Abstract

The Pacific decadal oscillation (PDO) is the leading mode of sea surface temperature (SST) variability over the North Pacific (north of 20°N). Its South Pacific counterpart (south of 20°S) is the South Pacific decadal oscillation (SPDO). The effects of tropical eastern Pacific (TEP) SST forcing and internal atmospheric variability are investigated for both the PDO and SPDO using a 10-member ensemble tropical Pacific pacemaker experiment. Each member is forced by the historical radiative forcing and observed SST anomalies in the TEP region. Outside the TEP region, the ocean and atmosphere are fully coupled and freely evolve. The TEP-forced PDO (54% variance) and SPDO (46% variance) are correlated in time and exhibit a symmetric structure about the equator, driven by the Pacific–North American (PNA) and Pacific–South American teleconnections, respectively. The internal PDO resembles the TEP-forced component but is related to internal Aleutian low (AL) variability associated with the Northern Hemisphere annular mode and PNA pattern. The internal variability is locally enhanced by barotropic energy conversion in the westerly jet exit region around the Aleutians. By contrast, barotropic energy conversion is weak associated with the internal SPDO, resulting in weak geographical preference of sea level pressure variability. Therefore, the internal SPDO differs from the TEP-forced component, featuring SST anomalies along ~60°S in association with the Southern Hemisphere annular mode. The limitations on isolating the internal component from observations are discussed. Specifically, internal PDO variability appears to contribute significantly to the North Pacific regime shift in the 1940s.

1. Introduction

The Pacific decadal oscillation (PDO) is the leading mode of monthly sea surface temperature (SST) anomalies in the North Pacific (poleward of 20°N) (Mantua et al. 1997; Zhang et al. 1997). It exhibits a horseshoe-like SST pattern, with anomalies of one sign in the Kuroshio–Oyashio extension (KOE) region (Nakamura et al. 1997; Nonaka et al. 2006) and of the opposite sign along the west coast of North America. The PDO predominantly varies on decadal-to-multidecadal time scales (Minobe 1997, 1999). It impacts U.S. summertime hydrologic variability (Barlow et al. 2001), North Pacific marine ecosystems, fisheries, and global climate (Miller and Schneider 2000; Mantua and Hare 2002).

Numerous studies have investigated the PDO mechanisms. A recent consensus is that the PDO is not a single mode but a combination of multiple physical processes (Schneider and Cornuelle 2005; Newman et al. 2016). Newman et al. (2016) decomposed the PDO into two components: tropical Pacific remote forcing and local North Pacific variability, both of which are associated with atmospheric Aleutian low (AL) variability. The warm phase of El Niño–Southern Oscillation (ENSO) excites the Pacific–North American (PNA) pattern, extending across the North Pacific to North America with a center of action at the AL (Horel and Wallace 1981). The strengthened AL over the North Pacific drives the positive PDO pattern through surface heat fluxes, wind-driven mixing, and Ekman advection (Alexander et al. 2002). AL variability also originates from atmospheric internal dynamics, forcing a reddened ocean response (Frankignoul and Hasselmann 1977). SST in the KOE region can vary through dynamical adjustment of upper-ocean gyre circulations excited by anomalous wind stress curl. The SST changes in turn feed back to the atmosphere, forming a coupled ocean–atmosphere process in the North Pacific (Zhang and Delworth 2015; Newman et al. 2016). Here we use a 10-member ensemble of partially coupled simulations to separate tropical forced and internal components of the PDO. Each member run shares the common radiative forcing and prescribed tropical Pacific SST variability and differs only in the initial condition. Here internal variability refers to the deviations from the ensemble mean and is unrelated to tropical Pacific variability. This internal component includes not only the atmospheric internal variability but also the coupled air–sea interaction process discussed above.

The South Pacific counterpart of the PDO (poleward of 20°S) is termed the South Pacific decadal oscillation (SPDO; e.g., Chen and Wallace 2015). The observed SPDO shows an equatorially symmetric SST structure with the PDO. Tropical Pacific forcing is suggested to be important for the SPDO (Shakun and Shaman 2009). The tropically excited Pacific–South American (PSA) pattern (Karoly 1989) is important in forcing the SPDO, mainly through the surface heat fluxes (Okumura 2013). It remains unclear if there is an SPDO component independent of tropical Pacific variability.

Previous studies have attempted to separate the PDO into tropical Pacific-forced and local internal variability components by means of statistical methods (e.g., Zhang et al. 1996; An and Wang 2005) and a hierarchy of coupled models. Ocean mixed layer models were used to study the ENSO-forced North Pacific variability, but they consider only thermodynamic coupling processes in the North Pacific (Alexander et al. 2002; Di Lorenzo et al. 2015). Partial coupling experiments were conducted to study the intrinsic North Pacific variability by prescribing the climatological SST in the tropics but with the ocean and atmosphere fully coupled in the extratropics (Zhang and Delworth 2015). Some studies investigated the impact of atmospheric noise on the North Pacific SST variability using an interactive ensemble coupling strategy (Kirtman and Shukla 2002; Yeh et al. 2007). Wang et al. (2012) compared the partial coupling experiment with a fully coupled run to investigate the influence of ENSO on the PDO. All the above studies have suggested that the PDO is made up of the tropical Pacific forced and intrinsic North Pacific components, both related to AL variability. However, not much attention has been given to possible differences in the atmospheric teleconnection pattern between the tropical forced and internal PDO. Even fewer studies examine the tropical-forced and internal components for the SPDO. It is unclear whether the SPDO components are similar in spatial structure to each other.

The present study uses a 10-member ensemble of the tropical Pacific pacemaker experiment (Kosaka and Xie 2016) to investigate the tropical eastern Pacific (TEP)-forced and internal components of the PDO and SPDO, with a focus on the associated atmospheric teleconnection patterns. We identify distinct features between the TEP-forced and internal components of the SPDO. We suggest that the small degrees of freedom limit our ability to isolate the internal component from observations. The comparison with the pacemaker experiment indicates that the PDO regime shift in the 1940s (Minobe 1997, 1999) is related to the internal variability.

The rest of the paper is organized as follows. Section 2 describes the pacemaker experiment, the method of isolating the TEP-forced and internal components, and the observational dataset used in this study. Section 3 compares the simulated PDO and SPDO variabilities by the pacemaker experiment with observations. Section 4 investigates the distinct characteristics between the TEP-forced and internal components for both the PDO and SPDO. Furthermore, barotropic energy conversion is diagnosed for atmospheric anomalies to explain the similarity (distinction) of the two components for the PDO (SPDO). Section 5 discusses the limitations on isolating the internal component from observations and identifies the relative contribution of TEP-forced and internal components to the PDO regime shifts. Section 6 is a summary with some discussion.

2. Data and methods

A long-term tropical Pacific pacemaker experiment, named Pacific Ocean–Global Atmosphere (POGA) (Kosaka and Xie 2016), is used to isolate the TEP-forced and internal components based on the Geophysical Fluid Dynamics Laboratory Coupled Model, version 2.1 (CM2.1; Delworth et al. 2006). The POGA ensemble has 10-member runs, each of which is forced by the identical historical radiative forcing and observed SST anomalies in the TEP domain (from 180° to the American coast, 15°S–15°N, with a 5° buffer zone north, south, and west of the domain), but with different atmospheric and oceanic initial conditions. Outside the TEP-restoring region, the ocean and atmosphere are fully coupled and freely evolve. The horizontal resolution of the ocean is approximately 1° longitude × 1° latitude, with finer meridional resolution equatorward of 30°S and 30°N (⅓° at the equator). The atmospheric resolution is about 2.5° longitude × 2° latitude. The time period of 1890–2014 is analyzed. We refer details on the model framework to Kosaka and Xie (2013, 2016).

To obtain the TEP-forced component, we calculate the ensemble mean of 10 POGA members and then subtract the ensemble mean of 20-member historical runs, each of which is forced by the historical and representative concentration pathway 4.5 (RCP4.5) radiative forcing and based on CM2.1, to remove the radiatively forced response. The internal component is obtained by subtracting the POGA ensemble mean from each POGA member. Note that the internal component here refers to variability independent of the prescribed TEP SST. Although 10 members may be not sufficient to extract the TEP-forced and internal components in a single event, the PDO and SPDO are obtained from empirical orthogonal function (EOF) analysis, which also exploits the degrees of freedom in the time dimension. Our analyses suggest that 10 members are sufficient for statistically isolating the TEP-forced and internal variability of the PDO and SPDO [see the strong (weak) connection of SST anomalies with the tropical Pacific in Figs. 3a and 5a (Figs. 3d and 5d)]. In addition to CM2.1 POGA, we also examine the POGA results based on the Community Earth System Model, version 1 (CESM POGA; Deser et al. 2017; see the  appendix).

We use the Extended Reconstructed Sea Surface Temperature product, version 5 (ERSSTv5; Huang et al. 2017). Other SST datasets show similar results. The analysis period begins from 1900 (until 2014) because of large uncertainties in observations prior to the early twentieth century. We also use the Southern Oscillation index (SOI) from 1890 to 2014 to represent long-term tropical Pacific variability. This index is based on annual standardization, which has shown to maximize the signal of large-scale variability (Trenberth 1984).

We focus on the PDO and SPDO variabilities on interannual-to-multidecadal time scales. In the internal component, we subtract climatology in each ocean variable of each member to remove initial perturbations that persist throughout the integration in the Southern Ocean. The annual mean centered on the local wintertime is calculated, from July to the following June in the Northern Hemisphere (NH), and January–December in the Southern Hemisphere (SH). The reemergence process occurs in local winter and is important for the low-frequency PDO and SPDO (Newman et al. 2003; Shakun and Shaman 2009).

To isolate the TEP-forced covariability mode, we conduct joint EOF analysis of annual-mean (July–June) SST anomalies between the North and South Pacific. The analyses based on the annual mean of January–December are similar (not shown). SST anomalies are normalized by the spatially averaged SST standard deviation in each basin. This normalization method assigns equal weight on SST variability between the North and South Pacific.

3. Evaluating POGA simulation

We first evaluate the POGA’s skill in simulating the spatial–temporal structures of the observed PDO and SPDO. Figure 1 shows the spatial patterns of the PDO and SPDO along with the corresponding SLP anomalies. In general, POGA well simulates the PDO, with a horseshoe-like SST pattern associated with strengthened AL variability, although some biases exist, such as a slight westward shift in the main center of the AL anomaly, stronger SST anomalies in the KOE region, and weaker SST anomalies in the subtropical northeastern Pacific (Fig. 1a vs Fig. 1b). POGA also well simulates the SPDO pattern, characterized by SST cooling below the South Pacific convergence zone (SPCZ) region and around New Zealand and SST warming in the southeastern South Pacific and Southern Ocean. It is associated with low pressure anomalies in the midlatitudes, with maxima south of New Zealand, and high pressure anomalies centered around 65°S, 120°W (Fig. 1c vs Fig. 1d).

Fig. 1.

Spatial patterns of the (a),(b) PDO and (c),(d) SPDO in POGA and observations. Patterns are the regressions of SST (shading; °C) and SLP (contours with 0.2-hPa interval; zero line thickened) anomalies against the normalized PC1 of SST. (top) POGA and (bottom) observations. EOF in POGA is performed based on the tandem 10-member runs to derive one EOF pattern and corresponding 10 PC1s. The explained variance fraction is shown in parentheses in each panel based on annual means of July–June in the NH and January–December in the SH.

Fig. 1.

Spatial patterns of the (a),(b) PDO and (c),(d) SPDO in POGA and observations. Patterns are the regressions of SST (shading; °C) and SLP (contours with 0.2-hPa interval; zero line thickened) anomalies against the normalized PC1 of SST. (top) POGA and (bottom) observations. EOF in POGA is performed based on the tandem 10-member runs to derive one EOF pattern and corresponding 10 PC1s. The explained variance fraction is shown in parentheses in each panel based on annual means of July–June in the NH and January–December in the SH.

We further compare the temporal characteristics of the PDO and SPDO between POGA and observations (Fig. 2). Overall, POGA well simulates the PDO (Fig. 2a) and SPDO spectra (Fig. 2b), although the latter of which shows strong interannual and weak decadal-to-multidecadal variability. The model also captures the correlation between ENSO and the PDO or SPDO, albeit biased somewhat in lead/lag (Figs. 2c,d). The correlation between the PDO and SPDO in POGA is lower than in observations (Fig. 2e). This may arise from too strong internal variability in POGA (see further discussion in section 5a).

Fig. 2.

Power spectra of the (a) PDO and (b) SPDO and cross correlations between the (c) PDO and ENSO, (d) SPDO and ENSO, and (e) PDO and SPDO. The ENSO index is winter SST anomalies (November–January mean for correlation with the PDO; June–August mean with the SPDO) averaged in the Niño-3.4 region (5°N–5°S, 170°–120°W). Each thin red line denotes the analysis based on each run. The bold red lines denote the mean of the analyses of each run. Black lines denote observations. The texts labeled in (c)–(e) indicate that one leads the other [(e.g., “PDO” in (c) denotes that the PDO leads ENSO].

Fig. 2.

Power spectra of the (a) PDO and (b) SPDO and cross correlations between the (c) PDO and ENSO, (d) SPDO and ENSO, and (e) PDO and SPDO. The ENSO index is winter SST anomalies (November–January mean for correlation with the PDO; June–August mean with the SPDO) averaged in the Niño-3.4 region (5°N–5°S, 170°–120°W). Each thin red line denotes the analysis based on each run. The bold red lines denote the mean of the analyses of each run. Black lines denote observations. The texts labeled in (c)–(e) indicate that one leads the other [(e.g., “PDO” in (c) denotes that the PDO leads ENSO].

Fig. 3.

TEP-forced and internal components of the PDO. Regression patterns of (a),(d) SST (°C); (b),(e) SLP (hPa); and (c),(f) Z200 (m) anomalies against the normalized PC1 of SST for both the TEP-forced and internal components. (top) TEP-forced and (bottom) internal components are indicated. The EOF domain is the North Pacific (20°–70°N, 120°E–100°W; green rectangular box) based on the annual mean of July–June. Note that the EOF of internal SST is performed based on the tandem 10-member internal components to derive one EOF pattern and the corresponding 10 PC1s. The regression patterns of SLP (Z200) are shown poleward of 20°N (equator). The explained variance fraction for the TEP-forced and internal components are shown in parentheses in (a) and (d), respectively.

Fig. 3.

TEP-forced and internal components of the PDO. Regression patterns of (a),(d) SST (°C); (b),(e) SLP (hPa); and (c),(f) Z200 (m) anomalies against the normalized PC1 of SST for both the TEP-forced and internal components. (top) TEP-forced and (bottom) internal components are indicated. The EOF domain is the North Pacific (20°–70°N, 120°E–100°W; green rectangular box) based on the annual mean of July–June. Note that the EOF of internal SST is performed based on the tandem 10-member internal components to derive one EOF pattern and the corresponding 10 PC1s. The regression patterns of SLP (Z200) are shown poleward of 20°N (equator). The explained variance fraction for the TEP-forced and internal components are shown in parentheses in (a) and (d), respectively.

Nevertheless, POGA well simulates both the spatial and temporal characteristics of the observed PDO and SPDO. Thus, we can further investigate the TEP-forced and internal components based on POGA.

4. TEP-forced versus internal variability

This section explores the characteristics of the TEP-forced and internal components separately for the PDO and SPDO. We identify distinct spatial patterns between these two components, especially for the SPDO.

a. PDO

Figure 3 shows the TEP-forced and internal components of the PDO. Generally, both the TEP-forced and internal PDOs exhibit a horseshoe-like SST pattern in the North Pacific, although the internal component features strong SST cooling over the KOE region (Fig. 3a vs Fig. 3d). The TEP-forced PDO is closely related to the ENSO-like SST pattern in the tropics, while the internal PDO is confined to the extratropical North Pacific. Both the TEP-forced and internal PDOs are associated with the anomalous AL over the North Pacific (Figs. 3b,e), consistent with previous findings (e.g., Newman et al. 2016). However, these two components are related to different NH atmospheric teleconnections. The TEP-forced PDO is driven by the PNA pattern (Wallace and Gutzler 1981) excited from the tropical Pacific (Fig. 3c), which differs from the tropical/Northern Hemisphere pattern in Barnston and Livezey (1987). From the SLP field, the internal PDO is not only associated with the anomalous AL, but also with an anomalously high SLP system over the Arctic polar cap region (Fig. 3e). This SLP pattern is reminiscent of the negative phase of the Northern Hemisphere annular mode (NAM), with the center of action over the entire Arctic Ocean surrounded by a zonal ring of the opposite sign centered near 45°N (Thompson and Wallace 1998, 2000). From the 200-hPa geopotential height (Z200) anomaly pattern, the internal PDO appears to be a hybrid of the NAM and PNA pattern (Fig. 3f).

To examine whether the atmospheric teleconnection patterns associated with the internal PDO are the dominant modes of internal atmospheric circulation in the NH, we obtain EOFs of the internal component of the annual-mean (July–June) SLP anomalies poleward of 20°N. Figures 4a and 4b show the regressions of SLP anomalies against the normalized principal components (PCs). The first EOF (SLP1) is the NAM (Fig. 4a), and the second EOF (SLP2) shows an anomalously low pressure over the Aleutian Islands (Fig. 4b). The SLP2 corresponds to the PNA pattern, which is the second EOF mode of the NH atmospheric circulation anomaly in observations, as reported by Wallace and Thompson (2002). Spatial correlations of SLP1 and SLP2 with the SLP pattern related to the internal PDO (Fig. 3e) demonstrate that the internal PDO is primarily associated with the internal NAM teleconnection, and secondarily with the internal PNA teleconnection (Table 1).

Fig. 4.

Leading modes of hemispheric internal SLP variability. Regression patterns of internal SLP anomalies (hPa) against normalized internal (a) NH SLP PC1, (b) NH SLP PC2, and (c) SH SLP PC1. EOFs and regressions are performed poleward of 20° latitude. The SLP associated with EOF1 and EOF2 are labeled as SLP1 and SLP2, respectively, based on annual means of July–June in the NH and January–December in the SH. The explained variance fraction is shown in parentheses in each panel.

Fig. 4.

Leading modes of hemispheric internal SLP variability. Regression patterns of internal SLP anomalies (hPa) against normalized internal (a) NH SLP PC1, (b) NH SLP PC2, and (c) SH SLP PC1. EOFs and regressions are performed poleward of 20° latitude. The SLP associated with EOF1 and EOF2 are labeled as SLP1 and SLP2, respectively, based on annual means of July–June in the NH and January–December in the SH. The explained variance fraction is shown in parentheses in each panel.

Table 1.

Spatial correlations between the SLP patterns associated with the internal PDO and SPDO (Figs. 3e, 5e) and the corresponding internal SLP regression patterns in Fig. 4. The SH SLP2 is also included for computing the spatial correlation with the internal SPDO SLP (not shown in Fig. 4).

Spatial correlations between the SLP patterns associated with the internal PDO and SPDO (Figs. 3e, 5e) and the corresponding internal SLP regression patterns in Fig. 4. The SH SLP2 is also included for computing the spatial correlation with the internal SPDO SLP (not shown in Fig. 4).
Spatial correlations between the SLP patterns associated with the internal PDO and SPDO (Figs. 3e, 5e) and the corresponding internal SLP regression patterns in Fig. 4. The SH SLP2 is also included for computing the spatial correlation with the internal SPDO SLP (not shown in Fig. 4).
Fig. 5.

As in Fig. 3, but for the SPDO. Based on the annual mean of January–December. The EOF domain is in the South Pacific (70°–20°S, 120°E–70°W; green rectangular box).

Fig. 5.

As in Fig. 3, but for the SPDO. Based on the annual mean of January–December. The EOF domain is in the South Pacific (70°–20°S, 120°E–70°W; green rectangular box).

b. SPDO

Figure 5 shows the TEP-forced and internal components of the SPDO. The TEP-forced SPDO connects to the ENSO-like variability and exhibits a rather symmetric structure about the equator with the TEP-forced PDO (Figs. 3a, 5a). It is characterized by an SST cooling below the SPCZ region surrounded by SST warming on both the equatorward and poleward sides (Fig. 5a). Analogous to the TEP-forced PDO, the TEP-forced SPDO is driven by the PSA pattern (Mo and Ghil 1987) excited from the tropical Pacific (Figs. 5b,c).

The internal SPDO is very different from the TEP-forced counterpart (Fig. 5a vs Fig. 5d). Notably, the internal SPDO exhibits strong SST warming along ~60°S with weak cooling around New Zealand, but without the cooling in the SPCZ region (Fig. 5d). The internal SPDO is not associated with the PSA pattern but with the Southern Hemisphere annular mode (SAM), which features a primary center of action over Antarctica that is regionally enhanced over the Amundsen Sea and western Antarctica, accompanied by a zonal ring of opposite sign along ~45°S (Figs. 5e,f) (Thompson and Wallace 2000). The SAM is the leading mode of internal SLP variability in the SH winter (Fig. 4c). High spatial correlation with the internal SAM (Fig. 5e) indicates a strong association with the internal SPDO (Table 1).

We evaluate the TEP-forced and internal components by projecting POGA anomalies onto the PDO (Fig. 1a) and SPDO (Fig. 1c) patterns. The variance of the PDO and SPDO explained by TEP forcing is 54% and 46%, respectively. This result suggests that both the TEP-forced and internal components are important for the PDO and SPDO variability.

c. Atmospheric forcing of the PDO and SPDO

This subsection examines the mechanism for SST variability in the PDO and SPDO. As both surface turbulent heat flux and ocean Ekman advection are related to surface wind variability, we perform 1-yr-lead regressions of the sum of these two terms against PC1s of the PDO and SPDO (contours in Fig. 6). Ekman advection is given by , where is the specific heat of seawater, is the Coriolis parameter, is the unit normal vector, and is the surface wind stress that is converted from the surface wind using quadratic stress law. Our calculation suggests that the Ekman advection is weak compared to heat flux. Thus, Fig. 6 can be mainly regarded as the surface heat flux effect.

Fig. 6.

The 1-yr-lead regressions of the sum of surface turbulent heat flux (latent and sensible heat fluxes) and ocean Ekman advection anomalies (contours with 1 W m−2 interval; positive downward; zero line thickened) against PC1s of the (a),(b) PDO and (c),(d) SPDO. (top) TEP-forced and (bottom) internal components. Simultaneous SST anomalies (°C) regressed onto PDO (Figs. 3a,d) and SPDO (Figs. 5a,d) are superimposed as shadings.

Fig. 6.

The 1-yr-lead regressions of the sum of surface turbulent heat flux (latent and sensible heat fluxes) and ocean Ekman advection anomalies (contours with 1 W m−2 interval; positive downward; zero line thickened) against PC1s of the (a),(b) PDO and (c),(d) SPDO. (top) TEP-forced and (bottom) internal components. Simultaneous SST anomalies (°C) regressed onto PDO (Figs. 3a,d) and SPDO (Figs. 5a,d) are superimposed as shadings.

For both the TEP-forced and internal PDOs (Figs. 6a,b), the overall SST pattern is consistent with surface heat flux forcing and Ekman advection. In the KOE, surface heat flux and Ekman advection damp SST anomalies, indicating that ocean dynamics plays a primary role. In the central-to-eastern North Pacific on the boundary between the SST cooling and warming regions, vertical entrainment due to the intensified wind speed (not shown) might also contribute to the SST cooling. For the TEP-forced and internal SPDOs (Figs. 6c,d), SST anomalies are basically driven by the PSA pattern and SAM, respectively. Thus, SST anomalies in the open ocean are driven by atmospheric wind variability through surface processes (heat flux and Ekman advection). In section 4d, we address the question of why the atmospheric circulation patterns associated with the TEP-forced and internal SPDOs are distinct but similar for the PDOs.

d. Barotropic energy conversion

For the PDO, both components are related to similar AL anomaly patterns in SLP (Fig. 7a). The SLP anomalies in the midlatitudes associated with the TEP-forced SPDO are geographically anchored south of Australia (negative value of shading in Fig. 7b). However, there is no geographical preference for SLP pattern in the midlatitudes associated with the internal SPDO (negative value of contour in Fig. 7b).

Fig. 7.

TEP-forced (shading; hPa) and internal (contours with 0.2-hPa interval; zero line thickened) components of the SLP pattern associated with the (a) PDO and (b) SPDO. SLP patterns in (a) are from Figs. 3b and 3e in the North Pacific and those in (b) are from Figs. 5b and 5e in the South Pacific.

Fig. 7.

TEP-forced (shading; hPa) and internal (contours with 0.2-hPa interval; zero line thickened) components of the SLP pattern associated with the (a) PDO and (b) SPDO. SLP patterns in (a) are from Figs. 3b and 3e in the North Pacific and those in (b) are from Figs. 5b and 5e in the South Pacific.

Over the North Pacific, previous studies suggested that the stationary AL variability is due partly to barotropic energy conversion (Simmons et al. 1983; Trenberth et al. 1998). It refers to a process through which disturbances can gain kinetic energy from the climatological zonally varying flow. This preferentially occurs in the westerly jet exit region as west of the Aleutians. Here we calculate the local barotropic energy conversion (CK) associated with the internal PDO and SPDO in Cartesian coordinates, following the formulation (cf. Simmons et al. 1983):

 
formula

Here is the zonal velocity of the basic state; and and are the zonal and meridional velocity anomalies, respectively, which are obtained by regression against normalized PC1 of the internal PDO or SPDO. Positive CK indicates the conversion of kinetic energy from the mean flow to atmospheric circulation perturbations. The calculation is based on the wintertime-centered annual mean at the 200-hPa level.

Over the North Pacific, positive CK is large in the westerly jet exit region (Fig. 8a), indicating that strong barotropic energy conversion anchors AL variability, which in turn induces the internal PDO. Because the AL anomaly pattern is similar, the TEP-forced and internal components of the PDO both feature a horseshoe-like SST pattern (Figs. 3a,d). By contrast, CK over the South Pacific is weak despite a comparable magnitude of the circulation anomaly (Fig. 8b), resulting in a weak zonal preference for the atmospheric circulation anomaly associated with the internal SPDO. Consequently, the internal SPDO features pronounced SST warming along ~60°S, associated with the negative phase of the atmospheric internal SAM (Figs. 5d–f). The TEP-forced component shows strong zonal variations, with low SLP anomalies of the PSA pattern (shading in Fig. 7b) driving the SST cooling in the SPCZ region and SST warming in the northeast–southwest direction on the poleward side (Fig. 5a).

Fig. 8.

The CK patterns (shading; ×10−7 m2 s−3) associated with the (a) internal PDO and (b) internal SPDO. Green contours denote climatology of annual mean (July–June in the NH; January–December in the SH) zonal wind (contours with 5 m s−1 interval; only ≥30 m s−1 are shown). Z200 patterns (black contours with 2-m interval; zero line thickened) in (a) are from Fig. 3f in the North Pacific and in (b) are from Fig. 5f in the South Pacific.

Fig. 8.

The CK patterns (shading; ×10−7 m2 s−3) associated with the (a) internal PDO and (b) internal SPDO. Green contours denote climatology of annual mean (July–June in the NH; January–December in the SH) zonal wind (contours with 5 m s−1 interval; only ≥30 m s−1 are shown). Z200 patterns (black contours with 2-m interval; zero line thickened) in (a) are from Fig. 3f in the North Pacific and in (b) are from Fig. 5f in the South Pacific.

Ultimately, the westerly jets vary more strongly in the Northern than in the Southern Hemisphere. The resultant interhemispheric difference in barotropic energy conversion explains why the internal PDO resembles the TEP-forced component while the two components of the SPDO are markedly distinct. The results based on the CESM POGA are similar (see the  appendix).

5. Discussion: Observed internal variability

a. Limitations on isolating the internal component

In the POGA ensemble mean, the temporal correlation is moderate between the PDO and SPDO (Table 2). This could be because they are sensitive to details in the structure and seasonality of TEP SST or because of remaining internal variability in the 10-member average. Alternatively, we can perform joint SST EOF between the North and South Pacific (see the method in section 2). The leading PC captures interhemispheric coherent variability and is highly correlated with both the TEP-forced PDO and SPDO (Table 2). This leading EOF resembles the so-called interdecadal Pacific oscillation (IPO; Power et al. 1999; Henley et al. 2015).

Table 2.

Correlation coefficients among the PC1s of the PDO, SPDO, and joint EOF of annual mean (July–June) SST anomalies in the POGA ensemble mean. Correlations among the 10-yr low-pass-filtered PC1s are listed in parentheses.

Correlation coefficients among the PC1s of the PDO, SPDO, and joint EOF of annual mean (July–June) SST anomalies in the POGA ensemble mean. Correlations among the 10-yr low-pass-filtered PC1s are listed in parentheses.
Correlation coefficients among the PC1s of the PDO, SPDO, and joint EOF of annual mean (July–June) SST anomalies in the POGA ensemble mean. Correlations among the 10-yr low-pass-filtered PC1s are listed in parentheses.

We apply the joint EOF method over the extratropical North and South Pacific basins to observations. Figure 9 shows the regressions of annual-mean (July–June) SST anomalies against the PCs of joint EOF, along with the POGA ensemble mean results. The results based on the January–December mean are similar (not shown). The joint EOF1 is highly correlated spatially and temporally between observations and the POGA ensemble mean (Figs. 9a and 9d; Table 3). This mode explains 38% and 24% of the variance in the POGA ensemble mean and observations, respectively. The second and higher modes disagree in spatial pattern between the POGA ensemble mean and observations (Figs. 9b,c vs Figs. 9e,f), indicating a limitation on reliably isolating higher modes by removing the TEP-forced variability over the North and South Pacific due to the small degrees of freedom in the instrumental record.

Fig. 9.

Regressions of annual-mean (July–June) SST anomalies (°C) in the (a)–(c) POGA ensemble mean and (d)–(f) observations against their individual PC1, PC2, and PC3 of the joint SST EOF. Joint EOF domains are in the North and South Pacific (green rectangular boxes). The explained variance of the total variance for the POGA ensemble mean and observations are listed in parentheses in each panel.

Fig. 9.

Regressions of annual-mean (July–June) SST anomalies (°C) in the (a)–(c) POGA ensemble mean and (d)–(f) observations against their individual PC1, PC2, and PC3 of the joint SST EOF. Joint EOF domains are in the North and South Pacific (green rectangular boxes). The explained variance of the total variance for the POGA ensemble mean and observations are listed in parentheses in each panel.

Table 3.

Correlation coefficients of the regression patterns in Fig. 9 and the PCs (parentheses) of the joint EOF between the POGA ensemble mean and ERSSTv5. “Reg” denotes the regression pattern.

Correlation coefficients of the regression patterns in Fig. 9 and the PCs (parentheses) of the joint EOF between the POGA ensemble mean and ERSSTv5. “Reg” denotes the regression pattern.
Correlation coefficients of the regression patterns in Fig. 9 and the PCs (parentheses) of the joint EOF between the POGA ensemble mean and ERSSTv5. “Reg” denotes the regression pattern.

By design, POGA does not address the origin of TEP variability. The extratropical anomalies propagating toward the tropical Pacific are eventually restored by the observed TEP SST anomalies. The internal component might induce the TEP variability via the tropical–extratropical interactions. For example, the extratropical Pacific meridional mode can propagate into the tropics via the seasonal footprinting mechanism (Vimont et al. 2001, 2003; Di Lorenzo et al. 2015; Ma et al. 2017). The South Pacific is suggested to be coupled with the tropical Pacific on decadal time scales via the oceanic tunnel (Luo and Yamagata 2001). Okumura (2013) pointed out that atmospheric internal variability over the subtropical South Pacific can thermodynamically force the tropical Pacific.

b. PDO regime shifts

The PDO features marked regime shifts in the 1940s, the late 1970s, and the late 1990s (Minobe 1997, 1999). The joint EOF1 of the POGA ensemble mean suggests that the covariability between the PDO and SPDO is dominated by TEP forcing (Fig. 9a). Indeed, temporal correlation between the internal PDO and SPDO is nearly zero (r = −0.01). In observations, the PDO and SPDO exhibit strong covariability in the late 1970s and 1990s, and the POGA ensemble mean captures these regime shifts (Figs. 10a, A3a). This suggests that the 1970s and 1990s regime shifts are dominated by TEP forcing, consistent with previous analyses (e.g., Nitta and Yamada 1989; Kosaka and Xie 2016). By contrast, the PDO–SPDO covariability is weak during the 1940s regime shift (black line in Fig. 10a vs Fig. 10b). The PDO exhibits a regime shift but the POGA ensemble mean fails to capture it (Figs. 10a, A3a). As a caveat, the TEP SST itself is somewhat uncertain during World War II because of sparse observations (Thompson et al. 2008) as discussed in Kosaka and Xie (2016). The SOI, which reflects large-scale tropical Pacific variability, exhibits a consistent phase shift with the PDO, but over a much shorter period (Fig. 10a). This suggests that the internal variability plays an important role in the 1940s regime shift of the PDO.

Fig. 10.

The 10-yr low-pass-filtered annual-mean (a) PDO PC1 and (b) SPDO PC1 in the POGA ensemble mean (red) and observations (black). The 10-yr low-pass-filtered annual-mean (July–June) SOI (normalized; sign reversed) is superimposed in (a). The correlation coefficient of the 10-yr low-pass-filtered PC1 between the POGA ensemble mean and observations is marked in the bottom left in each panel.

Fig. 10.

The 10-yr low-pass-filtered annual-mean (a) PDO PC1 and (b) SPDO PC1 in the POGA ensemble mean (red) and observations (black). The 10-yr low-pass-filtered annual-mean (July–June) SOI (normalized; sign reversed) is superimposed in (a). The correlation coefficient of the 10-yr low-pass-filtered PC1 between the POGA ensemble mean and observations is marked in the bottom left in each panel.

6. Summary

We have used a 10-member ensemble of a long-term tropical Pacific pacemaker experiment and showed that the TEP-forced and internal components are distinct for both the PDO and SPDO. The TEP-forced PDO and SPDO exhibit an interhemispherically symmetric structure about the equator, forced by the PNA and PSA patterns excited from the tropical Pacific, respectively. The internal PDO resembles the TEP-forced component but is related to internal AL variability associated with the NAM and PNA pattern. The internal variability is locally enhanced by barotropic energy conversion in the westerly jet exit region around the Aleutians, driving a horseshoe-like SST pattern similar to the TEP-forced PDO. By contrast, barotropic energy conversion is rather weak and ineffective to form the preferred spatial pattern associated with the internal SPDO, resulting in weak geographical preference of SLP variability. The resultant SST anomaly pattern differs from the TEP-forced SPDO, featuring pronounced SST variability along ~60°S in association with the SAM teleconnection. These atmospheric patterns associated with the internal PDO and SPDO are the dominant modes of the internal atmospheric variability in each hemisphere. The more efficient barotropic energy conversion over the North than South Pacific explains why the internal PDO resembles the TEP-forced component while the internal and TEP-forced SPDOs are markedly distinct. The internal variability in the PDO appears to contribute significantly to the North Pacific climate regime shift in the 1940s, although short observations limit the confidence in the attribution.

As discussed in the introduction, coupled ocean–atmosphere interaction could be important for the PDO. The atmospheric resolution of CM2.1 is insufficient to adequately represent the crucial mesoscale ocean–atmosphere interaction in the KOE region (Ma et al. 2016). Future research needs to examine detailed physical processes of the PDO and SPDO—both the TEP-forced and internal components—with higher-resolution POGA experiments.

Acknowledgments

The ERSST, version 5, product is provided by the NOAA/Earth System Research Laboratory (www.esrl.noaa.gov/psd/data/gridded/data.noaa.ersst.v5.html), and the SOI time series is provided by Climate and Global Dynamics Division, NCAR (www.cgd.ucar.edu/cas/catalog/climind/soiAnnual.html). CESM1 historical, RCP8.5, and POGA runs have been obtained from the Earth System Grid (www.earthsystemgrid.org). We sincerely thank the three anonymous reviewers for their constructive comments and suggestions. This work is supported by the Natural Science Foundation of China (41490640 and 41490641), China Scholarship Council (201706330016), U.S. National Science Foundation (AGS 1637450), Japanese Ministry of Education, Culture, Sports, Science and Technology (the Integrated Research Program for Advancing Climate Models and Arctic Challenge for Sustainability), the Japan Society for the Promotion of Science (15H05466 and 18H01278), and the Japan Science and Technology Agency (Belmont Forum CRA “InterDec”).

APPENDIX

The TEP-Forced and Internal Components of the PDO and SPDO in CESM POGA

We also use 10-member runs of CESM POGA to investigate the characteristics of the TEP-forced and internal components of the PDO and SPDO. The prescribed SST dataset is identical to that in CM2.1 POGA. The horizontal resolutions of both atmosphere and ocean are approximately 1° longitude × 1° latitude. The time period of the experiment is from 1920 to 2013. We use CESM 40-member historical and RCP8.5 runs to remove the response to the radiative forcing.

Both the TEP-forced and internal PDOs exhibit similar horseshoe-like SST patterns and are associated with AL variability (Fig. A1), consistent with the results based on the CM2.1 POGA. The TEP-forced SPDO is also characterized by an interhemispherically symmetric SST pattern with the TEP-forced PDO and is driven by the PSA pattern (Figs. A2a,b). As in CM2.1, the internal SPDO in CESM1 markedly differs from the TEP-forced SPDO, without SST cooling in the SPCZ region (Fig. A2c) and is also associated with the zonally symmetric SAM teleconnection (Fig. A2d). Likewise, the internal SPDO is notably distinct from the TEP-forced SPDO, associated with little geographical preference of SLP anomalies because of the weak barotropic energy conversion. Regarding the observed PDO regime shifts, the CESM POGA ensemble mean captures the late 1970s and 1990s regime shifts but not the one in the 1940s (Fig. A3a). This result is consistent with CM2.1 POGA, further suggesting that the internal variability in the PDO partly contributes to the 1940s regime shift.

Fig. A1.

As in Fig. 3, but for the CESM POGA. The regression patterns of Z200 field are omitted. Note that the internal PDO in CESM POGA is extracted as in the second EOF mode.

Fig. A1.

As in Fig. 3, but for the CESM POGA. The regression patterns of Z200 field are omitted. Note that the internal PDO in CESM POGA is extracted as in the second EOF mode.

Fig. A2.

As in Fig. A1, but for the SPDO.

Fig. A2.

As in Fig. A1, but for the SPDO.

Fig. A3.

As in Fig. 10, but for the comparison of the PC1 between the CESM POGA ensemble mean and observations. The SOI time series is not superimposed.

Fig. A3.

As in Fig. 10, but for the comparison of the PC1 between the CESM POGA ensemble mean and observations. The SOI time series is not superimposed.

REFERENCES

REFERENCES
Alexander
,
M. A.
,
I.
Bladé
,
M.
Newman
,
J.
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
, https://doi.org/10.1175/1520-0442(2002)015<2205:TABTIO>2.0.CO;2.
An
,
S.-I.
, and
B.
Wang
,
2005
:
The forced and intrinsic low-frequency modes in the North Pacific
.
J. Climate
,
18
,
876
885
, https://doi.org/10.1175/JCLI-3298.1.
Barlow
,
M.
,
S.
Nigam
, and
E. H.
Berbery
,
2001
:
ENSO, Pacific decadal variability, and U.S. summertime precipitation, drought, and stream flow
.
J. Climate
,
14
,
2105
2128
, https://doi.org/10.1175/1520-0442(2001)014<2105:EPDVAU>2.0.CO;2.
Barnston
,
A. G.
, and
R. E.
Livezey
,
1987
:
Classification, seasonality and persistence of low-frequency atmospheric circulation patterns
.
Mon. Wea. Rev.
,
115
,
1083
1126
, https://doi.org/10.1175/1520-0493(1987)115<1083:CSAPOL>2.0.CO;2.
Chen
,
X.
, and
J. M.
Wallace
,
2015
:
ENSO-like variability: 1900–2013
.
J. Climate
,
28
,
9623
9641
, https://doi.org/10.1175/JCLI-D-15-0322.1.
Delworth
,
T.
, and Coauthors
,
2006
:
GFDL’s CM2 global coupled climate models. Part I: Formulation and simulation characteristics
.
J. Climate
,
19
,
643
674
, https://doi.org/10.1175/JCLI3629.1.
Deser
,
C.
,
R.
Guo
, and
F.
Lehner
,
2017
:
The relative contributions of tropical Pacific sea surface temperatures and atmospheric internal variability to the recent global warming hiatus
.
Geophys. Res. Lett.
,
44
,
7945
7954
, https://doi.org/10.1002/2017GL074273.
Di Lorenzo
,
E.
,
G.
Liguori
,
N.
Schneider
,
J. C.
Furtado
,
B. T.
Anderson
, and
M. A.
Alexander
,
2015
:
ENSO and meridional modes: A null hypothesis for Pacific climate variability
.
Geophys. Res. Lett.
,
42
,
9440
9448
, https://doi.org/10.1002/2015GL066281.
Frankignoul
,
C.
, and
K.
Hasselmann
,
1977
:
Stochastic climate models. Part II: Application to sea‐surface temperature anomalies and thermocline variability
.
Tellus
,
29A
,
289
305
, https://doi.org/10.1111/j.2153-3490.1977.tb00740.x.
Henley
,
B. J.
,
J.
Gergis
,
D. J.
Karoly
,
S. B.
Power
,
J.
Kennedy
, and
C. K.
Folland
,
2015
:
A tripole index for the interdecadal Pacific oscillation
.
Climate Dyn.
,
45
,
3077
3090
, https://doi.org/10.1007/s00382-015-2525-1.
Horel
,
J.
, and
J.
Wallace
,
1981
:
Planetary-scale atmospheric phenomena associated with the Southern Oscillation
.
Mon. Wea. Rev.
,
109
,
813
829
, https://doi.org/10.1175/1520-0493(1981)109<0813:PSAPAW>2.0.CO;2.
Huang
,
B.
, and Coauthors
,
2017
:
Extended Reconstructed Sea Surface Temperature, version 5 (ERSSTv5): Upgrades, validations, and intercomparisons
.
J. Climate
,
30
,
8179
8205
, https://doi.org/10.1175/JCLI-D-16-0836.1.
Karoly
,
D. J.
,
1989
:
Southern Hemisphere circulation features associated with El Niño–Southern Oscillation events
.
J. Climate
,
2
,
1239
1252
, https://doi.org/10.1175/1520-0442(1989)002<1239:SHCFAW>2.0.CO;2.
Kirtman
,
B. P.
, and
J.
Shukla
,
2002
:
Interactive coupled ensemble: A new coupling strategy for CGCMs
.
Geophys. Res. Lett.
,
29
,
1367
, https://doi.org/10.1029/2002GL014834.
Kosaka
,
Y.
, and
S.-P.
Xie
,
2013
:
Recent global-warming hiatus tied to equatorial Pacific surface cooling
.
Nature
,
501
,
403
407
, https://doi.org/10.1038/nature12534.
Kosaka
,
Y.
, and
S.-P.
Xie
,
2016
:
The tropical Pacific as a key pacemaker of the variable rates of global warming
.
Nat. Geosci.
,
9
,
669
673
, https://doi.org/10.1038/ngeo2770.
Luo
,
J.-J.
, and
T.
Yamagata
,
2001
:
Long-term El Niño–Southern Oscillation (ENSO)-like variation with special emphasis on the South Pacific
.
J. Geophys. Res.
,
106
,
22 211
22 227
, https://doi.org/10.1029/2000JC000471.
Ma
,
J.
,
S.-P.
Xie
, and
H.
Xu
,
2017
:
Contributions of the North Pacific meridional mode to ensemble spread of ENSO prediction
.
J. Climate
,
30
,
9167
9181
, https://doi.org/10.1175/JCLI-D-17-0182.1.
Ma
,
X.
, and Coauthors
,
2016
:
Western boundary currents regulated by interaction between ocean eddies and the atmosphere
.
Nature
,
535
,
533
537
, https://doi.org/10.1038/nature18640.
Mantua
,
N. J.
, and
S. R.
Hare
,
2002
:
The Pacific decadal oscillation
.
J. Oceanogr.
,
58
,
35
44
, https://doi.org/10.1023/A:1015820616384.
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
, https://doi.org/10.1175/1520-0477(1997)078<1069:APICOW>2.0.CO;2.
Miller
,
A. J.
, and
N.
Schneider
,
2000
:
Interdecadal climate regime dynamics in the North Pacific Ocean: Theories, observations and ecosystem impacts
.
Prog. Oceanogr.
,
47
,
355
379
, https://doi.org/10.1016/S0079-6611(00)00044-6.
Minobe
,
S.
,
1997
:
A 50–70 year climatic oscillation over the North Pacific and North America
.
Geophys. Res. Lett.
,
24
,
683
686
, https://doi.org/10.1029/97GL00504.
Minobe
,
S.
,
1999
:
Resonance in bidecadal and pentadecadal climate oscillations over the North Pacific: Role in climatic regime shifts
.
Geophys. Res. Lett.
,
26
,
855
858
, https://doi.org/10.1029/1999GL900119.
Mo
,
K. C.
, and
M.
Ghil
,
1987
:
Statistics and dynamics of persistent anomalies
.
J. Atmos. Sci.
,
44
,
877
902
, https://doi.org/10.1175/1520-0469(1987)044<0877:SADOPA>2.0.CO;2.
Nakamura
,
H.
,
G.
Lin
, and
T.
Yamagata
,
1997
:
Decadal climate variability in the North Pacific during the recent decades
.
Bull. Amer. Meteor. Soc.
,
78
,
2215
2225
, https://doi.org/10.1175/1520-0477(1997)078<2215:DCVITN>2.0.CO;2.
Newman
,
M.
, and Coauthors
,
2016
:
The Pacific decadal oscillation, revisited
.
J. Climate
,
29
,
4399
4427
, https://doi.org/10.1175/JCLI-D-15-0508.1.
Newman
,
M.
,
G.
Compo
, and
M.
Alexander
,
2003
:
ENSO-forced variability of the Pacific decadal oscillation
.
J. Climate
,
16
,
3853
3857
, https://doi.org/10.1175/1520-0442(2003)016<3853:EVOTPD>2.0.CO;2.
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
383
, https://doi.org/10.2151/jmsj1965.67.3_375.
Nonaka
,
M.
,
H.
Nakamura
,
Y.
Tanimoto
,
T.
Kagimoto
, and
H.
Sasaki
,
2006
:
Decadal variability in the Kuroshio–Oyashio Extension simulated in an eddy-resolving OGCM
.
J. Climate
,
19
,
1970
1989
, https://doi.org/10.1175/JCLI3793.1.
Okumura
,
Y. M.
,
2013
:
Origins of tropical Pacific decadal variability: Role of stochastic atmospheric forcing from the South Pacific
.
J. Climate
,
26
,
9791
9796
, https://doi.org/10.1175/JCLI-D-13-00448.1.
Power
,
S.
,
T.
Casey
,
C.
Folland
,
A.
Colman
, and
V.
Mehta
,
1999
:
Inter-decadal modulation of the impact of ENSO on Australia
.
Climate Dyn.
,
15
,
319
324
, https://doi.org/10.1007/s003820050284.
Schneider
,
N.
, and
B. D.
Cornuelle
,
2005
:
The forcing of the Pacific decadal oscillation
.
J. Climate
,
18
,
4355
4373
, https://doi.org/10.1175/JCLI3527.1.
Shakun
,
J. D.
, and
J.
Shaman
,
2009
:
Tropical origins of North and South Pacific decadal variability
.
Geophys. Res. Lett.
,
36
,
L19711
, https://doi.org/10.1029/2009GL040313.
Simmons
,
A. J.
,
J. M.
Wallace
, and
G.
Branstator
,
1983
:
Barotropic wave propagation and instability, and atmospheric teleconnection patterns
.
J. Atmos. Sci.
,
40
,
1363
1392
, https://doi.org/10.1175/1520-0469(1983)040<1363:BWPAIA>2.0.CO;2.
Thompson
,
D.
, and
J.
Wallace
,
1998
:
The Arctic Oscillation signature in the wintertime geopotential height and temperature fields
.
Geophys. Res. Lett.
,
25
,
1297
1300
, https://doi.org/10.1029/98GL00950.
Thompson
,
D.
, and
J.
Wallace
,
2000
:
Annular modes in the extratropical circulation. Part I: Month-to-month variability
.
J. Climate
,
13
,
1000
1016
, https://doi.org/10.1175/1520-0442(2000)013<1000:AMITEC>2.0.CO;2.
Thompson
,
D.
,
J. J.
Kennedy
,
J. M.
Wallace
, and
P. D.
Jones
,
2008
:
A large discontinuity in the mid-twentieth century in observed global-mean surface temperature
.
Nature
,
453
,
646
649
, https://doi.org/10.1038/nature06982.
Trenberth
,
K. E.
,
1984
:
Signal versus noise in the Southern Oscillation
.
Mon. Wea. Rev.
,
112
,
326
332
, https://doi.org/10.1175/1520-0493(1984)112<0326:SVNITS>2.0.CO;2.
Trenberth
,
K. E.
,
G. W.
Branstator
,
D.
Karoly
,
A.
Kumar
,
N.
Lau
, and
C.
Ropelewski
,
1998
:
Progress during TOGA in understanding and modeling global teleconnections associated with tropical sea surface temperatures
.
J. Geophys. Res.
,
103
,
14 291
14 324
, https://doi.org/10.1029/97JC01444.
Vimont
,
D. J.
,
D. S.
Battisti
, and
A. C.
Hirst
,
2001
:
Footprinting: A seasonal connection between the tropics and mid‐latitudes
.
Geophys. Res. Lett.
,
28
,
3923
3926
, https://doi.org/10.1029/2001GL013435.
Vimont
,
D. J.
,
J. M.
Wallace
, and
D. S.
Battisti
,
2003
:
The seasonal footprinting mechanism in the Pacific: Implications for ENSO
.
J. Climate
,
16
,
2668
2675
, https://doi.org/10.1175/1520-0442(2003)016<2668:TSFMIT>2.0.CO;2.
Wallace
,
J. M.
, and
D. S.
Gutzler
,
1981
:
Teleconnections in the geopotential height field during the Northern Hemisphere winter
.
Mon. Wea. Rev.
,
109
,
784
812
, https://doi.org/10.1175/1520-0493(1981)109<0784:TITGHF>2.0.CO;2.
Wallace
,
J. M.
, and
D. W. J.
Thompson
,
2002
:
The Pacific center of action of the Northern Hemisphere annular mode: Real or artifact?
J. Climate
,
15
,
1987
1991
, https://doi.org/10.1175/1520-0442(2002)015<1987:TPCOAO>2.0.CO;2.
Wang
,
H.
,
A.
Kumar
,
W.
Wang
, and
Y.
Xue
,
2012
:
Influence of ENSO on Pacific decadal variability: An analysis based on the NCEP Climate Forecast System
.
J. Climate
,
25
,
6136
6151
, https://doi.org/10.1175/JCLI-D-11-00573.1.
Yeh
,
S. W.
,
B. P.
Kirtman
, and
S.-I.
An
,
2007
:
Local versus non-local atmospheric weather noise and the North Pacific SST variability
.
Geophys. Res. Lett.
,
34
,
L14706
, https://doi.org/10.1029/2007GL030206.
Zhang
,
L.
, and
T. L.
Delworth
,
2015
:
Analysis of the characteristics and mechanisms of the Pacific decadal oscillation in a suite of coupled models from the Geophysical Fluid Dynamics Laboratory
.
J. Climate
,
28
,
7678
7701
, https://doi.org/10.1175/JCLI-D-14-00647.1.
Zhang
,
Y.
,
J. M.
Wallace
, and
N.
Iwasaka
,
1996
:
Is climate variability over the North Pacific a linear response to ENSO?
J. Climate
,
9
,
1468
1478
, https://doi.org/10.1175/1520-0442(1996)009<1468:ICVOTN>2.0.CO;2.
Zhang
,
Y.
,
J. M.
Wallace
, and
D. S.
Battisti
,
1997
:
ENSO-like interdecadal variability: 1900–93
.
J. Climate
,
10
,
1004
1020
, https://doi.org/10.1175/1520-0442(1997)010<1004:ELIV>2.0.CO;2.

Footnotes

© 2018 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).