1. Introduction
The South China Sea (SCS) summer monsoon (SCSSM) is one of the most important subsystems of the Asian summer monsoon (Tao and Chen 1987; Ding et al. 2004), first emerging over the SCS and then propagating northward and westward to form the East Asian summer monsoon and South Asian summer monsoon (Hu et al. 2003; Ding et al. 2004). According to previous studies, a strong SCSSM is always accompanied by strong convection and heavy rainfall over southern China, as well as the countries around the SCS (Li and Zhang 1999; Li and Qu 2000; Liang et al. 2007). Therefore, further studies on the SCSSM are of great importance for the prediction of summer precipitation over SCS-related areas.
The strength of the SCSSM is influenced by many atmospheric and oceanic climate systems (Zhou and Chan 2007; B. Wang et al. 2009; Ding et al. 2010; Mao et al. 2011; S. Chen et al. 2015), among which El Niño–Southern Oscillation (ENSO) is one of the most dominant (Wang et al. 2008; B. Wang et al. 2009; Ge et al. 2017). Typically, the mature phase of ENSO is locked to a particular season, specifically boreal winter. Therefore, El Niño (La Niña) events usually develop (decay) in the seasons before (after) winter. An western North Pacific anomalous anticyclone (cyclone), which is related to the thermal conditions over the central and eastern Pacific during the decaying phase of El Niño (La Niña), has a considerable effect on the transportation of water vapor from the Pacific to continental Asia in summer, influencing the variability of the SCSSM (Wang et al. 2000; Wu et al. 2010; Zhang et al. 2015). Stronger ENSO tends to favor a longer persistence of the circulation anomaly over the western North Pacific during the decaying phase of ENSO (Wang et al. 2000, 2008; B. Wang et al. 2009), and thus the influence of the decaying phase of ENSO on the SCSSM is highly dependent on the strength of ENSO events. Usually, relatively larger ENSO variability leads to a stronger relationship between ENSO and the SCSSM (Lu et al. 2008; Fan and Fan 2017).
Importantly, the relationship between ENSO and the SCSSM is an interactive process. ENSO not only impacts the SCSSM during its decaying phase, but is also influenced by the summer monsoon–related anomalous tropical westerly during its developing phase (Huang et al. 1998; Kirtman and Shukla 2000; Huang et al. 2001; Wu and Kirtman 2003). Wu and Kirtman (2003) pointed out that, although the monsoon-induced wind anomalies are smaller than those directly associated with the development of El Niño, they can be amplified and propagate eastward through ongoing air–sea interaction in the Pacific. Strong East Asian summer monsoon could strengthen the low-level cyclonic anomaly and associated westerly flow over the tropical western Pacific, speeding up the eastward-propagating warm Kelvin waves and influencing the ENSO cycle (Huang et al. 1998; Huang et al. 2001). It has also been noted that La Niña events experience much weaker impacts from monsoon anomalies (Wu and Kirtman 2003).
Additionally, the relationship between ENSO and the SCSSM is unstable. Li and Ting (2015) suggested that the ENSO–monsoon relationship was mainly modulated by natural processes during the twentieth century. Large-scale climate systems, for instance the Atlantic multidecadal oscillation (AMO), play important roles in the modulation of ENSO and monsoon variability (Chan and Zhou 2005; Dong et al. 2006; Lu et al. 2006; Lu et al. 2008; He et al. 2013; Geng et al. 2017; Fan and Fan 2017). The AMO, with a long periodicity of 50–70 yr, has been emphasized as the “pacemaker” of climate in the Northern Hemisphere (Kerr 2000). Persistent basin-scale sea surface temperature (SST) anomalies in the North Atlantic have significant impacts not only on the climate over the North Atlantic Ocean and its surroundings (Sutton and Hodson 2005; Knight et al. 2006), but also on the climate in remote areas such as the tropical Pacific and East Asia (Zhang and Delworth 2005; Dong et al. 2006; Lu et al. 2006; Li et al. 2009; Y. Wang et al. 2009; Wang et al. 2013; Guo et al. 2016; Hao and He 2017).
The AMO impacts the climate of East Asia mainly via four processes. First, eastward-propagating wave trains, in response to the anomalous heating over the North Atlantic, could influence the East Asian winter monsoon (Geng et al. 2017; Zhou et al. 2015), as well as the rainfall over India (Zhang and Delworth 2006; Li et al. 2008; Y. Wang et al. 2009; Luo et al. 2011) and East Asia (Orsolini et al. 2015; Zhu et al. 2011). Second, the anomalous SST over the North Atlantic modulates the intertropical convergence zone and Walker circulation over the Atlantic and the Pacific, leading to anomalous SST and circulation over the western tropical Pacific through a series of air–sea interactions (Zhang and Delworth 2005; Dong et al. 2006; Lu et al. 2006; Sun et al. 2017). Third, the AMO may modulate the frequency and intensity of the North Atlantic Oscillation (Goswami et al. 2006), which has been proven to be closely related with the East Asian climate (Zhao et al. 2012; Zhou and Wang 2015; Chen et al. 2017). And fourth, the AMO can also modulate the variability of ENSO (Dong et al. 2006; Lu et al. 2006; Lu et al. 2008; W. Chen et al. 2015; Geng et al. 2017). It was found that anomalous SSTs over the Atlantic influence the Pacific through atmospheric and oceanic processes (Timmermann et al. 2005; Zhang and Delworth 2005, 2007), and the atmospheric feedbacks could spread the influence much more rapidly and efficiently than oceanic processes (Goswami et al. 2006; Li and Bates 2007; Y. Wang et al. 2009; Wang et al. 2011; Zhou and Wu 2016; Hao and He 2017). Anomalous Atlantic Ocean SSTs modulate both the mean climate and climate variability of the SST over the tropical Pacific via atmospheric teleconnections (Dong et al. 2006), and the monsoon–ENSO relationship intensifies during negative phases of the AMO because of enhanced ENSO variability (Lu et al. 2008; W. Chen et al. 2015).
Given the oceanic and atmospheric anomalies induced by the AMO, we hypothesize that the AMO exerts an important influence on the interaction between ENSO and the SCSSM. Previous investigations have focused mainly on the changing relationship between the decaying phase of ENSO and the SCSSM (Wang et al. 2008; Lu et al. 2008; B. Wang et al. 2009; W. Chen et al. 2015; Fan et al. 2016). However, the feedbacks of monsoon circulation to the development of ENSO during different AMO phases need to be further studied. B. Wang et al. (2009) indicated that the SCSSM is primarily correlated with the developing phase of ENSO before the late 1970s, but with the decaying phase of ENSO after the late 1970s. To date, however, the physical mechanism behind this changing relationship has not been comprehensively demonstrated. Considering the insufficiency of investigations on how the relationship between different phases of ENSO and the SCSSM changes during different AMO phases, the present study focuses on three main issues: First, what role does the AMO play in the interaction between the SCSSM and the different phases of ENSO? Second, if the AMO does change the interaction between ENSO and the SCSSM, what is the mechanism responsible for this modulation? And third, can climate models reasonably simulate the modulation of the interaction between ENSO and the SCSSM by the AMO?
The rest of this paper is structured as follows: The datasets and methods used are described in section 2. An analysis of the ENSO–SCSSM relationship during different phases of the AMO is made in section 3. Further investigation of the mechanism responsible for the changing relationship is given in section 4, and model simulation results are presented in section 5. Finally, a summary and discussion of the study is provided in section 6.
2. Datasets and methods
a. Reanalysis data
Because the AMO has a periodicity of 50–70 yr, long-term datasets are needed to investigate the modulation of ENSO–SCSSM interaction by the AMO. In this study, we use Twentieth Century Reanalysis version 2c data (Compo et al. 2006) during 1851–2014 (https://www.esrl.noaa.gov/psd/data/gridded/data.20thC_ReanV2c.html). The SST data are from the NOAA ERSST version 4 dataset, from 1854 to 2014 (Huang et al. 2015; https://www.esrl.noaa.gov/psd/data/gridded/data.noaa.ersst.v4.html). Both these datasets have a resolution of 2° × 2°.
b. Simulation data
The simulation datasets used in this study are the output from phase 5 of the Coupled Model Intercomparison Project (CMIP5; Taylor et al. 2012). After calculating the correlation coefficients (CCs) between the model-simulated and observed SCSSM indices, 6 out of 32 models showing reasonable simulation ability with respect to the SCSSM are selected for further investigation. Each of them covers the period from 1900 to 2005. For brevity, we use the model name in Table 1 to indicate the model. (Further details can be found at https://esgf-node.llnl.gov/search/cmip5/.)
Details of the CMIP5 models used in this study, along with the CCs between the model-simulated and reanalysis SCSSMIs during 1900–2005 (according to the Student’s t test, boldface values are statistically significant at the 90% confidence level).
c. Methods
In this study, winter refers to the monthly average of December in the previous year and January and February of the current year (DJF); spring refers to the monthly average of March, April, and May (MAM); summer refers to the monthly average of June, July, and August (JJA); and autumn refers to the monthly average of September, October, and November (SON).
A detrended annual mean AMO index from 1856 to 2017 computed from the area-weighted average over the North Atlantic (0°–70°N) is derived from data available online at https://www.esrl.noaa.gov/psd/data/timeseries/AMO. Positive and negative phases of the AMO (+AMO and −AMO, respectively) are selected based on the smoothed (15-yr running average) AMO index. Periods with positive (negative) smoothed AMO index are defined as positive (negative) phases of the AMO. Based on this criterion, two negative AMO periods (1904–29 and 1968–93) and three positive AMO periods (1871–96, 1938–63, and 1995–2014) are defined. It should be noted that although the smoothed (15-yr running average) AMO index ends in 2010, we consider the years 2011, 2012, 2013, and 2014 to all be in a positive AMO phase because the unsmoothed AMO index from 2011 to 2014 is positive. Additionally, after illustrating the AMO index in Fig. 1, because of the poor quality of data before 1900 we remove the data before 1900 for the analyses in the following sections.
The detrended, smoothed AMO index (dashed line) and the running correlation with a 19-yr window between the JJA Niño-3.4 index and SCSSMI (bars) during 1864–2005, in which the solid line (x axis) represents the 0.1 significance level according to the Student’s t test.
Citation: Journal of Climate 31, 8; 10.1175/JCLI-D-17-0448.1
The sSCSSM and wSCSSM events, as characterized by the SCSSMI being greater than 1.00 (less than −1.00), and the El Niño (La Niña) events, as characterized by the Niño-3.4 index being greater than 1.00 (less than −1.00), during +AMO and −AMO.
The Niño-3.4 index is calculated through the monthly mean of the SST anomalies over the Niño-3.4 region (5°S–5°N, 170°–120°W). An El Niño (a La Niña) event is defined as the detrended normalized winter Niño-3.4 index being higher (lower) than 1.00 (−1.00). The years with winter ENSO events under different phases of the AMO are listed in Table 2.
Since the anomalies associated with positive and negative phases of the AMO tend to have opposite polarities and ENSO analogously, composite analysis is carried out in this study mainly by using the positive phases minus the negative phases. Composite analysis between strong and weak SCSSM years is also used, to amplify the anomalies corresponding to the SCSSM. Years with a normalized AMO index greater (less) than 1.00 (−1.00) are defined as positive (negative) AMO years. The composite between positive and negative AMO phases is based on 26 positive AMO years during positive phases and 30 negative AMO years during negative phases. The composite analyses for ENSO and the SCSSM are based on the information listed in Table 2.
The two-tailed F test is used to test the statistical significance of the difference between standard deviations of two periods. The Student’s t test is used to verify the statistical significance of the correlation and composite difference.
3. Unstable relationship between ENSO and the SCSSM against different AMO backgrounds
The 19-yr running correlation coefficient between the JJA Niño-3.4 and SCSSM indices during 1864–2005, as well as the detrended, smoothed annual mean AMO index, are shown in Fig. 1. The time series end in 2005 because the 19-yr running correlation ends in 2005. The correlation coefficients between the JJA Niño-3.4 index and the SCSSM index during the negative AMO phases (1904–29 and 1968–93) do not pass the significance test, whereas the coefficients for the positive AMO periods 1871–96, 1938–63, and 1995–2014 are 0.50, 0.54, and 0.57, respectively, all passing the 95% confidence level according to the Student’s t test. The result demonstrates that the positive relationship between the SCSSM and the summer Niño-3.4 index is significant (nonsignificant) during positive (negative) phases of the AMO.
Because the summer could be under the influence of either the developing or the decaying phase of ENSO, it is hard to determine which phase of ENSO has the changing relationship with the SCSSM, as shown in Fig. 1. Therefore, the correlations between the SCSSM index and the Niño-3.4 index in different seasons are calculated for positive and negative phases of the AMO. As shown in Table 3, the correlations between the SCSSMI and the Niño-3.4 index in summer [JJA(0)] and the following seasons [SON(0) and D(0)JF(1)] are positive and significant during positive phases of the AMO, but are nonsignificant during negative phases of the AMO. This suggests that the SCSSM has a significant positive correlation with ENSO development during positive phases of the AMO only.
Correlation between the SCSSMI from JJA(0) and Niño-3.4 index from DJF(0) to DJF(1), during −AMO and +AMO. According to the Student’s t test, boldface values are statistically significant at the 95% confidence level.
Table 3 also illustrates that the correlations between the SCSSMI in summer [JJA(0)] and the Niño-3.4 index in the preceding seasons [D(−1)JF(0) and MAM(0)] are negative and significant during negative phases of the AMO, but are nonsignificant during positive phases of the AMO. In other words, the SCSSM is influenced by the decaying phase of ENSO during negative phases of the AMO only.
In summary, two results can be concluded from Fig. 1 and Table 3. First, during negative phases of the AMO, the decaying phase of ENSO has a negative influence on the following SCSSM, but such a relationship disappears during positive AMO phases. Second, during positive phases of the AMO, the SCSSM has a significant positive correlation with the subsequent ENSO development, whereas such a relationship cannot be observed during negative phases of the AMO.
4. Mechanistic analysis
a. Influence of the decaying phase of ENSO on the SCSSM
As mentioned in section 1, the preceding ENSO and the SCSSM are connected via the circulation anomaly lasting from winter to summer over the western North Pacific (Kug et al. 2006; Xie et al. 2009; Wu et al. 2010; Jiang et al. 2013). In general, a stronger ENSO tends to favor a longer maintenance of the circulation anomaly during the decaying phase of ENSO (Wang et al. 2000, 2008; B. Wang et al. 2009), and thus the impact of the ENSO decaying phase on the SCSSM is stronger. Studies have pointed out that the variability of ENSO is larger during negative phases of the AMO than during positive phases (Dong et al. 2006). As shown in Fig. 2, compared with negative phases of the AMO, the SST over the North Atlantic, North Pacific, and tropical western Pacific increases (Fig. 2a) and the variability of the SST over the central and eastern tropical Pacific decreases (Fig. 2b) during positive phases of the AMO. Moreover, the number of ENSO events during negative phases of the AMO is larger than during positive phases (Table 2). The results indicate that the variability of ENSO is larger during negative phases of the AMO than during positive phases. The change in ENSO variability is attributable to the AMO-induced wind anomaly over the tropical Pacific and the subsequent air–sea interaction (Dong and Sutton 2002; Zhang and Delworth 2005; Dong et al. 2006). The easterly (westerly) wind anomalies over the western tropical Pacific and westerly (easterly) wind anomalies over the eastern tropical Pacific induced by positive (negative) phases of the AMO (Fig. 2c) could deepen (reduce) the thermocline in the tropical Pacific (Dong et al. 2006), reducing (enhancing) the variability of ENSO (Dong et al. 2006; Lu et al. 2008; W. Chen et al. 2015). Consequently, during positive phases of the AMO, the western North Pacific anomalous anticyclone (cyclone) during the decaying phase of El Niño (La Niña) only persists until spring (Figs. 3a,b) and then disappears in summer (Fig. 3c). Thus, the influence of El Niño (La Niña) on the SCSSM is nonsignificant in summer. On the contrary, during negative phases of the AMO, the western North Pacific anomalous anticyclone (cyclone) during the decaying phase of El Niño (La Niña) persists into summer (Figs. 3d–f), significantly influencing the SCSSM. The SCSSM index used in this study represents the strength of a cyclonic or an anticyclonic anomaly over the SCS: a positive (negative) SCSSMI represents a cyclonic (an anticyclonic) anomaly over the SCS and hence a strong (weak) SCSSM. Therefore, the western North Pacific anomalous anticyclone (cyclone) associated with the decaying phase of El Niño (La Niña) weakens (strengthens) the SCSSM. This may explain why the negative correlation between the preceding ENSO and the SCSSM is significant (nonsignificant) during negative (positive) phases of the AMO.
Composite difference between +AMO and −AMO in JJA for (a) SST (K), (b) standard deviation of SST, and (c) 850-hPa wind (m s−1). The letters A and C in (c) represent the anticyclone and cyclone, respectively. According to the Student’s t test, the hatched regions in (a) and shaded regions in (c) are statistically significant at the 95% confidence level. According to the F test, the hatched regions in (b) are statistically significant at the 95% confidence level.
Citation: Journal of Climate 31, 8; 10.1175/JCLI-D-17-0448.1
Composite difference in the 850-hPa wind (m s−1) between El Niño (EL) and La Niña (LA) years during (a)–(c) +AMO and (d)–(f) −AMO in DJF, MAM, and JJA. The letter A represents the anticyclone. According to the Student’s t test, the light and dark gray shading indicates statistical significance at the 90% and 95% confidence level, respectively.
Citation: Journal of Climate 31, 8; 10.1175/JCLI-D-17-0448.1
b. Relationship between the SCSSM and the subsequent developing phase of ENSO
As discussed above, during positive phases of the AMO, the SCSSM has a significant positive correlation with the subsequent developing phase of ENSO, whereas the SCSSM is not significantly associated with ENSO development during negative phases of the AMO. To determine why the relationship between the SCSSM and the subsequent ENSO development changes with the transformation of the AMO background, two issues must be investigated. First, what is the key factor that connects the SCSSM with the developing phase of ENSO? Second, how is this key factor influenced by the different phases of the AMO?
It has been pointed out that the zonal wind anomaly over the western tropical Pacific is a key factor that connects the SCSSM and subsequent ENSO development (Huang et al. 2001; Huang et al. 1998). Corresponding to a strong SCSSM, a cyclonic anomaly dominates over the SCS. The westerly wind anomaly to the south of the anomalous cyclone weakens the tropical northeasterly trade winds. Hereafter, the oceanic warm Kelvin wave excited by the westerly wind stress anomaly, as well as the westerly wind anomaly itself, propagates eastward, causing a disturbance of the mixed layer depth in the equatorial eastern Pacific, which favors the occurrence of an El Niño event (Huang et al. 2001; Huang et al. 1998). It should be noted that La Niña events experience much weaker impacts from monsoon anomalies (Wu and Kirtman 2003). As shown in Table 2, 11 (4) strong SCSSM events with an averaged SCSSMI of 1.4 (1.3) occurred during positive (negative) phases of the AMO, indicating that the strength of the SCSSM is larger during positive phases of the AMO than during negative phases. During positive phases of the AMO, corresponding to a strong SCSSM, the cyclonic anomaly and the westerly wind anomaly over the western tropical Pacific are significant in summer (Fig. 4a). Moreover, the westerly wind anomaly can also be detected in the following autumn and winter (Figs. 4b and 4c, respectively), indicating the development and maturation of El Niño. On the contrary, during negative phases of the AMO, the westerly wind anomaly over the tropical western Pacific associated with the SCSSM is not statistically significant in summer and the following autumn and winter (Figs. 4d–f). In other words, during positive (negative) phases of the AMO, the bridge between the SCSSM and the subsequent development of El Niño, namely, the low-level westerly wind anomaly over the tropical western Pacific, is strong (weak). Therefore, the positive correlation between the SCSSM and the subsequent development of ENSO is significant (nonsignificant) during positive (negative) phases of the AMO.
Composite difference in the 850-hPa wind (m s−1) between sSCSSM and wSCSSM years during (a)–(c) +AMO and (d)–(f) −AMO in JJA, SON, and DJF(1). According to the Student’s t test, the gray shading indicates statistical significance at the 95% confidence level.
Citation: Journal of Climate 31, 8; 10.1175/JCLI-D-17-0448.1
There are predominantly two AMO-related pathways by which the SCSSM can enhance and thereafter strengthens the westerly wind anomaly. The first pathway is related to the anomalous eastward-propagating large-scale waves induced by the anomalous heating over the North Atlantic. As shown in Fig. 2a, during positive phases of the AMO, the SST over the North Atlantic is significantly higher than that during negative phases. Previous studies have pointed out that anomalous heating over the North Atlantic could generate an arching wave train (Li et al. 2008; Zhu et al. 2011), which could be the atmospheric response to the tropospheric diabatic heating over the warmer North Atlantic (Li et al. 2008) and anomalous transient eddy activity over the North Atlantic (Zhu et al. 2011). As shown in Fig. 5a, during positive phases of the AMO, the SCSSM-related summer atmospheric circulation is characterized by a wave train with a series of alternate anomalous high and low pressure centers, which generates from the extratropical North Atlantic and propagates on a “great circle route” through Eurasia to the SCS. Consequently, a low pressure anomaly at the lower level dominates over the SCS (Fig. 5a). Against this low pressure anomaly background, the enhanced SCSSM could bring about a stronger westerly wind anomaly (Fig. 4) over the western tropical Pacific. It should be noted that the Tibetan Plateau may also play an important role in the process. Anomalous wave activities at 200 hPa converge over the Tibetan Plateau (Fig. 5a), thus strengthening the South Asian high. Thereafter, the stronger than normal South Asian high is associated with stronger than normal northerly flow to the southeast of the high (figure not shown). The strengthened northerly could increase positive geostrophic vorticity advection, and further contribute to ascending motion at low levels (Ren et al. 2007). The anomalous low pressure at 700 hPa over the SCS (Fig. 5a) may be a result of enhanced ascending motion.
Composite difference between sSCSSM and wSCSSM years for JJA in terms of 200-hPa wave activity flux (vectors; m2 s−2) and 700-hPa geopotential height (contours; gpm) during (a) +AMO and (b) −AMO.
Citation: Journal of Climate 31, 8; 10.1175/JCLI-D-17-0448.1
In conclusion for the first pathway, during positive phases of the AMO, an arching wave train propagating from the North Atlantic to Southeast Asia leads to a low pressure anomaly over the SCS (Fig. 5a), which could strengthen the westerly wind anomaly associated with a strong SCSSM (Fig. 4a). During negative phases of the AMO, however, no such wave train exists. Figure 5b suggests that during negative phases of the AMO, the anomalous large-scale wave train propagates along the subpolar jet, exerting limited influence on the SCS region.
The second pathway is associated with enhanced Walker circulation. During positive phases of the AMO, a dipole SST anomaly pattern in the Atlantic (warmer in the northern Atlantic and colder in the southern Atlantic) leads to a northward shift of the Atlantic intertropical convergence zone (Dong and Sutton 2002), resulting in a local convergence (divergence) anomaly at the lower (upper) level over the tropical Atlantic, and correspondingly anomalous flow diverges (converges) at the upper (lower) level over the eastern tropical Pacific (Fig. 6). Thereafter, there is a divergence anomaly at the upper level (Fig. 6a) and convergence anomaly at the lower level (Fig. 6b) over the SCS, inducing a cyclonic anomaly over the SCS (Fig. 2c). Climatologically, the ascending branches of the Walker circulation rise over the Maritime Continent and tropical Atlantic, and the descending branch sinks over the eastern tropical Pacific. Therefore, it can be concluded that, compared with negative phases of the AMO, the Walker circulation is stronger during positive phases of the AMO, resulting in increased ascending motion and lower-level cyclonic anomaly over the SCS. Consequently, the westerly wind anomaly corresponding to the enhanced SCSSM tends to be stronger during positive phases of the AMO than during negative phases (Fig. 4).
Composite difference between +AMO and −AMO in terms of velocity potential (shading; 106 m2 s−1) and divergent wind (vectors; m s−1) at the (a) 200- and (b) 850-hPa level.
Citation: Journal of Climate 31, 8; 10.1175/JCLI-D-17-0448.1
In summary, during positive AMO phases, an anomalous arching wave train and anomalous Walker circulation result in a cyclonic and low pressure anomaly over the SCS, which in turn strengthens the SCSSM (Table 2 and Fig. 4). Thereafter, the westerly wind anomaly accompanied by the stronger SCSSM intensifies the eastward-propagating warm Kelvin waves and further influences the ENSO cycle (Huang et al. 1998). Consequently, the relationship between the SCSSM and the subsequent ENSO development is closer during positive phases of the AMO than during negative phases.
5. Model-simulated result
As shown in Table 1, after calculating the correlation coefficients between the simulated and observed SCSSM indices, six models showing a reasonable simulation of the SCSSM are selected for further investigation: CCSM4, EC-EARTH, GFDL CM3, GISS-E2-H, MPI-ESM-LR, and MPI-ESM-MR.
It can be concluded from section 4 that the AMO-induced SST anomalies, ENSO-related western North Pacific circulation anomaly and the SCSSM-associated westerly wind anomaly are the most critical factors in the interaction process between ENSO and the SCSSM during different AMO phases. Therefore, we further investigate how the six aforementioned state-of-the-art coupled models perform with respect to these factors.
The simulations derived from CCSM4, GFDL CM3, and MPI-ESM-MR show reasonable reproductions of the anomalous SST and low-level circulation induced by the AMO. During positive (negative) phases of the AMO, significant warm (cold) SST anomalies locate over the North Atlantic and the western North Pacific (Fig. 7). Besides, the correlation coefficients between the model-simulated and reanalysis AMO index during 1900–2014 are calculated. The result derived from MPI-ESM-MR (0.18) and GFDL CM3 (0.21) are statistically significant at the 90% confidence level, based on the Student’s t test. The significant easterly (westerly) wind anomaly over the western (eastern) tropical Pacific, the AMO-induced anticyclonic anomaly over the North Pacific (Fig. 8), and the cyclonic anomaly over the SCS (Figs. 8a,c) all closely resemble the AMO-induced atmospheric anomalies discussed above with the reanalysis data.
Composite difference in the SST for JJA (K) between +AMO and −AMO. Results are derived from (a) CCSM4, (b) GFDL CM3, and (c) MPI-ESM-MR. According to the Student’s t test, hatched regions are statistically significant at the 95% confidence level.
Citation: Journal of Climate 31, 8; 10.1175/JCLI-D-17-0448.1
Regression of the 850-hPa winds for JJA onto the AMO index during 1900–2005. Results are derived from (a) CCSM4, (b) GFDL CM3, and (c) MPI-ESM-MR. The letters A and C represent the anticyclone and cyclone, respectively. Based on the Student’s t test, shaded regions are statistically significant at the 95% confidence level.
Citation: Journal of Climate 31, 8; 10.1175/JCLI-D-17-0448.1
It was shown in section 4a that, during negative phases of the AMO, the El Niño–induced (La Niña induced) western North Pacific anomalous anticyclone (cyclone) persists in the summer of ENSO decaying years, whereas during positive phases it disappears in the summer. Therefore, the influence of ENSO decay on the SCSSM is stronger during negative phases of the AMO than positive phases. Both GFDL CM3 (Figs. 9a,b) and MPI-ESM-MR (Figs. 9c,d) can basically reproduce the significant (nonsignificant) western North Pacific anomalous anticyclone (cyclone) in the summer of the El Niño (La Niña) decaying phase during negative (positive) phases of the AMO. To be noted, the anticyclones extracted from the two model simulations (Fig. 9) shift eastward to the reanalysis results (Fig. 3), which might be related to the model capability (He and Zhou 2014).
Composite difference in the 850-hPa wind for JJA (m s−1) between the decaying JJA of El Niño (EL) and La Niña (LA) years during (a),(c) +AMO and (b),(d) −AMO. Results are derived from (a),(b) GFDL CM3 and (c),(d) MPI-ESM-MR. The letter A represents the anticyclone. According to the Student’s t test, the light and dark gray shading indicates statistical significance at the 90% and 95% confidence level, respectively.
Citation: Journal of Climate 31, 8; 10.1175/JCLI-D-17-0448.1
As mentioned in section 4b, the westerly wind anomaly over the western tropical Pacific is a crucial factor connecting the SCSSM and the subsequent ENSO development. The westerly wind anomalies associated with an enhanced SCSSM are stronger during positive phases of the AMO than negative phases. These characteristics are also simulated reasonably well by GFDL CM3 and MPI-ESM-MR. The westerly wind anomalies associated with an enhanced SCSSM are stronger and cover a larger area over the tropical Pacific during positive phases of the AMO than negative phases (Fig. 10).
Composite difference in 850-hPa wind from JJA (m s−1) between sSCSSM and wSCSSM years during (a),(c) +AMO and (b),(d) −AMO. Results are derived from (a),(b) GFDL CM3 and (c),(d) MPI-ESM-MR. According to the Student’s t test, the light and dark gray shading indicates statistical significance at the 90% and 95% confidence level, respectively.
Citation: Journal of Climate 31, 8; 10.1175/JCLI-D-17-0448.1
In summary, GFDL CM3 and MPI-ESM-MR can simulate the main features of the AMO-induced anomalous SST and circulation reasonably. In addition, both can reproduce the interaction between ENSO and the SCSSM during the different AMO phases. On one hand, the two models basically capture the ENSO-related summer circulation anomalies over the SCS to the western North Pacific during negative phases of the AMO. On the other hand, their simulation of the strong (weak) SCSSM-associated westerly wind anomaly over the western tropical Pacific during positive (negative) phases of the AMO is in agreement with the reanalysis results. Therefore, it can be considered that the SCSSM–ENSO interaction is intelligently reproduced in these models. By comparing each model’s results, it is found that only those models that appropriately simulate the AMO-induced anomalous SST and circulation can also reasonably simulate the interaction between ENSO and the SCSSM against different AMO backgrounds. This further emphasizes the important roles of the AMO-induced anomalous SST and circulation, which supports the views proposed in section 4.
Importantly, the strong performance of GFDL CM3 and MPI-ESM-MR could be attributable to their rational simulations of the SST and oceanic processes in the North Atlantic. For instance, the ocean component of MPI-ESM-MR can simulate the North Atlantic overturning circulation in real time (Rahmstorf et al. 2015), and the simulation of the SST and sea ice anomaly over the Arctic Ocean and North Atlantic have both been significantly improved in GFDL CM3 (Griffies et al. 2011).
6. Summary and discussion
The interaction between ENSO and the SCSSM in the different phases of the AMO is investigated based on reanalysis data and simulations from CMIP5 models. The results show that both the influence of the ENSO decaying phase on the SCSSM and the relationship between the SCSSM and the subsequent ENSO development change with a transformation of the AMO background (Fig. 11).
Schematic diagram showing the interaction between ENSO and the SCSSM during negative phases of the AMO (colder than normal North Atlantic) and positive phases of the AMO (warmer than normal North Atlantic). (a) A colder than normal North Atlantic can enhance the variability of ENSO. Consequently, the anomalous anticyclone (cyclone) over the western North Pacific induced by El Niño (La Niña) can persist in the summer of ENSO decaying year, weakening (strengthening) the SCSSM. (During positive phases of the AMO, ENSO with weak variability cannot significantly influence the SCSSM.) (b) A warmer than normal North Atlantic can strengthen low pressure and cyclonic anomalies over the SCS. Consequently, a stronger than normal SCSSM is accompanied by significant westerly wind anomalies over the western Pacific, which favors the development of El Niño events. (During negative phases of the AMO, the SCSSM-induced westerly wind anomalies are rather weak, having a nonsignificant relationship with ENSO development.)
Citation: Journal of Climate 31, 8; 10.1175/JCLI-D-17-0448.1
An analysis of low-level circulation anomalies between positive and negative phases of the AMO shows that the easterly (westerly) wind anomalies over the western Pacific and the westerly (easterly) wind anomalies over the eastern tropical Pacific induced by positive (negative) phases of the AMO (Fig. 2c) will deepen (reduce) the thermocline in the Pacific, thus reducing (enhancing) the variability of ENSO (Dong et al. 2006). Consequently, during positive phases of the AMO, the circulation anomalies induced by the ENSO decaying phase disappear in summer (Fig. 3c), having no significant influence on the SCSSM. On the contrary, during negative phases of the AMO, the El Niño–induced (La Niña induced) western North Pacific anomalous anticyclone (cyclone) lasts during winter (Fig. 3d), spring (Fig. 3e), and summer (Fig. 3f), significantly weakening (strengthening) the SCSSM. Therefore, the negative correlation between the decaying phase of ENSO and the SCSSM is significant (nonsignificant) during negative (positive) phases of the AMO.
During positive phases of the AMO, anomalous wave activities (Fig. 5) and enhanced Walker circulation (Fig. 6) induced by the warmer North Atlantic favors anomalous low pressure and cyclonic circulation over the SCS. Consequently, the westerly wind anomaly over the western tropical Pacific corresponding to an enhanced SCSSM is stronger during positive phases of the AMO than during negative phases. Previous studies have demonstrated that westerly wind anomalies can intensify the eastward-propagating warm Kelvin wave, contributing to subsequent El Niño development (Huang et al. 1998; Huang et al. 2001; Kirtman and Shukla 2000). It has also been noted that La Niña events experience much weaker impacts from monsoon anomalies (Wu and Kirtman 2003). Therefore, during positive phases of the AMO, an enhanced SCSSM accompanied by strong westerly wind anomalies over the western tropical Pacific favors the development of subsequent El Niño development. Consequently, the positive relationship between the SCSSM and the subsequent development of ENSO is stronger during positive phases of the AMO than during negative phases.
To further support the results from the reanalysis data, six state-of-the-art climate models from CMIP5 showing a reasonable simulation of the SCSSMI are employed. Among them, CCSM4, GFDL CM3, and MPI-ESM-MR can simulate the SST and circulation anomalies induced by the AMO reasonably well. GFDL CM3 and MPI-ESM-MR can further reproduce the changing relationship between the SCSSM and the different phases of ENSO during positive and negative AMO phases. On one hand, the models can reasonably simulate the long-lasting western North Pacific anomalous anticyclone or cyclone during the decaying phase of ENSO for negative phases of the AMO; on the other hand, the strong SCSSM and enhanced westerly wind anomaly over the western tropical Pacific during positive phases of the AMO are both simulated reasonably well. The good performance of GFDL CM3 and MPI-ESM-MR might be attributable to their reasonable simulation of the SST and oceanic processes in the North Atlantic.
It should be noted that conclusions in this paper are based on correlation and composite analysis. Considering that the correlation does not necessarily imply the causation, we should be alert to possible overstretching of the results. Besides, it is well known that the AMO cycle has a long periodicity of up to 70 yr, but reliable datasets cover only around 100 years and the history of high-quality data only stretches back about 40 years. Therefore, the data span may be a major source of uncertainty in this study. In addition, the observed AMO signal is superimposed on global warming, and it is difficult to separate them effectively with statistical methods. Therefore, in the future, model experiments that exclude the global warming trend will be carried out to verify the theory put forward in this paper.
Acknowledgments
This research was jointly supported by the National Natural Science Foundation of China (Grant 41325018), the National Key R&D Program of China (2017YFA0603802), the National Natural Science Foundation of China (Grant 41421004), and the CAS/SAFEEA International Partnership Program for Creative Research Team “Regional environmental high resolution numerical simulation.” We thank the reviewers for their insightful comments and suggestions.
REFERENCES
Chan, J. C. L., and W. Zhou, 2005: PDO, ENSO and the early summer monsoon rainfall over South China. Geophys. Res. Lett., 32, L08810, https://doi.org/10.1029/2004GL022015.
Chen, J., Z. Wen, R. Wu, X. Wang, C. He, and Z. Chen, 2017: An interdecadal change in the intensity of interannual variability in summer rainfall over southern China around early 1990s. Climate Dyn., 48, 191–207, https://doi.org/10.1007/s00382-016-3069-8.
Chen, S., W. Chen, and R. Wu, 2015: An interdecadal change in the relationship between boreal spring Arctic Oscillation and the East Asian summer monsoon around the early 1970s. J. Climate, 28, 1527–1542, https://doi.org/10.1175/JCLI-D-14-00409.1.
Chen, W., R. Lu, and B. Dong, 2015b: Intensified anticyclonic anomaly over the western North Pacific during El Niño decaying summer under a weakened Atlantic thermohaline circulation. J. Geophys. Res. Atmos., 119, 13 637–13 650, https://doi.org/10.1002/2014JD022199.
Compo, G. P., J. S. Whitaker, and P. D. Sardeshmukh, 2006: Feasibility of a 100-year reanalysis using only surface pressure data. Bull. Amer. Meteor. Soc., 87, 175–190, https://doi.org/10.1175/BAMS-87-2-175.
Ding, R., K. J. Ha, and J. Li, 2010: Interdecadal shift in the relationship between the East Asian summer monsoon and the tropical Indian Ocean. Climate Dyn., 34, 1059–1071, https://doi.org/10.1007/s00382-009-0555-2.
Ding, Y., C. Li, and Y. Liu, 2004: Overview of the South China Sea monsoon experiment. Adv. Atmos. Sci., 21, 343–360, https://doi.org/10.1007/BF02915563.
Dong, B., and R. T. Sutton, 2002: Adjustment of the coupled ocean–atmosphere system to a sudden change in the thermohaline circulation. Geophys. Res. Lett., 29, 1728, https://doi.org/10.1029/2002GL015229.
Dong, B., R. T. Sutton, and A. A. Scaife, 2006: Multidecadal modulation of El Niño–Southern Oscillation (ENSO) variance by Atlantic Ocean sea surface temperatures. Geophys. Res. Lett., 33, L08705, https://doi.org/10.1029/2006GL025766.
Fan, Y., and K. Fan, 2017: Pacific decadal oscillation and the decadal change in the intensity of the interannual variability of the South China Sea summer monsoon. Atmos. Ocean. Sci. Lett., 10, 162–167, https://doi.org/10.1080/16742834.2016.1256189.
Fan, Y., K. Fan, and B. Tian, 2016: Has the prediction of the South China Sea summer monsoon improved since the late 1970s? J. Meteor. Res., 30, 833–852, https://doi.org/10.1007/s13351-016-6052-8.
Ge, F., X. Zhi, Z. A. Babar, W. Tang, and P. Chen, 2017: Interannual variability of the summer monsoon precipitation over the Indochina Peninsula in association with ENSO. Theor. Appl. Climatol., 128, 523–531, https://doi.org/10.1007/s00704-015-1729-y.
Geng, X., W. Zhang, M. F. Stuecker, P. Liu, F. Jin, and G. Tan, 2017: Decadal modulation of the ENSO–East Asian winter monsoon relationship by the Atlantic multidecadal oscillation. Climate Dyn., 49, 2531–2544, https://doi.org/10.1007/s00382-016-3465-0.
Goswami, B. N., M. S. Madhusoodanan, C. P. Neema, and D. Sengupta, 2006: A physical mechanism for North Atlantic SST influence on the Indian summer monsoon. Geophys. Res. Lett., 33, L02706, https://doi.org/10.1029/2005GL024803.
Griffies, S. M., and Coauthors, 2011: The GFDL CM3 coupled climate model: Characteristics of the ocean and sea ice simulations. J. Climate, 24, 3520–3544, https://doi.org/10.1175/2011JCLI3964.1.
Guo, Y., J. Li, J. Feng, F. Xie, C. Sun, and J. Zheng, 2016: The multidecadal variability of the asymmetric mode of the boreal autumn Hadley circulation and its link to the Atlantic multidecadal oscillation. J. Climate, 29, 5625–5641, https://doi.org/10.1175/JCLI-D-15-0025.1.
Hao, X., and S. He, 2017: Combined effect of ENSO-like and Atlantic multidecadal oscillation SSTAs on the interannual variability of the East Asian winter monsoon. J. Climate, 30, 2697–2716, https://doi.org/10.1175/JCLI-D-16-0118.1.
He, C., and T. Zhou, 2014: The two interannual variability modes of the western North Pacific subtropical high simulated by 28 CMIP5–AMIP models. Climate Dyn., 43, 2455–2469, https://doi.org/10.1007/s00382-014-2068-x.
He, S., H. Wang, and J. Liu, 2013: Changes in the relationship between ENSO and Asia–Pacific midlatitude winter atmospheric circulation. J. Climate, 26, 3377–3393, https://doi.org/10.1175/JCLI-D-12-00355.1.
Hu, F., J. Yang, Y. Li, and S. Qu, 2003: Some characteristics of air–sea exchange for the SCS summer monsoon during the SCSMEX in 1998. J. Meteor. Res., 17, 155–170, http://www.cmsjournal.net:8080/Jweb_jmr/EN/abstract/abstract1544.shtml.
Huang, B., and Coauthors, 2015: Extended reconstructed sea surface temperature version 4 (ERSST.v4). Part I: Upgrades and intercomparisons. J. Climate, 28, 911–930, https://doi.org/10.1175/JCLI-D-14-00006.1.
Huang, R., X. Zang, R. Zhang, and J. Chen, 1998: The westerly anomalies over the tropical Pacific and their dynamical effect on the ENSO cycles during 1980–1994. Adv. Atmos. Sci., 15, 135–151, https://doi.org/10.1007/s00376-998-0035-3.
Huang, R., R. Zhang, and B. Yan, 2001: Dynamical effect of the zonal wind anomalies over the tropical western Pacific on ENSO cycles. Sci. China Earth Sci., 44, 1089–1098, https://doi.org/10.1007/BF02906865.
Jiang, X., S. Yang, J. Li, Y. Li, H. Hu, and Y. Lian, 2013: Variability of the Indian Ocean SST and its possible impact on summer western North Pacific anticyclone in the NCEP Climate Forecast System. Climate Dyn., 41, 2199–2212, https://doi.org/10.1007/s00382-013-1934-2.
Kerr, R. A., 2000: A North Atlantic climate pacemaker for the centuries. Science, 288, 1984–1986, https://doi.org/10.1126/science.288.5473.1984.
Kirtman, B. P., and J. Shukla, 2000: Influence of the Indian summer monsoon on ENSO. Quart. J. Roy. Meteor. Soc., 126, 213–239, https://doi.org/10.1002/qj.49712656211.
Knight, J. R., C. K. Folland, and A. A. Scaife, 2006: Climate impacts of the Atlantic multidecadal oscillation. Geophys. Res. Lett., 33, L17706, https://doi.org/10.1029/2006GL026242.
Kug, J., T. Li, S. An, I. Kang, J. Luo, S. Masson, and T. Yamagata, 2006: Role of the ENSO–Indian Ocean coupling on ENSO variability in a coupled GCM. Geophys. Res. Lett., 33, L09710, https://doi.org/10.1029/2005GL024916.
Li, C., and L. Zhang, 1999: Summer monsoon activities in the South China Sea and its impacts. Sci. Meteor. Sin., 23, 257–266.
Li, C., and X. Qu, 2000: Large scale atmospheric circulation evolutions associated with summer monsoon onset in the South China Sea. Chin. J. Atmos. Sci., 24, 1–14.
Li, S., and G. T. Bates, 2007: Influence of the Atlantic multidecadal oscillation on the winter climate of East China. Adv. Atmos. Sci., 24, 126–135, https://doi.org/10.1007/s00376-007-0126-6.
Li, S., J. Perlwitz, X. Quan, and M. P. Hoerling, 2008: Modelling the influence of North Atlantic multidecadal warmth on the Indian summer rainfall. Geophys. Res. Lett., 35, L05804, https://doi.org/10.1029/2007GL032901.
Li, S., Y. Wang, and Y. Gao, 2009: A review of the researches on the Atlantic multidecadal oscillation (AMO) and its climate influence. Trans. Atmos. Sci., 32, 458–465.
Li, X., and M. Ting, 2015: Recent and future changes in the Asian monsoon–ENSO relationship: Natural or forced? Geophys. Res. Lett., 42, 3502–3512, https://doi.org/10.1002/2015GL063557.
Liang, J., S. Yang, C. H. Li, and X. Li, 2007: Long-term changes in the South China Sea summer monsoon revealed by station observations of the Xisha Islands. J. Geophys. Res., 112, D10104, https://doi.org/10.1029/2006JD007922.
Lu, R., B. Dong, and H. Ding, 2006: Impact of the Atlantic multidecadal oscillation on the Asian summer monsoon. Geophys. Res. Lett., 33, L24701, https://doi.org/10.1029/2006GL027655.
Lu, R., W. Chen, and B. Dong, 2008: How does a weakened Atlantic thermohaline circulation lead to an intensification of the ENSO–South Asian summer monsoon interaction? Geophys. Res. Lett., 35, L08706, https://doi.org/10.1029/2008GL033394.
Luo, F., S. Li, and T. Furevik, 2011: The connection between the Atlantic multidecadal oscillation and the Indian summer monsoon in Bergen climate model version 2.0. J. Geophys. Res., 116, D19117, https://doi.org/10.1029/2011JD015848.
Mao, J., J. C. L. Chan, and G. Wu, 2011: Interannual variations of early summer monsoon rainfall over south China under different PDO backgrounds. Int. J. Climatol., 31, 847–862, https://doi.org/10.1002/joc.2129.
Orsolini, Y. J., L. Zhang, D. H. W. Peters, K. Fraedrich, X. Zhu, A. Schneidereit, and B. van den Hurk, 2015: Extreme precipitation events over north China in August 2010 and their link to eastward‐propagating wave‐trains across Eurasia: Observations and monthly forecasting. Quart. J. Roy. Meteor. Soc., 141, 3097–3105, https://doi.org/10.1002/qj.2594.
Rahmstorf, S., J. E. Box, G. Feulner, M. E. Mann, A. Robinson, S. Rutherford, and E. J. Schaffernicht, 2015: Exceptional twentieth-century slowdown in Atlantic Ocean overturning circulation. Nat. Climate Change, 5, 475–480, https://doi.org/10.1038/nclimate2554.
Ren, R., Y. Liu, and G. Wu, 2007: Impact of South Asia high on the short-term variation of the subtropical anticyclone over western Pacific in July 1998. Acta Meteor. Sin., 65, 193–197.
Sun, C., F. Kucharski, J. Li, F. Jin, I. Kang, and R. Ding, 2017: Western tropical Pacific multidecadal variability forced by the Atlantic multidecadal oscillation. Nat. Commun., 8, 15998, https://doi.org/10.1038/ncomms15998.
Sutton, R. T., and D. L. R. Hodson, 2005: Atlantic Ocean forcing of North American and European summer climate. Science, 309, 115–118, https://doi.org/10.1126/science.1109496.
Takaya, K., and H. Nakamura, 2001: A formulation of a phase-independent wave-activity flux of stationary and migratory quasigeostrophic eddies on a zonally varying basic flow. J. Atmos. Sci., 58, 608–627, https://doi.org/10.1175/1520-0469(2001)058<0608:AFOAPI>2.0.CO;2.
Tao, S., and L. Chen, 1987: A review of recent research on the East Asian summer monsoon in China. Monsoon Meteorology, C. P. Chang and T. N. Krishnamurti, Eds., Oxford University Press, 60–92.
Taylor, K. E., R. J. Stouffer, and G. A. Meehl, 2012: A summary of the CMIP5 experiment design. Bull. Amer. Meteor. Soc., 93, 485–498, https://doi.org/10.1175/BAMS-D-11-00094.1.
Timmermann, A., S. I. An, U. Krebs, and H. Goosse, 2005: ENSO suppression due to weakening of the North Atlantic thermohaline circulation. J. Climate, 18, 3122–3139, https://doi.org/10.1175/JCLI3495.1.
Wang, B., R. Wu, and X. Fu, 2000: Pacific–East Asian teleconnection: How does ENSO affect East Asian climate? J. Climate, 13, 1517–1536, https://doi.org/10.1175/1520-0442(2000)013<1517:PEATHD>2.0.CO;2.
Wang, B., J. Yang, and T. Zhou, 2008: Interdecadal changes in the major modes of Asian-Australian monsoon variability: Strengthening relationship with ENSO since the late 1970s. J. Climate, 21, 1771–1789, https://doi.org/10.1175/2007JCLI1981.1.
Wang, B., F. Huang, Z. Wu, J. Yang, X. Fu, and K. Kikuchi, 2009: Multi-scale climate variability of the South China Sea monsoon: A review. Dyn. Atmos. Oceans, 47, 15–37, https://doi.org/10.1016/j.dynatmoce.2008.09.004.
Wang, B., J. Liu, H. J. Kim, P. J. Webster, S.-Y. Yim, and B. Xiang, 2013: Northern Hemisphere summer monsoon intensified by mega-El Niño/Southern Oscillation and Atlantic multidecadal oscillation. Proc. Natl. Acad. Sci. USA, 110, 5347–5352, https://doi.org/10.1073/pnas.1219405110.
Wang, X., C. Wang, W. Zhou, D. Wang, and J. Song, 2011: Teleconnected influence of North Atlantic sea surface temperature on the El Niño onset. Climate Dyn., 37, 663–676, https://doi.org/10.1007/s00382-010-0833-z.
Wang, Y., S. Li, and D. Luo, 2009: Seasonal response of Asian monsoonal climate to the Atlantic multidecadal oscillation. J. Geophys. Res., 114, D02112, https://doi.org/10.1029/2008JD010929.
Wu, B., T. Li, and T. J. Zhou, 2010: Relative contributions of the Indian Ocean and local SST anomalies to the maintenance of the western North Pacific anomalous anticyclone during the El Niño decaying summer. J. Climate, 23, 2974–2986, https://doi.org/10.1175/2010JCLI3300.1.
Wu, R., and B. P. Kirtman, 2003: On the impacts of the Indian summer monsoon on ENSO in a coupled GCM. Quart. J. Roy. Meteor. Soc., 129, 3439–3468, https://doi.org/10.1256/qj.02.214.
Xie, S., K. Hu, J. Hafner, H. Tokinaga, Y. Du, G. Huang, and T. Sampe, 2009: Indian Ocean capacitor effect on Indo-western Pacific climate during the summer following El Niño. J. Climate, 22, 730–747, https://doi.org/10.1175/2008JCLI2544.1.
Zhang, L., K. Fraedrich, X. Zhu, F. Sielmann, and X. Zhi, 2015: Interannual variability of winter precipitation in southeast China. Theor. Appl. Climatol., 119, 229–238, https://doi.org/10.1007/s00704-014-1111-5.
Zhang, R., and T. L. Delworth, 2005: Simulated tropical response to a substantial weakening of the Atlantic thermohaline circulation. J. Climate, 18, 1853–1860, https://doi.org/10.1175/JCLI3460.1.
Zhang, R., and T. L. Delworth, 2006: Impact of Atlantic multidecadal oscillations on India/Sahel rainfall and Atlantic hurricanes. Geophys. Res. Lett., 33, L17712, https://doi.org/10.1029/2006GL026267.
Zhang, R., and T. L. Delworth, 2007: Impact of the Atlantic multidecadal oscillation on North Pacific climate variability. Geophys. Res. Lett., 34, L23708, https://doi.org/10.1029/2007GL031601.
Zhao, P., S. Yang, R. Wu, Z. Wen, J. Chen, and H. Wang, 2012: Asian origin of interannual variations of summer climate over the extratropical North Atlantic Ocean. J. Climate, 25, 6594–6609, https://doi.org/10.1175/JCLI-D-11-00617.1.
Zhou, B., and Z. Wang, 2015: On the significance of the interannual relationship between the Asian‐Pacific Oscillation and the North Atlantic Oscillation. J. Geophys. Res., 120, 6489–6499, https://doi.org/10.1002/2015JD023328..
Zhou, W., and J. C. L. Chan, 2007: ENSO and the South China Sea summer monsoon onset. Int. J. Climatol., 27, 157–167, https://doi.org/10.1002/joc.1380.
Zhou, X., S. Li, F. Luo, Y. Gao, and T. Furevik, 2015: Air–sea coupling enhances the East Asian winter climate response to the Atlantic multidecadal oscillation. Adv. Atmos. Sci., 32, 1647–1659, https://doi.org/10.1007/s00376-015-5030-x.
Zhou, Y., and Z. Wu, 2016: Possible impacts of mega-El Niño/Southern Oscillation and Atlantic multidecadal oscillation on Eurasian heatwave frequency variability. Quart. J. Roy. Meteor. Soc., 142, 1647–1661, https://doi.org/10.1002/qj.2759.
Zhu, X., O. Bothe, and K. Fraedrich, 2011: Summer atmospheric bridging between Europe and East Asia: Influences on drought and wetness on the Tibetan Plateau. Quat. Int., 236, 151–157, https://doi.org/10.1016/j.quaint.2010.06.015.