1. Introduction
Interannual variability of tropical Atlantic sea surface temperature (SST) is organized primarily in two modes (Xie and Carton 2004; Chang et al. 2006): the equatorial zonal mode (Atlantic Niño; Zebiak 1993) and the Atlantic meridional mode (AMM; Chiang and Vimont 2004). The Atlantic Niño is closely tied to the zonal slope of the equatorial thermocline and shifts in the Atlantic Walker circulation, similar to El Niño–Southern Oscillation (ENSO). The Atlantic Niño is phase-locked to boreal summer and involves the Bjerknes feedback (Bjerknes 1969; Carton and Huang 1994; Okumura and Xie 2006; Richter et al. 2017). The AMM involves interannual variability in interhemispheric SST gradient in the tropical Atlantic, with the anomalous cross-equatorial flow toward the anomalously warm hemisphere (Nobre and Shukla 1996; Chang et al. 1997). It modulates the seasonal march of the intertropical convergence zone (ITCZ) (Foltz et al. 2012).
The AMM consists of a northern lobe [the north tropical Atlantic meridional mode (NAMM)] and a southern lobe [the south tropical Atlantic meridional mode (SAMM)]. The AMM involves the wind–evaporation–SST (WES) feedback between the trade wind and SST anomalies (Xie and Philander 1994; Chang et al. 1997). The trade wind anomalies can be induced by large-scale climate modes including ENSO and the North Atlantic Oscillation (NAO) (Chiang et al. 2002; Amaya and Foltz 2014). In the WES feedback, the cross-equatorial flow decelerates the trade winds in the warmer hemisphere and accelerates the trade winds in the cooler hemisphere, and the resultant surface latent heat flux anomalies further amplify the cross-equatorial SST gradient.
As the Pacific counterpart of NAMM, the North Pacific meridional mode (NPMM; Chiang and Vimont 2004) refers to a subtropical SST warming (or cooling) coupled with the weakened (or strengthened) trade winds in the subtropical northeast Pacific. The coupled pattern propagates southwestward and influences ENSO variability via a seasonal footprinting mechanism (SFM) (Vimont et al. 2001, 2003; Wu et al. 2010; Amaya et al. 2019). Recent studies confirm that both the AMM and NPMM are the conduits by which extratropical atmospheric variability influences equatorial SST variability, thereby triggering Atlantic Niño and ENSO events, respectively (Vimont et al. 2001, 2003; Foltz and McPhaden 2010; Larson and Kirtman 2013, 2014; Di Lorenzo and Mantua 2016).
ENSO is the dominant mode of interannual variability in the tropical Pacific, with impacts on the global climate via atmospheric teleconnections (Cayan et al. 1999; Deser et al. 2010; Xie et al. 2016). ENSO can influence north tropical Atlantic (NTA) SST in the ensuing spring [March–May (MAM)]. The mechanisms include the atmospheric Rossby wave train crossing the North Pacific–American sector (Enfield and Mayer 1997; Giannini et al. 2000, 2001), the modulations of the Atlantic Walker circulation (Klein et al. 1999; García–Serrano et al. 2017), and tropospheric temperature via the eastward-propagating Kelvin wave (Chiang and Sobel 2002; Chang et al. 2006; Lin et al. 2007; Lintner and Chiang 2007).
SST variability in the Atlantic Ocean may influence ENSO and its predictability (Dommenget et al. 2006; Jansen et al. 2009; Frauen and Dommenget 2012; Ham et al. 2013a,b; Park et al. 2019). The Atlantic Niño in boreal summer can influence ENSO development by modulating the Walker circulation (Rodríguez-Fonseca et al. 2009; Ding et al. 2012). Ham et al. (2013a,b) argue that the NTA SST warming in boreal spring can trigger a La Niña event in the subsequent winter by inducing western Pacific easterlies through both equatorial and north-tropical pathways. The equatorial pathway modulates western Pacific wind through the eastward-propagating Kelvin wave, while the north equatorial pathway involves a low-level anomalous cyclone over the northeastern tropical Pacific Ocean as a Rossby wave response to the NTA SST warming that triggers an NPMM. Some studies also identified an eastward-propagating teleconnection from the tropical Atlantic to the Indian Ocean and adjacent regions (Kucharski et al. 2008, 2009; Li et al. 2016; Kamae et al. 2017). The tropical Atlantic SST warming can modify the zonally overturning circulation and cause the anomalous descent over the central Pacific (Hong et al. 2014, 2015; Jin and Huo 2018). While much progress has been made in Atlantic to Pacific teleconnection, the relative contributions of these processes remain to be quantified.
The dominance of ENSO in interannual variability and the short observation record render it difficult to isolate the Atlantic influence of Pacific variability. Here we turn to the seasonal forecast using a coupled model initialized with observations. The ensemble mean is often used as the forecast, while the ensemble spread (defined as deviations from the ensemble mean due to differences in initial condition) is an important measure of prediction uncertainty. The ensemble spread arising from different initial conditions can be used for investigations into the intrinsic variability of the ocean and atmosphere (Kosaka et al. 2013; Ma et al. 2017). With greatly increased degrees of freedom compared to those available in observations, the perturbed initial condition ensemble (PICE) spread offers a close look into coupled ocean–atmosphere dynamics. Here we focus on the impact of tropical Atlantic SST anomalies on the other tropical regions.
This study investigates the spatiotemporal coevolution characteristics of SST and low-level winds over the tropical Atlantic in both the ensemble mean and spread of the hindcasts, to reveal the interbasin interactions between the tropical Atlantic and Pacific. The rest of this paper is organized as follows. Section 2 introduces the data and methods. Section 3 investigates the spatiotemporal coevolution of ensemble-mean SST and low-level winds over the tropical Atlantic with the interbasin interaction characteristics. Section 4 examines the spatiotemporal coevolution of ensemble spread of the SST and low-level winds over the tropical Atlantic, and elucidates the underlying mechanism for the delayed effect of an NTA SST anomaly on ENSO. Section 5 provides a summary and further discussion.
2. Data and methods
This study uses the hindcasts of ENSEMBLES, a multimodel ensemble system developed by the European Union (van der Linden and Mitchell 2009). The ENSEMBLES includes five fully coupled atmosphere–ocean–land models from the European Centre for Medium-Range Weather Forecasts (ECMWF), the Leibniz Institute of Marine Sciences at Kiel University (IFM-GEOMAR), Météo-France (MF), the Met Office (UKMO), and the Euro-Mediterranean Center for Climate Change (CMCC-INGV). Weisheimer et al. (2009) found that SST biases in the whole tropics are small in the ENSEMBLES hindcasts. Li et al. (2012, 2014) showed that the ENSEMBLES hindcasts successfully reproduce the northwest Pacific summer climate and ENSO including its phase evolution. They also found that, among the five models, the ECMWF model exhibits the best skill in simulating the climatological 850-hPa winds and precipitation. Therefore, the hindcasts of the ECMWF model for the period of 1980–2005 are used in this study, including SST, 850-hPa horizontal winds, sea level pressure (SLP), precipitation, and surface heat flux. The atmosphere and ocean are initialized using realistic state estimates from observations on the first day of February, May, August, and November each year. Each hindcast consists of nine members that differ only slightly in initial conditions. Further details on the ENSEMBLES multimodel project and the initial condition perturbations are provided by Weisheimer et al. (2009) and van der Linden and Mitchell (2009). In this study, we mainly use the hindcasts initialized from November, which last for 14 months. Shorter (7 months) hindcasts initialized in February, May, and August are also used as necessary.
Singular value decomposition (SVD) finds covariant patterns between two fields. SVD operates on the covariance matrix between two fields and provides pairs of spatial modes with high temporal covariance. Further details on an SVD analysis in meteorology can be found in Deser and Timlin (1997). To examine the spatiotemporal evolution in the covariability of SST and low-level winds, we conduct month-reliant SVD analyses (Ma et al. 2017). An SVD analysis is performed on a concatenated 9 members × 26 years × 14 months record of the SST and wind fields in the ECMWF hindcasts. For example, the SST matrix is (Nx, Ny, Nmo, Nens, Nyr), where Nx and Ny are the grid point numbers in the zonal and meridional directions, respectively, Nmo is the number of months, Nens the ensemble size, and Nyr the number of years. First, we focus on the ensemble-mean interannual variability by obtaining the ensemble-mean anomalies defined as the deviations from the climatological mean (Nx, Ny, Nmo). Note that the original values are detrended to remove the impact of global warming. Then, the matrix (Nx × Ny × Nmo, Nyr) is used for the SVD analysis.
Then we obtain the ensemble spread by subtracting the ensemble mean (Nx, Ny, Nmo, Nyr) from the raw hindcast (Nx, Ny, Nmo, Nens, Nyr). The matrix (Nx × Ny × Nmo, Nens × Nyr) is used to conduct the SVD analysis. Hence, the conventional time dimension is enlarged by the ensemble size. Note that we investigate the SST and wind anomalies in a 14-month-long sequence from November (0) to December (1), where the numerals in parentheses indicate the year relative to the initialization year. A covariance matrix is constructed by treating the SST or wind anomalies in the monthly sequence as one yearly step. Thus, we can obtain the heterogeneous fields consisting of 14 sequential patterns representing monthly evolution of the ensemble mean or spread for each mode. As these patterns share the same principal component (PC), the heterogeneous fields reflect the temporal evolution characteristics. In this study, multivariate SVD analyses are conducted for covariability between two (zonal and meridional winds) atmospheric variables and SST over the tropical Atlantic (30°S–30°N, 90°W–0°).
To compare the PC1 and PC2 of ensemble-mean interannual anomalies with the counterparts of nine ensemble members, we also conduct the SVD analyses to the interannual anomalies defined as the deviations (Nx, Ny, Nmo, Nens, Nyr) from the climatological mean (Nx, Ny, Nmo, Nens). The original values are also detrended to remove the impact of global warming. The matrix (Nx × Ny × Nmo, Nens × Nyr) is used for the SVD analysis. Thus, we can obtain the PC1 and PC2 for each ensemble member.
To map the leading SVD modes, we correlate the PCs with SST, SLP, precipitation, surface heat fluxes, and other atmospheric fields. The Student’s t test is used to determine the significance of correlations.
3. Evolution of ensemble-mean anomalies
This section presents the ensemble-mean results. We investigate the spatiotemporal coevolution of ensemble-mean SST and 850-hPa wind variability over the tropical Atlantic (30°S–30°N, 90°W–0°) by conducting the SVD analyses. The monthly wind and SST anomalies are used as the left and right fields, respectively. Figure 1 shows the left and right heterogeneous fields of the first month-reliant SVD mode (explained covariance: 45.01%) of the ensemble-mean variability from November to December of the following year. SST anomalies in the NTA region are weak from November to January, with negative SST anomalies near the equatorial Atlantic. In January, anomalous southwesterlies appear over the NTA region associated with a local SST warming via the WES mechanism (Xie and Philander 1994). The coupling between the anomalous southwesterlies and SST warming in the NTA region persists through June. Positive SST anomalies also appear near 0°–10°N of the North Atlantic from May to September. In the south tropical Atlantic (STA) off the equator, positive SST anomalies occur from May to December with the reduced southeasterly trade wind, peaking in April–July. From February to December, SST anomalies in much of the STA is of the same sign as those the NTA, in response to El Niño (Enfield and Mayer 1997). In addition, Atlantic Niño-like negative SST anomalies develop from July to September, which may be driven by the Bjerknes feedback (Bjerknes 1969).
Left (vectors) and right (color shading) heterogeneous fields of the first month-reliant SVD mode of the ensemble-mean 850-hPa winds and SST in the tropical Atlantic region from November (0) to December (1) at an interval of 2 months using the hindcasts initialized from November (the same hereafter unless otherwise stated). Small vectors are omitted for clarity.
Citation: Journal of Climate 34, 7; 10.1175/JCLI-D-20-0140.1
The second SVD mode (explained covariance: 20.51%) of the ensemble-mean 850-hPa winds and SST in the tropical Atlantic region shows an NTA SST warming that persists through the analysis period and peaks from January to March (Fig. 2). An SST warming develops in the equatorial Atlantic in June–August (JJA) to September–November (SON), associated with the westerly anomalies and the reduced upwelling near the equator due to the Bjerknes feedback. Notably, positive SST anomalies are located near 0°–10°N from November to March. Over the STA region, SST anomaly remains positive from November to May, but turns negative from July to December. The interhemispheric SST gradient in the tropical Atlantic, accompanied by cross-equatorial 850-hPa flow toward the anomalously warm hemisphere, is clear from July to September (Figs. 2e,f).
As in Fig. 1, but for the second mode.
Citation: Journal of Climate 34, 7; 10.1175/JCLI-D-20-0140.1
Figures 3a and 3b show PC1 and PC2 of the right heterogeneous (SST) fields of the month-reliant SVD modes, respectively. Each PC of the ensemble-mean variability is close to the average of the PC of nine ensemble members.
The (a) first principal component (PC1) and (b) second principal component (PC2) of the right heterogeneous fields of the month-reliant SVD modes of the ensemble-mean 850-hPa winds and SST in the tropical Atlantic region from November to December of the following year (black dots). Circles with different colors indicate the PCs of nine ensemble members. The years along the x axis correspond to those with November initialization.
Citation: Journal of Climate 34, 7; 10.1175/JCLI-D-20-0140.1
Figure 4 shows the lagged correlation with the Niño-3.4 index of November (0) to December (1) for the ensemble-mean PCs. The PC1 is significantly positively correlated with the Niño-3.4 index from November (0) to June (1), and the correlation coefficients decrease sharply in the following months. This indicates that the first SVD mode represents the transition of El Niño from the peak to decay phase. This is consistent with the finding that an Atlantic warming occurs 4–5 months after the mature phase of a Pacific warm event (Enfield and Mayer 1997). The ensemble-mean PC2 is significantly negatively correlated to the Niño-3.4 index from June (1) to December (1), implying that it favors a La Niña event in the following winter.
Lagged correlation with the Niño-3.4 index of November (0) to December (1) for the ensemble-mean PC1 (solid curve) and PC2 (dashed curve). The thicker lines represent the values significant at the 95% level.
Citation: Journal of Climate 34, 7; 10.1175/JCLI-D-20-0140.1
To further investigate the interbasin interaction, Fig. 5 shows the correlation maps of SST and 850-hPa winds with PC1 of the ensemble-mean SVD mode from November to December of the following year at an interval of two months. An El Niño event reaches its peak in winter, persists until May, and begins to dissipate after June. In response to the El Niño event, SST exhibits a tropical-wide warming pattern. Notably positive SST anomalies occur in the tropical Atlantic and Indian Oceans during January to July. The area with positive SST anomalies shrinks following the decay of El Niño. Our result is consistent with the general features of the ENSO forced mode in the Pacific Ocean–Global Atmosphere (POGA) pacemaker experiments of Yang et al. (2018).
First SVD mode of the ensemble-mean 850-hPa winds (vectors) and SST (shading) in the tropical Atlantic, expressed as correlations with PC1 from November (0) to December (1) at an interval of two months. Only the correlations significant at the 90% level are shown. The same threshold is applied hereafter.
Citation: Journal of Climate 34, 7; 10.1175/JCLI-D-20-0140.1
Figure 6 shows the evolution of the ensemble-mean SVD2 mode. Significantly positive SST anomalies occur in the NTA from November to May, accompanied by the tropical Indian Ocean warming. The NTA warming excites the eastward-propagating tropospheric Kelvin wave (Kucharski et al. 2008, 2009; Li et al. 2016; Kamae et al. 2017) and causes the anomalous descent over the central Pacific by modifying the zonal overturning circulation (Jin and Huo 2018). A pronounced SST warming develops in the equatorial Atlantic in JJA–SON, consistent with Fig. 2. In the central-eastern equatorial Pacific, negative SST anomalies appear in June and grows through December. This indicates that the NTA SST warming acts as a precursor for a La Niña event, consistent with the significantly negative correlation between the ensemble-mean PC2 and the Niño-3.4 index from June (1) to December (1). While this result is in agreement with Ham et al. (2013a,b), the detailed adjustments from a NTA warming to La Niña are hard to identify from the noisy ensemble-mean variations of limited degrees of freedom.
4. Ensemble spread
In this section, we turn attention to the spatiotemporal evolution of ensemble spread. We conduct month-reliant SVD analyses of the ensemble spread of 850-hPa winds and SST in the tropical Atlantic (30°S–30°N, 90°W–0°) from November to December of the following year. The first mode explains 18.2% of the total covariance, and shows that an NTA warming triggers a subsequent La Niña event via a north-tropical pathway through an NPMM. We have repeated the SVD analyses for the entire tropical Atlantic–Pacific domain. The first mode shows a similar development of La Niña with an antecedent NTA warming, but the NPMM appears nearly simultaneously with the NTA warming, making it difficult to isolate the impact of NTA anomalies on ENSO (not shown). Therefore, we choose to present the SVD analyses for the tropical Atlantic.
a. An NTA SST warming triggers a negative NPMM
Figure 7 shows the evolution of the leading SVD mode of ensemble spread in SST, 850-hPa winds, and latent heat flux, expressed as the correlation coefficients with PC1, from December (0) to December (1) at an interval of two months. NTA SST anomalies are initially weak, induced by the relaxed trade winds. The 850-hPa wind anomalies in the North Atlantic from winter to spring resemble the NAO pattern with a meridional dipole between a southern cyclone and a northern anticyclone (not shown). The weakened northeast trade winds in the NTA drive an anomalous flux of latent energy into the ocean, increasing SST (Fig. 7a). The coupling between positive SST and southwesterly anomalies in the deep tropics is indicative of the WES feedback (Xie and Philander 1994). The strong coupling occurs in MAM, consistent with the result of Amaya et al. (2017), who found that the leading mode of SST and wind covariability during MAM is much more pronounced in magnitude and spatial extent. Reduced upward surface latent heat flux (yellow contours in Fig. 7) contributes to the NTA SST warming in December–February (DJF) and MAM. The NTA SST warming amplifies into JJA. Moreover, an interhemispheric antisymmetric pattern of SST anomalies between the NTA and STA is clear. The SST cooling in the STA is due to the WES mechanism related to the acceleration of the southeast trade winds. Enhanced upward surface latent heat flux gives rise to the STA SST cooling in DJF–MAM (Fig. 7). Low-level winds blow toward the warm NTA region, with southerlies near the equator. This is consistent with previous studies (Nobre and Shukla 1996; Chang et al. 1997), which identified that the interhemispheric SST gradient in the tropical Atlantic is closely tied to the cross-equatorial flow toward the anomalously warm hemisphere. SST and wind anomalies are not limited to the tropical Atlantic, which is the domain for our SVD analysis. In Fig. 7, coherent anomalies emerge in February over the subtropical northeast Pacific and expand across the entire tropical Pacific.
First SVD mode of 850-hPa winds (vectors) and SST (shading) ensemble spread in the tropical Atlantic, expressed as correlations with PC1 from December (0) to December (1) at an interval of two months, along with downward surface latent heat flux (contours, negative dashed; yellow and magenta contours represent the significant positive and negative correlations, respectively). A 1–2–1 running mean is applied in time, and the same is true in the following figures.
Citation: Journal of Climate 34, 7; 10.1175/JCLI-D-20-0140.1
Figure 8 zooms in for a close-up of rotational wind and streamfunction anomalies at 850 hPa associated with SVD1. In November, anomalies are not well organized over the Pacific, presumably because the NTA SST anomalies are still small in magnitude. Weak westerly wind anomalies develop across low-lying Central America while the mountains block low-level wind response at other latitudes. As the NTA SST forcing grows from December to January, an anomalous cyclonic circulation develops west of Central America, presumably as the baroclinic Rossby response to the increased convective heating over the NTA region. The anomalous northerly winds on the west flank of the anomalous cyclone drive an SST cooling through the WES feedback and cold-dry advection from higher latitudes. Negative SST anomalies grow and migrate southwestward, coupled with northerly wind anomalies. The coupling of SST and low-level winds in the subtropical northeast Pacific is suggestive of the NPMM. Negative surface latent heat flux anomalies (dashed purple contours in Fig. 7) are displaced southwest of the SST cooling in the subtropical northeast Pacific, causing the southwestward copropagation as discussed in Vimont et al. (2001, 2003) and Wu et al. (2010).
As in Fig. 7, but for correlations of SST (shading), rotational winds (vectors), and streamfunction (contours) at 850 hPa with PC1 from November (0) to April (1) at an interval of one month in the region of 10°S–30°N, 160°E–50°W.
Citation: Journal of Climate 34, 7; 10.1175/JCLI-D-20-0140.1
Positive and negative velocity potential anomalies are associated with a convergent inflow and divergent outflow, respectively. Positive velocity potential anomalies at 850 hPa occur over the NTA region (shadings in Fig. 9), indicative of the low-level convergence. The NTA SST warming enhances the convective activity and diabatic heating, which produces the anomalous low over the NTA region as a Rossby wave response (Gill 1980). The anomalous low is accompanied by rainfall increase north of the equatorial Atlantic (stippling in Fig. 9), indicating that the ITCZ is shifted northward in response to a warm NTA region. This is in agreement with Chiang et al. (2002), who found that the ITCZ is sensitive to variations in the meridional SST gradient over the tropical Atlantic.
As in Fig. 7, but for velocity potential (shading) and divergent winds (vectors) at 850 hPa, along with precipitation (stippling; blue for positive and red for negative) with PC1.
Citation: Journal of Climate 34, 7; 10.1175/JCLI-D-20-0140.1
A low-level divergence occurs over the subtropical northeast Pacific in response to the convective heating with low-level convergent flow over the NTA region from December to April (Fig. 9). The divergence over the subtropical northeast Pacific is accompanied by local precipitation decreases (stippling in Fig. 9). The reduced convective heating excites a low-level anomalous anticyclone as an atmospheric Rossby wave response (Fig. 8). This anomalous anticyclone contributes to the northerly wind anomalies on the east flank, contributing to the SST cooling through the WES mechanism (Xie and Philander 1994). The velocity potential anomalies at 200 hPa (not shown) are generally opposite to those at 850 hPa, indicating a deep anomalous zonal–vertical circulation with ascending motion over the NTA region, and descending motion over the subtropical North Pacific. The low-level divergence occurs over the northeast tropical Pacific in June (Fig. 9d), possibly because of the seasonal SST warming and intensification of the ITCZ (Amaya et al. 2019).
The ensemble spread is small at the beginning of the hindcast. Figure 10 shows the growth of the ensemble spread in SST and low-level wind during the first four months. The first mode using the February initialization ensemble captures an evolution pattern similar to that in the November ensemble, whereas the pattern appears as the second mode in the May and August ensembles. Figure 10 is based on the correlation coefficients with PC1 (PC2) for November and February (May and August) ensembles. In all the four ensembles, SST cooling amplifies over the subtropical northeast Pacific following the NTA SST warming in the first four months. The cooling is coupled with northerly anomalies to its northeast, induced by negative surface latent heat flux anomalies. Additionally, low-level divergence over the subtropical northeast Pacific responds to the notable convergent flows over the NTA region in the fourth month in all the hindcasts (Fig. 11). The divergence is accompanied by an anomalous anticyclone over the subtropical North Pacific. These results show that the NTA warming can trigger a negative NPMM. The SST cooling and northerly anomalies over the subtropical northeast Pacific are more significant in spring–summer than those in autumn–winter. This seasonality of the NPMM response is primarily due to the deep convective feedback (Amaya et al. 2019).
Evolution of tropical Atlantic influence on the Pacific at months (left) 2 and (right) 4 after initialization: Correlations of 850-hPa winds (vectors), SST (shading), and surface latent heat flux (contours; positive downward; dashed for negative; yellow and magenta contours represent the significant positive and negative correlations, respectively) with the PC of the SVD mode. The Atlantic-to-Pacific mode is SVD1 in hindcasts with [a(1)],[b(1)] November and [a(2)],[b(2)] February initialization, and SVD2 in hindcasts with [a(3)],[b(3)] May and [a(4)],[b(4)] August initialization.
Citation: Journal of Climate 34, 7; 10.1175/JCLI-D-20-0140.1
As in Fig. 10, but for precipitation (stippling; blue for positive and red for negative), velocity potential (shading), and divergent winds (vectors) at 850 hPa.
Citation: Journal of Climate 34, 7; 10.1175/JCLI-D-20-0140.1
In the November-initialized hindcasts, the second ensemble spread mode accounts for 12.1% of the total covariance over the tropical Atlantic, featuring SST anomalies that start from the STA and expand into the NTA with a weak extension into the tropical Pacific (not shown). This mode does not involve the NPMM.
b. A negative NPMM induces La Niña
This subsection examines the role of the NPMM in the development of La Niña based on the correlation coefficients with PC1 of the ensemble spread in the tropical Atlantic using the November-initialized hindcasts. Negative SST anomalies coupled with the anomalous northeasterlies in the subtropical northeast Pacific arrive at the central equatorial Pacific in June (Fig. 7d). In the subsequent months, a La Niña event develops. This implies that the NTA SST warming acts as a precursor for a La Niña event by triggering a negative NPMM. Low-level divergence expands from the subtropical northeast Pacific to the equatorial Pacific from April to August. A rainfall deficit occurs over the central-eastern equatorial Pacific from August to December, in response to the developing La Niña event (Figs. 9e–g).
Figure 12 shows the longitude–month cross section of the correlations of SST, precipitation, and velocity potential with PC1 of the ensemble spread SVD mode, averaged over 0°–20°N. The NTA SST warming grows in the first four months and persists throughout the whole period. SST cooling in the subtropical northeast Pacific grows and expands westward from December to May, and persists thereafter. The cooling is mainly caused by enhanced upward latent heat flux. Low-level divergence appears in the subtropical North Pacific soon after the initialization, following the convergence over the NTA region, with consistent rainfall anomalies. This implies a zonal–vertical anomalous circulation associated with the baroclinic Rossby wave response to the NTA warming. The coupling of SST cooling and low-level winds, anomalous anticyclones, and rainfall decrease in the subtropical northeast Pacific is characteristic of the NPMM. Previous studies show that the NPMM can affect the subsequent ENSO (Vimont et al. 2001, 2003; Chang et al. 2006; Ma et al. 2017). Here, we show that an NTA warming can trigger a La Niña in the central equatorial Pacific through the NPMM pathway across Central America.
Longitude–month cross section of the correlation coefficients of SST (shading), precipitation (stippling; blue for positive and red for negative), and velocity potential (contours; solid for positive, and dashed for negative; yellow and magenta contours represent the significant positive and negative correlations, respectively) with PC1 of the SVD mode of the ensemble spread of 850-hPa winds and SST in the tropical Atlantic averaged along 0°–20°N.
Citation: Journal of Climate 34, 7; 10.1175/JCLI-D-20-0140.1
Ham et al. (2013a) proposed that the NTA SST warming can trigger a La Niña event via a north-tropical pathway, but they could not rule out the Kelvin wave effect on winds over the western equatorial Pacific because of ENSO’s dominance in limited observational records. Our study unequivocally identifies the north-tropical pathway from the ensemble hindcasts that increase the sample size by one order of magnitude compared to available observations.
In our ensemble spread analysis, zonal wind anomalies are insignificant over the equatorial western Pacific for the first four months after initialization (Fig. 7). This indicates that the Kelvin wave pathway is unimportant, perhaps because the tropical SST anomalies are antisymmetric about the equator, a pattern unfavorable for the Kelvin wave response that peaks on the equator. By contrast, the Kelvin wave pathway is evident in the second SVD mode of the ensemble-mean variability (Figs. 6b–d), as indicated by the marked easterly wind anomalies in the western equatorial Pacific, because the SST anomalies are of the same sign over the entire tropical Atlantic. The SST pattern seems important in selecting between the equatorial Kelvin and northern Rossby wave pathways.
Finally, we present the temporal evolution of the standard deviations (STD) of SST and 850-hPa zonal wind in the regions of NTA (0°–30°N, 90°W–0°) and Niño-4 (5°S–5°N, 160°E–150°W) (Fig. 13). The STD of NTA SST shows a rapid growth during the first few months, peaking in April–June and dipping slightly in August–September. The STD of 850-hPa zonal wind in the NTA peaks in February and dips in August, consistent with Yang et al. (2018). The rapid growth of Niño-4 SST STD from July to November coincides with the La Niña growth following the arrival of the NTA-induced NPMM at the central equatorial Pacific (Fig. 7).
Evolution of the ensemble spread of (a) SST (units: K) and (b) 850-hPa zonal wind (units: m s−1) in the regions of NTA (solid curves) and Niño-4 (dashed curves) as a function of calendar months.
Citation: Journal of Climate 34, 7; 10.1175/JCLI-D-20-0140.1
5. Summary and discussion
This study investigates the spatiotemporal coevolution of SST and low-level winds over the tropical Atlantic by applying a month-reliant SVD method to the ensemble mean and spread of the ECMWF hindcasts. The first SVD mode of the ensemble-mean variability generally shows the ENSO-forced variability (Enfield and Mayer 1997), with a coherent coevolution of SST and winds over the tropical Atlantic related to the transition of El Niño from the peak to decay phase. SST anomalies in the NTA and STA regions tend to be of the same sign from February to December. The second mode of the ensemble mean is associated with a phase transition from El Niño to La Niña, indicating that the NTA SST warming serves as a precursor for a La Niña event, in agreement with previous studies (Ham et al. 2013a,b).
Tapping into a much larger sample size than available observations, the ensemble spread analysis enables the detection of weak coupled modes that would be otherwise overwhelmed by energetic ENSO. The first SVD mode of the ensemble spread confirms that an NTA SST anomaly can trigger a subsequent ENSO event. We identify a north-tropical pathway for the NTA SST anomaly to impact the subsequent ENSO event. The NTA SST anomaly induces a signal in the subtropical northeast Pacific that resembles the NPMM. The NPMM grows and migrates southwestward due to the WES feedback, triggering eventually a La Niña event upon arrival at the central equatorial Pacific. Our study shows that the NTA teleconnection involves a zonal–vertical circulation, in which low-level divergence occurs in the subtropical North Pacific in response to the convective heating over the anomalous warm NTA region. The NPMM is characterized by the coupling of an SST cooling, a low-level anomalous anticyclone, and decreased rainfall in the subtropical northeast Pacific.
The eastward-propagating tropospheric Kelvin waves are an alternative pathway for tropical Atlantic SST anomalies to affect the Indian and Pacific Oceans (Ham et al. 2013a; Li et al. 2016; Kamae et al. 2017). The SST pattern over the tropical Atlantic seems crucial in selecting the mechanism for the teleconnection to the tropical Pacific. A cross-equatorial SST dipole over the tropical Atlantic, as in the ensemble-spread SVD1 (Fig. 7), is unfavorable for Kelvin wave excitation and ends up choosing the northern Rossby wave pathway, while a tropical wide warming in the Atlantic, as in the ensemble-mean SVD2 (Fig. 6), chooses the Kelvin wave pathway.
The tropical oceans are a tightly interconnected system (Cai et al. 2019). Our study shows an important role of the Atlantic in the pan-tropical ocean–atmosphere interactions. Specifically NTA SST variability may aid a fast phase transition of ENSO, as suggested by Ham et al. (2013a). When an El Niño event induces an NTA warming, the NTA warming may result in a fast termination of El Niño and possibly a rapid development of a La Niña event. In observations for the period of 1950–2018, 12 out of 24 El Niño events are followed by La Niña. Finally we note that this study uses only the hindcasts from the ECMWF model. SVD analyses are also applied to other model hindcasts in the ENSEMBLES. The patterns and evolution are similar to those in the ECMWF model, albeit with some differences. A close examination of the multimodel hindcasts will be the focus of our future study.
Acknowledgments
NUIST authors are jointly supported by the National Natural Science Foundation of China (41805051, 42030605, 41975106), and the Startup Foundation for Introducing Talent of NUIST (2017r057). S.P.X. is supported by the National Science Foundation (AGS 1637450). J.Z. is funded by the Korea Meteorological Administration Research and Development Program under Grant KMIPA 2018-03212.
REFERENCES
Amaya, D. J., and G. R. Foltz, 2014: Impacts of canonical and Modoki El Niño on tropical Atlantic SST. J. Geophys. Res. Oceans, 119, 777–789, https://doi.org/10.1002/2013JC009476.
Amaya, D. J., M. J. DeFlorio, A. J. Miller, and S.-P. Xie, 2017: WES feedback and the Atlantic meridional mode: Observations and CMIP5 comparisons. Climate Dyn., 49, 1665–1679, https://doi.org/10.1007/s00382-016-3411-1.
Amaya, D. J., Y. Kosaka, W. Zhou, Y. Zhang, S.-P. Xie, and A. Miller, 2019: The North Pacific pacemaker effect on historical ENSO and its mechanisms. J. Climate, 32, 7643–7661, https://doi.org/10.1175/JCLI-D-19-0040.1.
Bjerknes, J., 1969: Atmospheric teleconnections from the equatorial Pacific. Mon. Wea. Rev., 97, 163–172, https://doi.org/10.1175/1520-0493(1969)097<0163:ATFTEP>2.3.CO;2.
Cai, W., and Coauthors, 2019: Pantropical climate interactions. Science, 363, eaav4236, https://doi.org/10.1126/science.aav4236.
Carton, J. A., and B. Huang, 1994: Warm events in the tropical Atlantic. J. Phys. Oceanogr., 24, 888–903, https://doi.org/10.1175/1520-0485(1994)024<0888:WEITTA>2.0.CO;2.
Cayan, D. R., K. T. Redmond, and L. G. Riddle, 1999: ENSO and hydrologic extremes in the western United States. J. Climate, 12, 2881–2893, https://doi.org/10.1175/1520-0442(1999)012<2881:EAHEIT>2.0.CO;2.
Chang, P., L. Ji, and H. Li, 1997: A decadal climate variation in the tropical Atlantic Ocean from thermodynamic air–sea interactions. Nature, 385, 516–518, https://doi.org/10.1038/385516a0.
Chang, P., Y. Fang, R. Saravanan, L. Ji, and H. Seidel, 2006: The cause of the fragile relationship between the Pacific El Niño and the Atlantic El Niño. Nature, 443, 324–328, https://doi.org/10.1038/nature05053.
Chiang, J., and A. Sobel, 2002: Tropical tropospheric temperature variations caused by ENSO and their influence on the remote tropical climate. J. Climate, 15, 2616–2631, https://doi.org/10.1175/1520-0442(2002)015<2616:TTTVCB>2.0.CO;2.
Chiang, J., and D. J. Vimont, 2004: Analogous Pacific and Atlantic meridional modes of tropical atmosphere–ocean variability. J. Climate, 17, 4143–4158, https://doi.org/10.1175/JCLI4953.1.
Chiang, J., Y. Kushnir, and A. Giannini, 2002: Deconstructing Atlantic Intertropical Convergence Zone variability: Influence of the local cross-equatorial sea surface temperature gradient and remote forcing from the eastern equatorial Pacific. J. Geophys. Res., 107, 4004, https://doi.org/10.1029/2000JD000307.
Deser, C., and M. S. Timlin, 1997: Atmosphere–ocean interaction on weekly timescales in the North Atlantic and Pacific. J. Climate, 10, 393–408, https://doi.org/10.1175/1520-0442(1997)010<0393:AOIOWT>2.0.CO;2.
Deser, C., M. Alexander, S.-P. Xie, and A. S. Phillips, 2010: Sea surface temperature variability: Patterns and mechanisms. Annu. Rev. Mar. Sci., 2, 115–143, https://doi.org/10.1146/annurev-marine-120408-151453.
Di Lorenzo, E., and N. Mantua, 2016: Multi-year persistence of the 2014/15 North Pacific marine heatwave. Nat. Climate Change, 6, 1042–1047, https://doi.org/10.1038/nclimate3082.
Ding, H., N. S. Keenlyside, and M. Latif, 2012: Impact of the equatorial Atlantic on the El Niño Southern Oscillation. Climate Dyn., 38, 1965–1972, https://doi.org/10.1007/s00382-011-1097-y.
Dommenget, D., V. Semenov, and M. Latif, 2006: Impacts of the tropical Indian and Atlantic Oceans on ENSO. Geophys. Res. Lett., 33, L11701, https://doi.org/10.1029/2006GL025871.
Enfield, D., and D. Mayer, 1997: Tropical Atlantic SST variability and its relation to El Niño–Southern Oscillation. J. Geophys. Res., 102, 929–945, https://doi.org/10.1029/96JC03296.
Foltz, G. R., and M. J. McPhaden, 2010: Abrupt equatorial wave-induced cooling of the Atlantic cold tongue in 2009. Geophys. Res. Lett., 37, L24605, https://doi.org/10.1029/2010GL045522.
Foltz, G. R., M. J. McPhaden, and R. Lumpkin, 2012: A Strong Atlantic meridional mode event in 2009: The role of mixed layer dynamics. J. Climate, 25, 363–380, https://doi.org/10.1175/JCLI-D-11-00150.1.
Frauen, C., and D. Dommenget, 2012: Influences of the tropical Indian and Atlantic Oceans on the predictability of ENSO. Geophys. Res. Lett., 39, L02706, https://doi.org/10.1029/2011GL050520.
García-Serrano, J., C. Cassou, H. Douville, A. Giannini, and F. J. Doblas-Reyes, 2017: Revisiting the ENSO teleconnection to the tropical North Atlantic. J. Climate, 30, 6945–6957, https://doi.org/10.1175/JCLI-D-16-0641.1.
Giannini, A., Y. Kushnir, and M. Cane, 2000: Interannual variability of Caribbean rainfall, ENSO, and the Atlantic Ocean. J. Climate, 13, 297–311, https://doi.org/10.1175/1520-0442(2000)013<0297:IVOCRE>2.0.CO;2.
Giannini, A., J. Chiang, M. Cane, Y. Kushnir, and R. Seager, 2001: The ENSO teleconnection to the tropical Atlantic Ocean: Contributions of the remote and local SSTs to rainfall variability in the tropical Americas. J. Climate, 14, 4530–4544, https://doi.org/10.1175/1520-0442(2001)014<4530:TETTTT>2.0.CO;2.
Gill, A. E., 1980: Some simple solutions for heat-induced tropical circulation. Quart. J. Roy. Meteor. Soc., 106, 447–462, https://doi.org/10.1002/qj.49710644905.
Ham, Y.-G., J.-S. Kug, J.-Y. Park, and F.-F. Jin, 2013a: Sea surface temperature in the north tropical Atlantic as a trigger for El Niño/Southern Oscillation events. Nat. Geosci., 6, 112–116, https://doi.org/10.1038/ngeo1686.
Ham, Y.-G., J.-S. Kug, and J.-Y. Park, 2013b: Two distinct roles of Atlantic SSTs in ENSO variability: North tropical Atlantic SST and Atlantic Niño. Geophys. Res. Lett., 40, 4012–4017, https://doi.org/10.1002/grl.50729.
Hong, C.-C., T.-C. Chang, and H.-H. Hsu, 2014: Enhanced relationship between the tropical Atlantic SST and the summertime western North Pacific subtropical high after the early 1980s. J. Geophys. Res. Atmos., 119, 3715–3722, https://doi.org/10.1002/2013JD021394.
Hong, C.-C., M.-Y. Lee, H.-H. Hsu, N.-H. Lin, and B.-J. Tsuang, 2015: Tropical SST forcing on the anomalous WNP subtropical high during July–August 2010 and the record–high SST in the tropical Atlantic. Climate Dyn., 45, 633–650, https://doi.org/10.1007/s00382-014-2275-5.
Jansen, M. F., D. Dommenget, and N. Keenlyside, 2009: Tropical atmosphere–ocean interactions in a conceptual framework. J. Climate, 22, 550–567, https://doi.org/10.1175/2008JCLI2243.1.
Jin, D., and L. Huo, 2018: Influence of tropical Atlantic sea surface temperature anomalies on the East Asian summer monsoon. Quart. J. Roy. Meteor. Soc., 144, 1490–1500, https://doi.org/10.1002/qj.3296.
Kamae, Y., X. Li, S.-P. Xie, and H. Ueda, 2017: Atlantic effects on recent decadal trends in global monsoon. Climate Dyn., 49, 3443–3455, https://doi.org/10.1007/s00382-017-3522-3.
Klein, S. A., B. J. Soden, and N.-C. Lau, 1999: Remote sea surface temperature variations during ENSO: Evidence for a tropical atmospheric bridge. J. Climate, 12, 917–932, https://doi.org/10.1175/1520-0442(1999)012<0917:RSSTVD>2.0.CO;2.
Kosaka, Y., S.-P. Xie, N. C. Lau, and G. A. Vecchi, 2013: Origin of seasonal predictability for summer climate over the northwestern Pacific. Proc. Natl. Acad. Sci. USA, 110, 7574–7579, https://doi.org/10.1073/pnas.1215582110.
Kucharski, F., A. Bracco, J. H. Yoo, and F. Molteni, 2008: Atlantic forced component of the Indian monsoon interannual variability. Geophys. Res. Lett., 35, L04706, https://doi.org/10.1029/2007GL033037.
Kucharski, F., A. Bracco, J. H. Yoo, A. M. Tompkins, L. Feudale, P. Ruti, and A. Dell’Aquila, 2009: A Gill–Matsuno–type mechanism explains the tropical Atlantic influence on African and Indian monsoon rainfall. Quart. J. Roy. Meteor. Soc., 135, 569–579, https://doi.org/10.1002/qj.406.
Larson, S. M., and B. P. Kirtman, 2013: The Pacific meridional mode as a trigger for ENSO in a high-resolution coupled model. Geophys. Res. Lett., 40, 3189–3194, https://doi.org/10.1002/grl.50571.
Larson, S. M., and B. P. Kirtman, 2014: The Pacific meridional mode as an ENSO precursor and predictor in the North American Multimodel Ensemble. J. Climate, 27, 7018–7032, https://doi.org/10.1175/JCLI-D-14-00055.1.
Li, C., R. Lu, and B. Dong, 2012: Predictability of the western North Pacific summer climate demonstrated by the coupled models of ENSEMBLES. Climate Dyn., 39, 329–346, https://doi.org/10.1007/s00382-011-1274-z.
Li, C., R. Lu, and B. Dong, 2014: Predictability of the western North Pacific summer climate associated with different ENSO phases by ENSEMBLES multi-model seasonal forecasts. Climate Dyn., 43, 1829–1845, https://doi.org/10.1007/s00382-013-2010-7.
Li, X., S.-P. Xie, S. T. Gille, and C. Yoo, 2016: Atlantic-induced pan-tropical climate change over the past three decades. Nat. Climate Change, 6, 275–279, https://doi.org/10.1038/nclimate2840.
Lin, H., J. Derome, and G. Brunet, 2007: The nonlinear transient atmospheric response to tropical forcing. J. Climate, 20, 5642–5665, https://doi.org/10.1175/2007JCLI1383.1.
Lintner, B., and J. Chiang, 2007: Adjustment of the remote tropical climate to El Niño conditions. J. Climate, 20, 2544–2557, https://doi.org/10.1175/JCLI4138.1.
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.
Nobre, P., and J. Shukla, 1996: Variations of sea surface temperature, wind stress, and rainfall over the tropical Atlantic and South America. J. Climate, 9, 2464–2479, https://doi.org/10.1175/1520-0442(1996)009<2464:VOSSTW>2.0.CO;2.
Okumura, Y., and S.-P. Xie, 2006: Some overlooked features of tropical Atlantic climate leading to a new Niño-like phenomenon. J. Climate, 19, 5859–5874, https://doi.org/10.1175/JCLI3928.1.
Park, J.-H., T. Li, S.-W. Yeh, and H. Kim, 2019: Effect of recent Atlantic warming in strengthening Atlantic–Pacific teleconnection on interannual timescale via enhanced connection with the Pacific meridional mode. Climate Dyn., 53, 371–387, https://doi.org/10.1007/s00382-018-4591-7.
Richter, I., S.-P. Xie, Y. Morioka, T. Doi, B. Taguchi, and S. Behera, 2017: Phase locking of equatorial Atlantic variability through the seasonal migration of the ITCZ. Climate Dyn., 48, 3615–3629, https://doi.org/10.1007/s00382-016-3289-y.
Rodríguez-Fonseca, B., I. Polo, J. García-Serrano, T. Losada, E. Mohino, C. R. Mechoso, and F. Kucharski, 2009: Are Atlantic Niños enhancing Pacific ENSO events in recent decades? Geophys. Res. Lett., 36, L20705, https://doi.org/10.1029/2009GL040048.
van der Linden, P., and J. F. B. Mitchell, Eds., 2009: ENSEMBLES: Climate change and its impact: Summary of research and results from the ENSEMBLES Project. Met Office Hadley Centre, 160 pp.
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., D. S. Battisti, and A. C. Hirst, 2003: The seasonal footprinting mechanism in the CSIRO general circulation models. J. Climate, 16, 2653–2667, https://doi.org/10.1175/1520-0442(2003)016<2653:TSFMIT>2.0.CO;2.
Weisheimer, A., and Coauthors, 2009: ENSEMBLES: A new multi-model ensemble for seasonal-to-annual predictions—Skill and progress beyond DEMETER in forecasting tropical Pacific SSTs. Geophys. Res. Lett., 36, L21711, https://doi.org/10.1029/2009GL040896.
Wu, S., L. Wu, Q. Liu, and S.-P. Xie, 2010: Development processes of the tropical Pacific meridional mode. Adv. Atmos. Sci., 27, 95–99, https://doi.org/10.1007/s00376-009-8067-x.
Xie, S.-P., and S. G. H. Philander, 1994: A coupled ocean–atmosphere model of relevance to the ITCZ in the eastern Pacific. Tellus, 46A, 340–350, https://doi.org/10.3402/tellusa.v46i4.15484.
Xie, S.-P., and J. A. Carton, 2004: Tropical Atlantic variability: Patterns, mechanisms, and impacts. Earth’s Climate: The Ocean–Atmosphere Interaction, Geophys. Monogr., Vol. 147, Amer. Geophys. Union, 121–142.
Xie, S.-P., Y. Kosaka, Y. Du, K. M. Hu, J. Chowdary, and G. Huang, 2016: Indo-western Pacific Ocean capacitor and coherent climate anomalies in post-ENSO summer: A review. Adv. Atmos. Sci., 33, 411–432, https://doi.org/10.1007/s00376-015-5192-6.
Yang, Y., S.-P. Xie, L. Wu, Y. Kosaka, and J. Li, 2018: ENSO forced and local variability of north tropical Atlantic SST: Model simulations and biases. Climate Dyn., 51, 4511–4524, https://doi.org/10.1007/s00382-017-3679-9.
Zebiak, S., 1993: Air–sea interaction in the equatorial Atlantic region. J. Climate, 6, 1567–1586, https://doi.org/10.1175/1520-0442(1993)006<1567:AIITEA>2.0.CO;2.