticyclonic atmospheric anomalies over the midlatitude eastern Pacific (Figs. 8b,c). Obviously, those anomalies are shifted southeast of the PNA-like signal correlated with the SAFZ-SST index, representing the variability of the subtropical anticyclone in association with (quasi) decadal SST anomalies in the tropical Pacific and along the subtropical oceanic front (Nakamura et al. 1997). Likewise, the AMO does not accompany any significant atmospheric anomalies in the extratropical Pacific, but exhibits strong negative correlation with SLP anomalies over a tropical belt extending zonally from Africa through the Indian Ocean to the subtropical western Pacific (Fig. 8e), the latter signal much like the SAFZ-SST correlated SLP anomalies (Fig. 6b). While the analysis in the previous sections does not totally exclude the influences of the remote tropical SST anomalies, this additional analysis further suggests that the equivalent barotropic PNA-like signals in the North Pacific is primarily induced by (quasi) decadal SST anomalies in the SAFZ, given that the remote SST signals in the tropical Pacific and Atlantic associated with SAFZ-SST (Figs. 6a,d) are much weaker than the ones associated with ENSO and AMO indices, respectively (Figs. 8a,d).
(a)–(c) As in Figs. 6a–c, but correlated/regressed with the November ENSO index, which is based on the first principal component of the 3-yr running averaged SST anomalies over the tropical Pacific (5°S–5°N, 120°E–90°W). The sign is chosen such that the positive index corresponds to the La Niña phase. (d)–(f) As in (a)–(c), but with the November AMO index, which is based on the 3-yr running mean SST anomalies averaged over the North Atlantic (0°–70°N, 60°W–0°). The purple rectangular domains in (a) and (c) designate those for defining the ENSO and AMO indices, respectively.
Citation: Journal of Climate 25, 1; 10.1175/JCLI-D-11-00046.1
d. Transient eddy feedback
Previous studies have suggested that atmospheric transient eddies migrating along a storm track play a major role in the maintenance of an equilibrium response of the atmosphere to extratropical SST anomalies through their heat and vorticity fluxes (Kushnir et al. 2002; Deser et al. 2007; and references therein). Furthermore, recent studies have elucidated that an oceanic frontal zone can be important in anchoring and maintaining a nearby storm track by allowing efficient restoration of cross-frontal gradient of surface air temperature (SAT) by differential heat supply from the ocean (Nakamura et al. 2004, 2008; Taguchi et al. 2009; Nonaka et al. 2009; Sampe et al. 2010; Hotta and Nakamura 2011). One may thus question whether SST anomalies associated with SAFZ variability can influence basin-scale atmospheric circulation by modulating storm-track activity and associated vorticity and heat fluxes.
Figures 9b,f (color shading) display anomalous low-level storm-track activity in January and February, respectively, as measured by monthly anomalies in meridional eddy heat flux at the 850-hPa level, correlated with the November SAFZ-SST index. Climatologically (red contours in Figs. 9b,f), the low-level storm-track core is along the SAFZ both in January and February (Nakamura et al. 2004). The January anomalies (Fig. 9b) represent a poleward shift of the storm track over the central and eastern sectors of the North Pacific basin associated with warm SST anomalies in the SAFZ. This storm-track shift, which is concomitant with the development of anticyclonic stationary anomalies over the Aleutian region, exerts positive feedback forcing on the lower-tropospheric anticyclonic anomaly, as inferred from the monthly anomalous ∂Z/∂t due to eddy vorticity and heat fluxes that is correlated positively with (regressed upon) the SAFZ-SST index (color shading in Fig. 9c). At the 250-hPa level, the corresponding positive correlation is distributed somewhat less coherently with the weaker regressed anomalies. In agreement with Tanimoto et al. (2003), however, the transient eddy feedback forcing is nevertheless coherent between the upper and lower troposphere, reinforcing the equivalent barotropic anomalies observed in January over the North Pacific. In contrast, the anomalous storm-track activity correlated with the November SAFZ-SST index is much weaker in February (color shading in Fig. 9f), despite the fact that no substantial differences exist climatologically between those two calendar months. Likewise, the associated transient eddy feedback forcing in February is substantially weaker and less organized spatially than in January (Figs. 9g,h).
(a) Lag-correlation (color shaded) and regression coefficients (contoured) of anomalous near-surface baroclinicity in January with the November SAFZ-SST index (averaging domain indicated with a black rectangular box) based on the ICOADS data. The baroclinicity is measured by meridional gradient of 925-hPa potential temperature based on the NCEP–NCAR reanalysis. (b) As in (a), but for anomalous storm-track activity measured as the 850-hPa poleward heat flux associated with synoptic-scale disturbances 
Citation: Journal of Climate 25, 1; 10.1175/JCLI-D-11-00046.1
The distinct January–February contrast observed in the anomalous storm-track activity and its feedback forcing on the stationary circulation anomalies may be attributable, at least in part, to the corresponding contrast in anomalous near-surface baroclinicity between the two calendar months, given its crucial role for the baroclinic growth of synoptic-scale disturbances (e.g., Hoskins et al. 1985; Nakamura et al. 2004, 2008). In association with positive SST anomalies in the SAFZ, the anomalous near-surface baroclinicity measured by meridional gradient of 925-hPa potential temperature in January (color shaded in Fig. 9a) is characterized by the northward shift of the baroclinicity zone, which climatologically forms slightly southward of the SAFZ (as indicated with purple contours). This shift of the baroclinic zone is consistent with the anomalous northward shift of the storm-track activity (Fig. 9b). In contrast, the corresponding anomalous baroclinicity in February (Fig. 9e) is spatially less organized, which is also consistent with the less coherent anomalies in the storm-track activity in February. The January–February contrast in the anomalous baroclinicity could arise from the corresponding contrast in anomalous turbulent heat fluxes. In December coherent upward turbulent heat flux anomalies (Fig. A2a, left) tend to be observed over the warm SST anomalies in the SAFZ (Fig. A1a, left), while virtually no net turbulent heat flux anomalies tend to be observed in January along the warm SAFZ (Fig A2a middle) in the presence of weakened surface westerlies due to the predominant AL anomalies as discussed in section 3b. Assuming that a few weeks are needed for the AL anomalies to be developed and adjust the surface baroclinicity, heat flux anomalies in December may be able to sustain anomalous baroclinicity effectively in January, while the incoherent heat flux anomalies in January may result in the incoherent anomalous baroclinicity in February.
5. Coupled model simulation
In this section, we analyze the long-term CFES-M integration, focusing on low-frequency anomalies at the ocean surface in the SAFZ and the associated atmospheric anomalies simulated, as a complement to the observational analysis discussed in the sections 3 and 4. Validation for the simulated fields against the observations is given in appendix.
a. Simulated decadal SST variability in the SAFZ
The CFES-M is found to capture salient features of the climatological-mean field of SST and its variability observed in the North Pacific. Specifically, the model can simulate a sharp frontal structure climatologically in the wintertime SST field east of Japan around 40°N that corresponds to the observed SAFZ (Fig. 1b). As actually observed (Fig. 1a), the interannual/decadal variability in the simulated SST is strongest along the model SAFZ, with weaker variability to its east and along subtropical frontal zone north of Hawaii. The SST variability in the SAFZ includes pronounced decadal-scale fluctuations (thick black line in Fig. 2b), which are highly correlated with the meridional excursion of the model SAFZ marked by the latitude of the maximum meridional SST gradient (thin gray line in Fig. 2b). These features are in agreement with the observations, but there are several unrealistic features in the simulated SST field. For example, the meridional gradients of the climatological SST across the model SAFZ are unrealistically tight, because the simulated Kuroshio Current overshooting before its detachment from the Japanese coast advects too much warm water into the mixed water region between the KE and Oyashio Extension. This is a typical model bias in the separation of western boundary currents simulated in noneddy permitting OGCMs included in most of the climate models. Nevertheless, decadal SST anomalies confined to a meridionally narrow frontal zone simulated in CFES-M are worth examining, since not many climate models can simulate such frontal SST gradient owing to their coarser horizontal resolutions. It should be stressed that investigating specific mechanisms for the generation of decadal SST variability in CFES-M is beyond the scope of the present study. Rather, we focus on the seasonal evolution of the simulated decadal SST anomalies and associated atmospheric anomalies, in comparison with the corresponding observed features presented in the previous section.
b. Simulated seasonal evolution of atmospheric anomalies associated with SAFZ-SST variability
The seasonal evolution of decadal SST anomalies in the SAFZ and the North Pacific SLP anomalies simulated in CFES-M are examined by constructing the autocorrelation of the 3-yr running-mean monthly SST anomalies averaged within the SAFZ (extracted in the model SAFZ-SST index; Fig. 2b) and its cross correlation with the simulated NPI (Fig. 4b). Again, the running mean was applied to the model data separately for the individual calendar months. For constructing the model SAFZ-SST index, the area (38°–43°N, 142°–160°E; as indicated in Fig. 1b with a black rectangular box) was chosen in which simulated SST variability is particularly strong. As evident in Fig. 4b, the CFES-M integration overall captures the observed features on the seasonality of the SAFZ-SST and North Pacific SLP anomalies, including the seasonally persistent SST anomalies in the SAFZ and the seasonal locking of its correlation with NPI. Compared with its observational counterpart (Fig. 4a), the model SAFZ-SST index (Fig. 4b) exhibits even higher persistence with its monthly autocorrelations exceeding 0.8 throughout the cold season from October to the following May. The persistence is still high (with autocorrelation above 0.6) even in late summer and early autumn, in which the persistency of the observed SST anomalies drops substantially. The particularly high persistence of the model SST anomalies is presumably due to an unrealistically deep mixed layer found around the SAFZ in CFES-M (not shown), which may be attributable to the overshooting Kuroshio Current. In addition, the high cross correlation between the SAFZ-SST index for any calendar month and the January NPI and the rapid decline of the corresponding correlation with the February NPI are both reproduced in the CFES-M simulation (Fig. 4b).
In the following, we compare the simulated atmospheric anomalies between January and February. Figure 10 contrasts geographical distributions of lag-correlation/regression coefficients of simulated SST, SLP, and Z250 anomalies between January and February both evaluated with the model SAFZ-SST index for November (as the model counterpart of Fig. 6). Both in January and February, the SST anomalies (Figs. 10a,d) exhibit their strongest signal along the SAFZ as observed, although the simulated SST anomalies are stronger than the observed (Fig. A1). Meanwhile, the signals in the tropics and North Atlantic are weaker than their observational counterpart (Figs. 6a,d).
As in Fig. 6, but based on the 120-yr CFES-M integration.
Citation: Journal of Climate 25, 1; 10.1175/JCLI-D-11-00046.1
In agreement with the observations (Figs. 6b,c), the anticyclonic anomalies simulated for January in association with the warm SAFZ in November exhibit equivalent barotropic structure, manifested as the weakening of the surface AL and the corresponding phase of the PNA pattern aloft (Figs. 10b,c). Though weaker than their observational counterpart, those signals in the 120-yr CFES integration are statistically significant at the 95% confidence level (color shaded region in Fig. 10) based on the two-sided Student’s t test with degrees of freedom assumed to be 39 for those 3-yr running mean anomalies. Compared to the observations, the simulated atmospheric anomalies in January are confined more strongly to the extratropics, with virtually no significant signals in the tropics. Furthermore, the three-dimensional structure of the simulated atmospheric anomalies (Fig. 7b) is essentially the same as its observed counterpart (Fig. 7a), although the amplitude of the former is ~50% smaller despite the larger temperature anomalies simulated over the SAFZ than the observed. The PNA-like anomaly pattern in the model is thus likely a manifestation of the coupled ocean–atmosphere variability in the North Pacific, and it may be forced and maintained by the SST anomalies in the SAFZ. In agreement with the observations, surface heat flux anomalies simulated in the SAFZ tend to be locally positive when SST is warmer (Fig. A2). In addition, enhanced precipitation around the SAFZ associated with anomalous storm-track activity can act as diabatic heating, whose magnitude is greater than in the reanalysis data (not shown). This is probably due to the oversensitivity of model precipitation to the underlying SST.
In agreement with the observations, no coherent atmospheric anomalies emerge in February in correlation with the model SAFZ-SST index (Figs. 10e,f). As suggested also by the observational analysis, a possible factor that gives rise to the January–February contrast may be subtle differences in modulations in the storm-track activity and its feedback on quasi-stationary anomalies. Figure 11 shows the meridional gradient of 925-hPa potential temperature, the 850-hPa poleward heat flux associated with synoptic-scale eddies, and geopotential height tendencies at the 850-hPa and 250-hPa levels due to eddy heat and vorticity fluxes based on (2) for the CFES-M simulation. Though slightly overestimated, the climatological-mean storm-track activity over the North Pacific is well reproduced in the model both for January and February (red contours in Figs. 11b,f, respectively), including the collocation of the storm-track core with the SAFZ. Moreover, the model can also reproduce the meridional shift of the near-surface baroclinicity and the storm track in January that follows the emergence of anomalous SST along the SAFZ in November. Consistent with the observational analysis, the northward shift of the storm track associated with positive SST anomalies in the SAFZ results in anticyclonic feedback forcing over the subpolar central Pacific (Fig. 11c), where the equivalent barotropic anticyclonic anomaly is reproduced in the model (Figs. 10b,c). No such transient eddy feedback forcing is simulated, however, in February (Figs. 11g,h), which is consistent with our observational analysis (Figs. 9g,h). Furthermore, unlike in the NCEP reanalysis, the simulated dipolar anomalies in near-surface baroclinicity remain in February as prominent as in January.
As in Fig. 9, but based on the 120-yr CFES-M integration.
Citation: Journal of Climate 25, 1; 10.1175/JCLI-D-11-00046.1
c. Phase realization of the intrinsic PNA mode
We have shown that the distinct January–February contrast in atmospheric response to decadal-scale SST anomalies in the SAFZ is a robust feature represented both in the observations and CFES-M integration. The confinement of the atmospheric anomalies simulated for January to the extratropics may be a manifestation of the PNA teleconnection pattern as an intrinsic dynamical mode in the extratropical atmosphere. While clarifying the specific mechanisms for triggering the PNA-like response is beyond the scope of this study, a hypothesis could be put forward that the sign of SST anomalies in the SAFZ could influence the realization of the preferred phase of the PNA-like anomaly pattern, presumably through the modulated storm-track activity and resultant feedback forcing (section 4d). In the following, we attempt to sort out the probability of the phase realization of the PNA-like pattern depending on the sign of the SAFZ-SST index, both in the observations and CFES-M integration.
Given that the first EOF of SLP variability well represents the AL variability as a surface manifestation of the PNA pattern in both January and February (Fig. A4), we examine the probability density function (PDF) of the corresponding principal component (SLP PC-1) with respect to the SAFZ-SST index for both the observations and CFES-M integration. In Fig. 12, the PDF histograms of SLP PC-1 are constructed separately for the positive (black lines) and negative (gray shaded) phases, for each of which the November SAFZ-SST index is greater in magnitude than its unit standard deviation. The histograms of SLP PC-1 for January are well separated between the positive and negative phases of the SAFZ-SST index. For both the observations (Fig. 12a) and in the model (Fig. 12c), January SLP PC-1 is inclined to be positive (negative), corresponding to the weakened (enhanced) AL, for positive (negative) SST anomalies in the SAFZ. The atmospheric state over the North Pacific in January thus shows certain dependency on decadal-scale SST anomalies in the SAFZ (i.e., anticyclonic SLP anomalies and the negative phase of the PNA pattern for the positive SAFZ-SST index and vice versa), which may be interpreted as the sensitivity of the PNA-like anomaly pattern in January to the SST anomalies as a preferred mode of the atmospheric variability. In February, however, the particular dependency is less significant or even absent (Figs. 12b,d).
PDF of the first principle component of monthly SLP anomalies shown in histograms separately for the two categories in which the November SAFZ-SST index exceeds its one standard deviation positively (black line) and negatively (shaded). The histograms are based on the (a),(b) NCEP–NCAR reanalysis and (c),(d) CFES-M integration for (a),(c) January and (b),(d) February.
Citation: Journal of Climate 25, 1; 10.1175/JCLI-D-11-00046.1
6. Summary and discussion
In this study anomalous atmospheric circulation that tends to occur in association with decadal SST variability in the North Pacific SAFZ has been investigated with historical observations for the period 1959–2006 and a 120-yr CGCM integration. Consistent with previous studies, our observational analysis has confirmed that pronounced decadal SST variations are confined to the SAFZ especially in late winter (January–March) in association with its meridional shift. Further analysis has revealed that the association between the SST anomalies in the SAFZ and a PNA-like atmospheric anomaly pattern is strongest and most robust in January, but this association breaks down in February despite the persistency of the SST anomalies throughout the cold season (October–March). With its coherent precursory signal that starts emerging in November, the PNA-like signal in January is so robust and prominent as to emerge even in association with the SST anomalies in autumn and early winter, suggestive of oceanic influence on the atmosphere. The PNA-like pattern is equivalent barotropic, modulating the surface AL in such a manner that it tends to be weakened in association with positive SST anomalies in the SAFZ. The anticyclonic feedback forcing due to modulated activity (northward shift) of the North Pacific storm track appears to be effective in maintaining the equivalent barotropic anomaly pattern.
The CGCM integration is found to well reproduce the observed features, including pronounced decadal SST variability confined to the SAFZ persistent throughout the cold season, its robust association with equivalent barotropic PNA-like anomaly pattern in January and the rapid breakdown of the association in February. The association of the January PNA-like signal in the model with SST anomalies in the SAFZ in the preceding calendar months appears to be even tighter than in the observations, accompanying virtually no SST signals in the tropics. This association suggests that the PNA-like anomaly pattern in January may be an atmospheric response to seasonally developing decadal-scale SST anomalies in the SAFZ.
In both the observations and CGCM integration, the North Pacific storm track in January tends to be shifted northward in association with positive SST anomalies along the SAFZ in late fall and early winter. The anomalous storm-track activity accompanies anomalous precipitation in the vicinity of the SAFZ, which acts as diabatic heating in the same sense as anomalous surface heat release from the ocean. In January the storm-track modulation induces eddy feedback forcing that reinforces the PNA-like equivalent barotropic anomalies, while the modulation and associated eddy feedback are weaker and spatially less organized in February. While the cause of this distinct subseasonal contrast has not definitely been identified, possible mechanisms can nevertheless be argued. First, a more coherent pattern of modulated storm-track activity in January than in February may arise from the more coherently organized anomalous near-surface baroclinicity observed around the storm track in January. Though no such distinct January–February contrast is simulated in anomalous near-surface baroclinicity in the CGCM, the observed contrast may be induced by subseasonal modulations of anomalous surface turbulent heat fluxes as discussed in section 4d. Second, subtle January–February differences in the background westerlies could be another factor. Brayshaw et al. (2008) suggested through aquaplanet AGCM experiments that the relative position of anomalous SST gradient to that of the background subtropical jet can be a crucial factor for the large-scale atmospheric response. As will be shown (see Fig. A3), the Asian–Pacific subtropical jet over the western Pacific basin takes its southernmost position in February both in the reanalysis and the CFES-M integration (white contour), while the SST anomalies are confined meridionally to the SAFZ and their latitudinal position (black contour) is thus unchanged on the poleward side of the jet from January to February. The anomalous SST gradient more distant from the subtropical jet in February may be less effective in modulating the storm-track activity and the midlatitude eddy-driven jet, which seems consistent with the argument by Brayshaw et al. (2008) and with the “eddy trapping” mechanism by Nakamura and Sampe (2002). Seasonal migration of the subtropical jet may also influence the near-surface baroclinicity around the SAFZ via interaction between the jet and the thermal contrast between the Asian continent and the Pacific, as shown for the North Atlantic basin through a series of hierarchical semirealistic AGCM experiments by Brayshaw et al. (2009). The sudden breakdown of the PNA signal from January into February revealed in both observations and our CGCM is reminiscent of AGCM experiments by Peng et al. (1997) and Peng and Whitaker (1999). They obtained two qualitatively different responses in a model atmosphere to an identical midlatitude SST anomaly pattern along the North Pacific SAFZ under perpetual January and February conditions. Specific mechanisms involved in the subseasonal sensitivity of the atmospheric anomaly pattern should be addressed in a future study.
Recently, Frankignoul et al. (2011) have demonstrated large-scale atmospheric responses to the anomalous paths of the western boundary current extensions through regression analysis. They extracted a significant equivalent barotropic atmospheric response in winter (November–March), which resembles the NPO–western Pacific (WP) pattern (Wallace and Gutzler 1981) in association with the meridional shift of the Oyashio Extension (OE) front, an index similar to our SAFZ index but based on a high-resolution satellite-measured SST for 1982–2008. The difference in the atmospheric response pattern between our analysis (the PNA-like response in January) and theirs may arise from the difference in analysis period. Relying on satellite-measured SST data, their analysis is limited to the period after the climate regime shift, whereas our analysis based on ship-observed SST data for the longer period 1959–2006 highlights the changes before and after the regime shift.
Lag correlation (color shaded) and regression coefficients (contoured) of anomalous Ekman pumping velocity based on the NCEP–NCAR reanalysis for (a) January, (b) February, (c) winter mean (January–March), and (d) annual mean with the November SAFZ-SST index. (e),(f) As in (c),(d), but for the CFES-M integration. The contour interval (10−6 m s−1) for the regression coefficients is shown at the right-bottom corner of (a)–(f).
Citation: Journal of Climate 25, 1; 10.1175/JCLI-D-11-00046.1
Despite its seasonal breakdown after January, the PNA-like response to the (quasi) decadal SST anomalies in the SAFZ tends to reoccur in the next January. This suggests that the PNA-like anomaly pattern may develop as a response to the persistent SST anomalies that are maintained by oceanic processes, including the reemergence of SST anomalies (Alexander et al. 1999) and other ocean dynamics. Investigation is under way on whether the ocean–atmosphere positive feedback, as inferred from our observational and model analyses of anomalous wE, indeed enhances the decadal SST variability in the North Pacific SAFZ via the aforementioned oceanic response in the CFES-M integration. Investigation is also under way on how the anomalous wE is induced around the SAFZ as evident in Fig. 13 and whether it can exert positive and negative feedback locally on SST anomalies in the SAFZ.
Acknowledgments
We thank Hiroki Tokinaga and Takafumi Miyasaka for providing us the originally gridded ICOADS dataset and a FORTRAN code for solving the feedback forcing of transient eddy fluxes, respectively. We also thank Niklas Schneider, Bo Qiu, Young-Oh Kwon, Shang-Ping Xie, Shoshiro Minobe, Fei-Fei Jin, In-Sik Kang, and Mike Alexander for fruitful discussions; Claude Frankignoul, David Brayshaw, and an anonymous reviewer for their thoughtful and constructive comments on the earlier version of the paper; and Wataru Ohfuchi, Hideharu Sasaki, and Takeshi Enomoto for their efforts in developing and improving CFES. The integration of the CFES-M was carried out on the Earth Simulator under the support of JAMSTEC. BT and MN are supported in part by the Agriculture, Forestry, and Fisheries Research Council of Japan through the research project POMAL (Population Outbreak of Marine Life) and by the Japan Society for Promotion for Science (JSPS) through a Grant-in-Aid for Scientific Research (C) 21540458. HN is supported in part by JSPS through a Grant-in-Aid for Scientific Research (B) 22340135 and by the Japanese Ministry of Environment through the Global Environment Research Fund (S-5). NK and AKY are supported in part by JSPS through a Grant-in-Aid for Science Research (A) 22244057. All the authors (except AG) are also supported in part by the Japanese Ministry of Education, Culture, Sports, Science and Technology (MEXT) through a Grant-in-Aid for Scientific Research in Innovative Areas 2205.
APPENDIX
Validation of the CGCM Simulation
In the following, we provide supplementary validations of the CFES-M integration against observations with respect to some key features on the decadal variability of the SAFZ and its influence on the large-scale atmospheric circulation.
a. SST and surface heat fluxes
Figure A1 compares the climatological-mean SST (purple contour) over the western North Pacific for December, January, and February, and the corresponding decadal anomalies correlated with the November SAFZ-SST index (color shading for the correlation coefficient and black contours for the regression coefficient) between the ICOADS data and CFES-M integration. The model can reproduce the meridional confinement of the observed SST anomalies into the SAFZ, but both the amplitude of the anomalies and the tightness of their meridional confinement are overestimated. This overestimation of the simulated anomalies is consistent with the overestimation of the frontal gradient in the climatological-mean SST (purple contour; see also Fig. 1), which acts to yield stronger SST anomalies in association with the latitudinal excursion of the SAFZ in the model that is in realistic magnitude (Fig. 2). In fact, both the SST anomalies and climatological-mean SST gradient in the model are nearly twice as large as their observational counterpart (Fig. 2; cf. Thompson and Kwon 2010).
(a) Lag correlation (color shaded) and regression coefficients (contoured) of SST anomalies based on the ICOADS data for (left) December, (middle) January, and (right) February with the November SAFZ-SST index. (b) As in (a), but for the CFES-M integration. The contour interval (°C) for the regression coefficients is shown at the right-bottom corner of each panel. The corresponding monthly climatologies of SST are superimposed with purple contours. A light 9-point smoothing is applied to each of the fields.
Citation: Journal of Climate 25, 1; 10.1175/JCLI-D-11-00046.1
Figure A2 (purple contours) indicates that the climatological-mean surface heat fluxes for the winter months also tend to be overestimated by ~50% in the CFES-M integration than in the ICOADS data. This overestimation is owing to the warm bias within the mixed water region due to the overshooting of the model Kuroshio (as mentioned in section 5a) in addition to the overestimation of the surface westerly wind speed (not shown). In agreement with the observations, simulated heat flux anomalies correlated with the November SAFZ-SST tend to be upward over the warm SST anomalies in the SAFZ. Reflecting the overestimated intensity of the model SST anomalies (Fig. A1), the amplitude of the heat flux anomalies are much larger than observed, leading to unrealistically strong thermal forcing on the atmosphere over the SAFZ (Fig. 7b).
As in Fig. A1, but for surface turbulent (sensible plus latent) heat fluxes.
Citation: Journal of Climate 25, 1; 10.1175/JCLI-D-11-00046.1
b. Storm-track activity and upper-level westerly jet
The climatological-mean seasonal march in the upper-tropospheric zonal wind over the western North Pacific (150°E–180°) based on the NCEP–NCAR reanalysis for the 1959–2006 period (Fig. A3a) is characterized by the southward migration and intensification of the westerly jet stream from autumn into midwinter in step with the development of the East Asian winter monsoon (e.g., Nakamura and Sampe 2002; Nakamura et al. 2002). Exept in midwinter, the jet axis is collocated with the storm-track axis (color shading in Fig. A3a), showing its eddy-driven nature. In midwinter, by contrast, the jet axis is displaced southward from the storm track, exhibiting its hybrid characteristics between the eddy-driven and subtropical jets (Nakamura et al. 2004). This dislocation between the jet and storm-track axes is concurrent with the midwinter suppression of storm-track activity (Nakamura 1992; Nakamura and Sampe 2002). In the CFES-M integration (Fig. A3b), the Pacific storm-track activity is overall overestimated by 30%–~40%, which may be attributable, at least in part, to the unrealistically strong SST gradient across the SAFZ. Otherwise, the model well reproduces the observed seasonal march of the jet and storm track, including the midwinter minimum of the storm-track activity in association with the southward-displaced upper-level jet relative to the storm track that is anchored along the SAFZ. Both in the observations and model, decadal SST anomalies correlated with the November SAFZ-SST index (black contours in Fig. A3) tend to remain in the SAFZ throughout the cold season and thus stay on the poleward side of the upper-level jet (~42°N) especially in midwinter.
(a) Latitude–month section of monthly climatologies of storm-track activity measured as the 850-hPa poleward eddy heat flux (color shaded) and the 250-hPa zonal wind velocity (white contour) averaged for (150°E–180°) based on the NCEP–NCAR reanalysis, and regression coefficients with the November SAFZ-SST of SST averaged for (147.5°–160.5°E) based on the ICOADS data (black contour).
Citation: Journal of Climate 25, 1; 10.1175/JCLI-D-11-00046.1
c. Sea level pressure
The climatological mean of SLP and standard deviation of its (quasi) decadal variability based on the NCEP–NCAR reanalysis are shown in Figs. A4a and A4d for January and February, respectively. The SLP variability is overall stronger in January with a well-defined maximum just east of the climatological AL center. For each of the calendar months, the major fraction of the variability is accounted for by the two leading EOFs: the first EOF (SLP EOF-1) representing AL variability (Figs. A4b,e) and the second EOF representing a meridional dipole of anomalies similar to the North Pacific Oscillation (NPO; Figs. A4c,f; Walker and Bliss 1932; Linkin and Nigam 2008). Compared to the January situation, the AL variability extracted in SLP EOF-1 is weaker and well displaced eastward in February, contributing to the formation of dual maxima of the variability. These features are well reproduced in CFES-M, except the AL and its variability are slightly overestimated and the eastward displacement of the AL variability from January into February is less apparent (Fig. A5).
(a) Climatological mean (contoured) and standard deviation (shaded) of January-mean SLP based on the NCEP–NCAR reanalysis. (b) The leading EOF of January-mean SLP anomalies for the North Pacific domain (10°–80°N, 100°E–100°W), represented as a linear regression map of the anomalies onto the corresponding principal component time series (contoured for every 1 hPa). The standard deviation of January SLP is shown in background with shading. (c) As in (b), but for the second EOF. (d)–(f) As in (a)–(c), but for February.
Citation: Journal of Climate 25, 1; 10.1175/JCLI-D-11-00046.1
As in Fig. A4, but for SLP based on the 120-yr CFES-M integration. Note the grayscale is slightly different from that in Fig. A4.
Citation: Journal of Climate 25, 1; 10.1175/JCLI-D-11-00046.1
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 re-emergence of SST anomalies in the North Pacific Ocean. J. Climate, 12, 2419–2431.
Alexander, M. A., I. Blade, M. Newman, J. R. Lanzante, N.-C. Lau, and J. D. Scott, 2002: The atmospheric bridge: The influence of ENSO teleconnections on air–sea interaction over the global oceans. J. Climate, 15, 2205–2231.
Bond, N. A., and M. F. Cronin, 2008: Regional weather patterns during anomalous air–sea fluxes at the Kuroshio Extension Observatory (KEO). J. Climate, 21, 1680–1697.
Brayshaw, D. J., B. J. Hoskins, and M. Blackburn, 2008: The storm-track response to idealized SST perturbations in an aquaplanet GCM. J. Atmos. Sci., 65, 2842–2860.
Brayshaw, D. J., B. J. Hoskins, and M. Blackburn, 2009: The basic ingredients of the North Atlantic storm track. Part I: Land–sea contrast and orography. J. Atmos. Sci., 66, 2539–2558.
Deser, C., R. Tomas, and S. Peng, 2007: The transient atmospheric circulation response to North Atlantic SST and sea ice anomalies. J. Climate, 20, 4751–4767.
Di Lorenzo, E., and Coauthors, 2008: North Pacific Gyre Oscillation links ocean climate and ecosystem change. Geophys. Res. Lett., 35, L08607, doi:10.1029/2007GL032838.
Enomoto, T., A. Kuwano-Yoshida, N. Komori, and W. Ohfuchi, 2008: Description of AFES2: Improvements for high-resolution and coupled simulations. High Resolution Numerical Modeling of the Atmosphere and Ocean, K. Hamilton and W. Ohfuchi, Eds., Springer, 77–97.
Frankignoul, C., 1985: Sea surface temperature anomalies, planetary waves and air-sea feedback in the middle latitudes. Rev. Geophys., 23, 357–390.
Frankignoul, C., and E. Kestenare, 2002: The surface heat flux feedback. Part 1: Estimates from observations in the Atlantic and the North Pacific. Climate Dyn., 19, 633–647.
Frankignoul, C., and N. Sennéchael, 2007: Observed influence of North Pacific SST anomalies on the atmospheric circulation. J. Climate, 20, 592–606.
Frankignoul, C., A. Czaja, and B. L’Heveder, 1998: Air–sea feedback in the North Atlantic and surface boundary conditions for ocean models. J. Climate, 11, 2310–2324.
Frankignoul, C., N. Sennéchael, Y.-O. Kwon, and M. A. Alexander, 2011: Influence of the meridional shifts of the Kuroshio and the Oyashio Extensions on the atmospheric circulation. J. Climate, 24, 762–777.
Gent, P., and J. McWilliams, 1990: Isopycnal mixing in ocean circulation models. J. Phys. Oceanogr., 20, 150–155.
Held, I. M., 1983: Stationary and quasi-stationary eddies in the extratropical troposphere: Theory. Large-Scale Dynamical Processes in the Atmosphere, B. J. Hoskins and R. P. Pearce, Eds., Academic Press, 127–168.
Horel, J. D., and J. M. Wallace, 1981: Planetary-scale atmospheric phenomena associated with the Southern Oscillation. Mon. Wea. Rev., 109, 813–829.
Hoskins, B. J., and D. J. Karoly, 1981: The steady linear response of a spherical atmosphere to thermal and orographic forcing. J. Atmos. Sci., 38, 1179–1196.
Hoskins, B. J., M. E. McIntyre, and A. Robertson, 1985: On the use and significance of isentropic potential vorticity maps. Quart. J. Roy. Meteor. Soc., 111, 877–946.
Hotta, D., and H. Nakamura, 2011: On the significance of sensible heat supply from the ocean in the maintenance of mean baroclinicity along storm tracks. J. Climate, 24, 3377–3401.
Jensen, T. G., T. J. Campbell, R. A. Allard, R. J. Small, and T. A. Smith, 2011: Turbulent heat fluxes during an intense cold-air outbreak over the Kuroshio Extension region: Results from a high-resolution coupled atmosphere–ocean model. Ocean Dyn., 61, 657–674.
Joyce, T. M., Y.-O. Kwon, and L. Yu, 2009: On the relationship between synoptic wintertime atmospheric variability and path shifts in the Gulf Stream and the Kuroshio Extension. J. Climate, 22, 3177–3192.
Kalnay, E., and Coauthors, 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437–471.
Kelly, K. A., and S. Dong, 2004: Heat budget in the Gulf Stream region: The importance of heat storage and advection. J. Phys. Oceanogr., 34, 1214–1231.
Kelly, K. A., R. Small, R. Samelson, B. Qiu, T. Joyce, Y.-O. Kwon, and M. Cronin, 2010: Western boundary currents and frontal air–sea interaction: Gulf Stream and Kuroshio Extension. J. Climate, 23, 5644–5667.
Komori, N., K. Takahashi, K. Komine, T. Motoi, X. Zhang, and G. Sagawa, 2005: Description of sea-ice component of coupled ocean-sea ice model for the Earth Simulator (OIFES). J. Earth Simul., 4, 31–45.
Komori, N., A. Kuwano-Yoshida, T. Enomoto, H. Sasaki, and W. Ohfuchi, 2008a: High-resolution simulation of the global coupled atmospheric-ocean system: Description and preliminary outcomes of CFES (CGCM for the Earth Simulator). High Resolution Numerical Modelling of the Atmosphere and Ocean, K. Hamilton and W. Ohfuchi, Eds., Springer, 241–260.
Komori, N., W. Ohfuchi, B. Taguchi, H. Sasaki, and P. Klein, 2008b: Deep ocean inertia-gravity waves simulated in a high-resolution global coupled atmosphere–ocean GCM. Geophys. Res. Lett., 35, L04610, doi:10.1029/2007GL032807.
Konda, M., H. Ichikawa, H. Tomita, and M. F. Cronin, 2010: Surface heat flux variations across the Kuroshio Extension as observed by surface flux buoys. J. Climate, 23, 5206–5221.
Kushnir, Y., W. A. Robinson, I. Brade, N. M. J. Hall, S. Peng, and R. Sutton, 2002: Atmospheric GCM response to extratropical SST anomalies: Synthesis and evaluation. J. Climate, 15, 2233–2256.
Kuwano-Yoshida, A., T. Enomoto, and W. Ohfuchi, 2010a: An improved cloud scheme for climate simulations. Quart. J. Roy. Meteor. Soc., 136, 1583–1597.
Kuwano-Yoshida, A., S. Minobe, and S.-P. Xie, 2010b: Precipitation response to the Gulf Stream in an atmospheric GCM. J. Climate, 23, 3676–3698.
Kwon, Y.-O., M. Alexander, N. Bond, C. Frankignoul, H. Nakamura, B. Qiu, and L. Thompson, 2010: Role of the Gulf Stream and Kuroshio–Oyashio systems in large-scale atmosphere–ocean interaction: A review. J. Climate, 23, 3249–3281.
Kwon, Y.-O., C. Deser, and C. Cassou, 2011: Coupled atmosphere-mixed layer ocean response to ocean heat flux convergence along the Kuroshio Current Extension. Climate Dyn., 36, 2295–2312, doi:10.1007/s00382-010-0764-8.
Lau, N.-C., 1997: Interactions between global SST anomalies and the midlatitude atmospheric circulation. Bull. Amer. Meteor. Soc., 78, 21–33.
Lau, N.-C., and E. O. Holopainen, 1984: Transient eddy forcing of the time-mean flow as identified by geopotential tendencies. J. Atmos. Sci., 41, 1363–1392.
Linkin, M. E., and S. Nigam, 2008: The North Pacific Oscillation–West Pacific teleconnection pattern: Mature-phase structure and winter impacts. J. Climate, 21, 1979–1997.
Liu, Z., and L. Wu, 2004: Atmospheric response to North Pacific SST: The role of ocean–atmosphere coupling. J. Climate, 17, 1859–1882.
Liu, Z., Y. Liu, L. Wu, and R. Jacob, 2007: Seasonal and long-term atmospheric responses to reemerging North Pacific Ocean variability: A combined dynamical and statistical assessment. J. Climate, 20, 955–980.
Madden, R. A., and R. H. Jones, 2001: A quantitative estimate of the effect of aliasing in climatological time series. J. Climate, 14, 3987–3993.
Mantua, N. J., S. R. Hare, Y. Zhang, J. M. Wallace, and R. C. Francis, 1997: A Pacific interdecadal climate oscillation with impacts on salmon production. Bull. Amer. Meteor. Soc., 78, 1069–1079.
Minobe, S., A. Kuwano-Yoshida, N. Komori, S.-P. Xie, and R. Small, 2008: Influence of the Gulf Stream on the troposphere. Nature, 452, 206–209.
Minobe, S., A. Kuwano-Yoshida, M. Miyashita, H. Tokinaga, and S.-P. Xie, 2010: Atmospheric response to the Gulf Stream: Seasonal variations. J. Climate, 23, 3699–3719.
Nakamura, H., 1992: Midwinter suppression of baroclinic wave activity in the Pacific. J. Atmos. Sci., 49, 1629–1642.
Nakamura, H., 1996: Year-to-year and interdecadal variability in the activity of intraseasonal fluctuations in the Northern Hemisphere wintertime circulation. Theor. Appl. Climatol., 55, 19–32.
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, 49–72.
Nakamura, H., and T. Sampe, 2002: Trapping of synoptic-scale disturbances into the North-Pacific subtropical jet core in midwinter. Geophys. Res. Lett., 29, 1761, doi:10.1029/2002GL015535.
Nakamura, H., and A. S. Kazmin, 2003: Decadal change in the North Pacific oceanic frontal zones as revealed in ship and satellite observation. J. Geophys. Res., 108, 3078, doi:10.1029/1999JC000085.
Nakamura, H., M. Tanaka, and J. M. Wallace, 1987: Horizontal structure and energetics of Northern Hemisphere wintertime teleconnection patterns. J. Atmos. Sci., 44, 3377–3391.
Nakamura, H., G. Lin, and T. Yamagata, 1997: Decadal climate variability in the North Pacific during recent decades. Bull. Amer. Meteor. Soc., 78, 2215–2225.
Nakamura, H., T. Izumi, and T. Sampe, 2002: Interannual and decadal modulations recently observed in the Pacific storm track activity and East Asian winter monsoon. J. Climate, 15, 1855–1874.
Nakamura, H., T. Sampe, Y. Tanimoto, and A. Shimpo, 2004: Observed associations among storm tracks, jet streams, and midlatitude oceanic fronts. Earth Climate: The Ocean-Atmosphere Interaction, Geophys. Monogr., Vol. 147, Amer. Geophys. Union, 329–345.
Nakamura, H., T. Sampe, A. Goto, W. Ohfuchi, and S.-P. Xie, 2008: On the importance of midlatitude oceanic frontal zones for the mean state and dominant variability in the tropospheric circulation. Geophys. Res. Lett., 35, L15709, doi:10.1029/2008GL034010.
Nakamura, H., T. Miyasaka, Y. Kosaka, K. Takaya, and M. Honda, 2010: Northern Hemisphere extratropical tropospheric planetary waves and their low-frequency variability: Their vertical structure and interaction with transient eddies and surface thermal contrasts. Climate Dynamics: Why Does Climate Vary? Geophys. Monogr., Vol. 189, Amer. Geophys. Union, 149–179.
Nakamura, M., and S. Yamane, 2010: Dominant anomaly patterns in the near-surface baroclinicity and accompanying anomalies in the atmosphere and oceans. Part II: North Pacific basin. J. Climate, 23, 6445–6467.
Nigam, S., 2003: Teleconnections. Encyclopedia of Atmospheric Sciences, J. R. Holton, J. Pyle, and J. A. Curry, Academic Press, 2243–2269.
Nishii, K., H. Nakamura, and T. Miyasaka, 2009: Modulations in the planetary wave field induced by upward-propagating Rossby wave packets prior to stratospheric sudden warming events: A case study. Quart. J. Roy. Meteor. Soc., 135, 39–52.
Nitta, T., and S. Yamada, 1989: Recent warming of tropical sea surface temperature and its relationship to the Northern Hemisphere circulation. J. Meteor. Soc. Japan, 67, 375–382.
Nonaka, M., H. Nakamura, Y. Tanimoto, T. Kagimoto, and H. Sasaki, 2006: North Pacific decadal variability in SST and frontal structure simulated in a high-resolution OGCM. J. Climate, 19, 1970–1989.
Nonaka, M., H. Nakamura, Y. Tanimoto, T. Kagimoto, and H. Sasaki, 2008: Interannual-to-decadal variability in the Oyashio current and its influence on temperature in the subarctic frontal zone: An eddy-resolving OGCM simulation. J. Climate, 21, 6283–6303.
Nonaka, M., H. Nakamura, B. Taguchi, N. Komori, A. Kuwano-Yoshida, and K. Takaya, 2009: Localization of surface heat flux in association with a sea surface temperature front in the south Indian Ocean in a high-resolution coupled GCM. J. Climate, 22, 6515–6535.
Numaguti, A., M. Takahashi, T. Nakajima, and A. Sumi, 1997: Description of CCSRNIES Atmospheric General Circulation model. Study on the Climate System and Mass Transport by a Climate Model, A. Numaguchi, Ed., Center for Global Environmental Research, National Institute for Environmental Studies, 1–48.
Ohfuchi, W., and Coauthors, 2004: 10-km mesh meso-scale resolving simulations of the global atmosphere on the Earth Simulator—Preliminary outcomes of AFES (AGCM for the Earth Simulator). J. Earth Simul., 1, 8–34.
Ohfuchi, W., H. Sasaki, Y. Masumoto, and H. Nakamura, 2007: “Virtual” atmospheric and oceanic circulations in the Earth Simulator. Bull. Amer. Meteor. Soc., 88, 861–866.
Pacanowski, R. C., and S. M. Griffies, 2000: Mom 3.0 manual. Tech. Rep., NOAA/Geophysical Fluid Dynamics Laboratory, 680 pp.
Peng, S., and J. S. Whitaker, 1999: Mechanisms determining the atmospheric response to midlatitude SST anomalies. J. Climate, 12, 1393–1408.
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.