Impact of Tropical Sea Surface Temperature Independent of the Preceding Winter ENSO on the Interannual Variability of South China Sea Summer Monsoon Intensity

Abstract

This study explores the impact of the tropical sea surface temperature (SST) independent of the preceding winter El Nino-Southern Oscillation (ENSO) events (ENSO-independent SST) on the interannual variability of the South China Sea Summer Monsoon (SCSSM) and the associated mechanisms. During summer, the ENSO-independent SST component dominates across tropical ocean regions. The tropical ENSO-independent SSTs during spring and summer in the Maritime Continent (MC), the equatorial central-eastern Pacific (CEP), and the tropical Atlantic Ocean (TAO) regions play a comparably significant role in the interannual variation of the SCSSM intensity, compared to the tropical SST dependent on the preceding winter ENSO. The ENSO-independent SST anomalies (SSTA) in the TAO during spring and summer exhibit significant persistence. They can influence the SCSSM through westward propagation of teleconnection, as well as through eastward-propagating Kelvin waves. In summer, the SSTA in the MC, CEP, and TAO regions contribute jointly to the variability of the SCSSM. The MC SSTA affects local convection and generates anomalous meridional circulation to impact the SCSSM intensity. The CEP SSTA directly influences the SCSSM via the Matsuno-Gill response mechanism and indirectly affects it via meridional circulation by modulating vertical motions over the MC through zonal circulation. The TAO SSTA impacts the SCSSM through both westward and eastward pathways, as well as by influencing zonal circulation patterns in the tropical and subtropical North Pacific. The results offer valuable insights into the factors influencing the interannual variability of the SCSSM intensity.

Headnote

Abstract: This study explores the impact of the tropical sea surface temperature (SST) independent of the preceding winter El Nino-Southern Oscillation (ENSO) events (ENSO-independent SST) on the interannual variability of the South China Sea Summer Monsoon (SCSSM) and the associated mechanisms. During summer, the ENSO-independent SST component dominates across tropical ocean regions. The tropical ENSO-independent SSTs during spring and summer in the Maritime Continent (MC), the equatorial central-eastern Pacific (CEP), and the tropical Atlantic Ocean (TAO) regions play a comparably significant role in the interannual variation of the SCSSM intensity, compared to the tropical SST dependent on the preceding winter ENSO. The ENSO-independent SST anomalies (SSTA) in the TAO during spring and summer exhibit significant persistence. They can influence the SCSSM through westward propagation of teleconnection, as well as through eastward-propagating Kelvin waves. In summer, the SSTA in the MC, CEP, and TAO regions contribute jointly to the variability of the SCSSM. The MC SSTA affects local convection and generates anomalous meridional circulation to impact the SCSSM intensity. The CEP SSTA directly influences the SCSSM via the Matsuno-Gill response mechanism and indirectly affects it via meridional circulation by modulating vertical motions over the MC through zonal circulation. The TAO SSTA impacts the SCSSM through both westward and eastward pathways, as well as by influencing zonal circulation patterns in the tropical and subtropical North Pacific. The results offer valuable insights into the factors influencing the interannual variability of the SCSSM intensity.

Key words: South China Sea Summer Monsoon; interannual variability; tropical sea surface temperature; ENSO-independent component

1 INTRODUCTION

The South China Sea (SCS) summer monsoon (SCSSM) is an important component of the East Asian summer monsoon system and an important bridge connecting the summer monsoon systems of East Asia and South Asia (Wang 1994; Wang et al. 2009). The SCSSM activity is closely related to the summer rainfall and climate in South China (He and Zhu 2015; Huang et al. 2006; Jiang et al. 2018) and also has a great effect on the weather and climate even on a broader domain in the Northern Hemisphere through teleconnection (Ren and

Huang 2002; Wang et al. 2001; Xu et al. 2019). Therefore, to fully understand the controlling factors of the interannual variation of the SCSSM intensity and the related underlying mechanisms is instructive for the improvement of the predictions of the SCSSM and the climate in South China.

The SCSSM exhibits significant interannual variability (Lau and Yang 1997; Luo et al. 2015; Mao et al. 2011; Wang et al. 2009). Due to its central location in the Asian-Australian monsoon region, the SCSSM is usually regulated by numerous factors (Chen et al. 2022), including tropical sea surface temperature (SST). For example, the El Nino-Southern Oscillation (ENSO) directly affects the South China Sea summer monsoon via the Walker circulation or stimulating the westward-propagating Rossby wave and air-sea interactions (Wang et al. 2000; Wang et al. 2009), and the SST anomalies in tropical Indian Ocean (TIO) force the SCSSM via the mechanism of Kelvin-wave-induced Ekman divergence or the anomalous Walker circulation (Xie et al. 2009; Yang et al. 2007; Yuan et al. 2008).

ENSO is the most prominent air-sea coupled mode in tropical regions (Mestas-Nunez and Enfield 2001) and has received widespread attention for its crucial impact on the SCSSM activity. The ENSO events usually develop to their mature phase in winter, and their evolution in the subsequent decaying phase can directly affect the SCSSM onset and the East Asian climate through the Pacific-East Asia teleconnection (Wang et al. 2000; Zhou and Chan 2007), and can also affect the SCSSM intensity by influencing the SST in the TIO (Xie et al. 2009). The SCSSM usually breaks out later (earlier) during the decaying phase of El Nino (La Nina) events, followed by a weaker (stronger) intensity in summer (Zhou and Chan 2007). Some studies suggested that the relationship between ENSO and the SCSSM is unstable and has tended to weaken in the past twenty years. It is associated with the interdecadal change of ENSO teleconnection and the tropical Pacific cold tongue mode triggered by enhanced warm pool convection due to global warming (Hu et al. 2018; Hu et al. 2020a; Jiang and Zhu 2020, 2021). Therefore, exploring the non-ENSO forcing factors of the SCSSM activities is very important.

Owing to the warm climatology, multiscale convective activities over the Maritime Continent (MC) region, surrounded by the warmest climatological SSTs, have notable impacts on both the local and global climate from intraseasonal to interannual time scales (Chung et al. 2011; Sui et al. 2007; Wu and Zhou 2008; Xie et al. 2022). The MC SST anomalies may affect the SCSSM circulation by modulating the local convection activities and inducing anomalous local meridional circulation (Wu and Zhou 2008; Xie and Wang 2020; Zhang et al. 2021).

In recent years, many studies have also concerned the connection between the SST anomalies (SSTA) in the tropical Atlantic Ocean (TAO) and the circulation anomalies over the tropical Pacific and East Asia (Ham et al. 2013; Hu et al. 2020b; Jin and Huo 2018; Rong et al. 2010). Some studies suggested that the TIO SSTA might affect the summer circulation over the Northwestern Pacific by stimulating the westward-propagating Gill-type Rossby wave teleconnection and the East Asian summer monsoon via eastward-propagating Kelvin-wave response pathway (Choi and Ahn 2019; Ham et al. 2013). Recently, Xu et al (2019) emphasized that the northern TAO SSTA may lead to anomalous circulation in the Northwestern Pacific through the aforementioned westward-propagating pathway, thereby causing variations in the SCSSM intensity.

Although the SST in MC and TAO regions and ENSO are closely related to the SCSSM, they are not independent of each other. On the one hand, the summer MC SSTA may stimulate the local anomalous convection, inducing wind anomalies in the Northwestern Pacific and ultimately triggering ENSO events (Gu et al. 2010). The TAO SSTA can affect subsequent ENSO events by forcing low-level zonal wind anomalies in the equatorial eastern Pacific region (Chikamoto et al. 2020; Ding et al. 2011; Ham et al. 2013; Jiang and Li 2021). On the other hand, ENSO events may also affect the MC SST via the Walker circulation (Gu et al. 2010; Sui et al. 2007; Wu and Zhou 2008; Zhang et al. 2021), and the TAO SST through zonal circulation and the Pacific-North American teleconnection (Alexander and Scott 2002; Enfield and Mayer 1997; Klein et al. 1999; Wu and He 2019). However, some studies showed that the MC and TAO SSTs are not forced fully by preceding ENSO events. After the pre-winter ENSO signals are removed, the MC SST still remains closely related to the East Asian climate (Xu and Guan 2017), and the northern TAO SST also remains significantly correlated with the SCSSM (Xu et al. 2019). Therefore, to explore the climatic effect of the tropical SST, we divide the SST into two components, i.e., ENSO-dependent and ENSO-independent components, which are dependent on and independent of the pre-winter ENSO, respectively.

There are still some unclear issues regarding the relationship between the intensity of the SCSSM and tropical SST. Specifically, what are the variance contributions of tropical SST components independent of preceding winter ENSO (named ENSO-independent SST) to the interannual variability of tropical SST in spring and summer? Where are the key domains in which the tropical ENSO-independent SSTA may exert critical forcing on the SCSSM intensity? And what are the related underlying physical mechanisms? The aim of this study is to address these issues. The remainder of this paper is organized as follows. The datasets and methods used in the study are introduced in Section 2. In Section 3, we explore the leading interannual modes of the SCSSM intensity. Then, the seasonal and spatial variation features of tropical ENSO-independent SSTA related to the SCSSM intensity are examined in Section 4. The impacts of tropical ENSO-independent SSTA in the key domains on the interannual variation of the SCSSM intensity and underlying mechanisms are discussed in Section 5. Finally, the conclusion and discussion are provided in Section 6.

2 DATA AND METHODS

The Fifth-Generation Atmospheric Reanalysis Dataset from the European Centre for Medium-Range Weather Forecasts (Hersbach et al. 2023) is employed in this study, including monthly mean winds, temperature, geopotential height, and other variables for the period 1958-2022 with a horizontal resolution of l°xl°. Monthly mean Hadley Center SST (HadlSST) with a horizontal resolution of l°xl° (Rayner et al. 2006) for the period 1958-2022 is obtained.

Based on previous studies (Chen et al. 2022; Wang et al. 2009), the onset of the SCSSM generally occurs in mid-to-late May, accompanied by the abrupt burst of monsoon rains across a large latitudinal range from 5°N to 22°N with a complete reversal of lower tropospheric zonal wind (from easterly to westerly) between the equator and 18°N in the SCS. In contrast to the abrupt onset, the mean summer monsoon retreat in the SCS is gradual, taking about two months (from mid-September to early November) to withdraw from the northern to the equatorial SCS. The climatological low-level southwesterly winds prevail continuously in the SCS from June to September, and this feature has also been confirmed with the ERA5 data in this study (not shown). Therefore, the SCSSM intensity refers to its summer average from June to September.

In this study, we focus on the interannual variability, and an 8-year high-pass filter has been applied to all variables with the harmonic decomposition procedure. We use the empirical orthogonal function (EOF) analysis to examine the principal modes of the SCSSM interannual variation.

The tropical SST is decomposed into the component dependent on the preceding winter ENSO (named the ENSO-dependent component) and the ENSO-independent component with the following procedure. Firstly, the ENSO-dependent component of SST is estimated with a linear regression equation of SST against the preceding winter SST averaged in the Nino3.4 region (5°S-5°N, 170°-120°W). Then, the residual of the original SST minus the ENSO-dependent component is obtained as the ENSO-independent component, which is unrelated to the preceding winter ENSO.

3 PRINCIPAL MODES OF THE SCSSM INTERANNUAL VARIATION

To clarify the interannual variation features of the

SCSSM intensity, we perform the EOF analysis on the summer 850-hPa wind over the SCSSM region. The first leading mode accounts for 71.3% of the total variance and is well distinguished from other modes according to the criterion of North et al. (1982). The spatial pattern of the first EOF mode (EOF1) (Fig. la) displays westerly anomalies over the central SCS and cyclonic circulation anomalies over the northern SCS to South China, with positive normalized principal component (PCI) in Fig.lb. The negative PCI represents opposite wind anomaly pattern over the SCS. The second mode is featured by the opposite zonal wind anomalies over the southern and northern SCS (Figs, lc and Id) and only accounts for 8.8% of the total variance.

The above EOF results actually indicate that the interannual variation of the SCSSM is predominantly characterized by the intensifying or weakening of the westerly wind component and imply that the first leading mode of the SCSSM well depicts interannual variation of the SCSSM intensity. Therefore, PCI is taken as an index of the interannual variation of the SCSSM intensity in this study, and the positive indices imply intensified SCSSM cases and vice versa. Moreover, PCI indicates a remarkable interdecadal enhancement of the interannual variability of the SCSSM since the late 1980s (Fig. lb).

Furthermore, the variance percentage contribution of the pre-winter Nino3.4 SST to the SCSSM intensity index (SCSSMI), estimated with the linear regression procedure, only accounts for 28%, implying that the effect of ENSO-independent component of SST on the SCSSM deserves further exploration.

4 KEY DOMAINS OF TROPICAL ENSO-IN-DEPENDENT SST RELATED TO THE SCSSM

It is essential to understand the variance percentage contributed by the tropical ENSO-independent component to the raw SST in spring and summer. In spring (Fig. 2a), except in the TIO, the MC, the equatorial central and eastern Pacific (CEP), and a zonal narrow belt in the central-eastern part of the northern tropical Pacific, the variance percentage of the tropical ENSO-independent component is more than 50% in most tropical oceans and especially more than 90% in the western Pacific and the TAO. During summer, the variance percentage of the tropical ENSO-independent SST is more than 50% in all tropical oceans. The above evidence indicates that the ENSO-independent component is predominant in the tropical SST interannual variability in the spring and summer seasons.

To identify the key regions of tropical ENSO-independent SST associated with the interannual variation of the SCSSM intensity, we examine the regression of seasonal mean ENSO-independent SST against the SCSSMI. In the preceding winter, the SCSSM is prominently related to the ENSO-dependent SST signals, featured by negative SSTA in the CEP and TIO and a horse-shoe-shaped positive SSTA in the western Pacific (Fig. 3a2). However, the SCSSM has a very weak connection with the ENSO-independent SST, with no obvious SSTA in the Indo-Pacific Oceans (Fig. 3a3). Eventually, the regressed pattern of raw SST against the SCSSMI (Fig. 3al) is quite similar to that of ENSO-dependent SST (Fig. 3a2).

The regressed patterns in subsequent spring resemble those in the preceding winter (Figs. 3bl-3b3), with weakened regressed SSTA in the CEP. In addition, the ENSO-dependent SSTA is partly offset by the ENSO-independent counterpart in the tropical central Pacific, leading to rather weak raw SSTA compared to the ENSO-dependent component (Figs. 3b 1 and 3b2). Furthermore, compared to winter, the domain with significant tropical SSTA becomes enlarged (Fig. 3b3), and both the ENSO-independent and ENSO-dependent components are equally critical to the SCSSM. The SCSSM-related ENSO-dependent SSTA are mainly concentrated in the northern TAO, while the ENSO-independent counterparts are mainly located in the TAO.

In summer, the SCSSM-related ENSO-dependent SSTA is predominant in the TIO, while both the ENSO-dependent and ENSO-independent components in the MC comparably have a significant correlation with the SCSSMI (Figs. 3c2 and 3c3). The SCSSM-related ENSO-independent SSTA are predominant in the CEP, offset partly by the ENSO-dependent component with opposite anomalies in the off-equatorial regions of tropical CEP. Meanwhile, both the SCSSM-related negative SST anomalies of ENSO-independent and ENSO-independent components persist from spring to summer in the TAO with weakened trend and area shrinkage.

Based on the above discussion, the SCSSM is closely related to the spring and summer ENSO-independent SSTs in the TAO and the summer counterparts in the MC and the CEP. Therefore, we have identified the MC region (10°S-5°N, 90°-115°E), the CEP Nmo3.4 region (5°S-5°N, 170°-120°W), and the TAO region (12°S-18°N, 55°W-10°E) as three key domains of the ENSO-mdependent SST. The SCSSM-related TAO key domain of the ENSO-independent SST identified in this study is quite different from the domain (0°-25°N, 90°W-15°E) suggested by Xu et al. (2019) based on the raw SST. It should be mentioned that although the summer ENSO-independent SST in the CEP is hardly related to the preceding winter ENSO events, it is actually related to the simultaneously developing ENSO events, which might be excited by the forcing of the spring TAO SSTA (Ham et al. 2013).

We further evaluate the variance percentages of the two SST components in the three aforementioned key domains and their correlation coefficients with the SCSSMI. As shown in Table 1, the variance percentage of the summer ENSO-independent SSTA in the MC accounts for 68%, about two times of the ENSO-dependent counterpart, indicating that the former is the predominant component of the summer SSTA in the MC, and has a significantly negative correlation with the SCSSM. In the TAO, the ENSO-independent SSTA is always predominant in both spring and summer, especially up to 96% in the later season. In the CEP, the ENSO-independent SSTA accounts for 98% of the total interannual variance and presents a significant positive correlation with the SCSSM, implying that the summer CEP SST has no significant correlation with the preceding winter one, which is confirmed by a correlation coefficient of-0.11.

It is worth noting that significant correlations are found between the ENSO-independent SSTs in the three key regions. Specifically, the spring ENSO-independent SST in the TAO has a significant correlation with the summer counterparts in the CEP and the TAO; the correlation coefficients are -0.44 and 0.66, respectively. Significant correlations are also found between the summer ENSO-independent SST in the MC and the SSTs in the CEP and TAO, with correlation coefficients of -0.32 and 0.32, respectively. The correlation coefficient between the summer ENSO-independent SSTs in the CEP and TAO is -0.3. All correlation coefficients are significant at the 95% confidence level based on Student's /-test. It indicates that the ENSO-independent SSTs in three key regions may interact with each other through zonal atmospheric circulation.

To evaluate the critical forcing effect of the tropical ENSO-independent SST on the SCSSM, we apply multiple linear regression procedure with five key factors, including the preceding winter Nino3.4 SST, the preceding spring ENSO-independent SST in the TAO and the summer ENSO-independent SSTs in the MC, Nino3.4 region and the TAO, to obtain a multiple linear regression equation for the SCSSMI (i.e., PCI). Based on the equation, the variance contributions of the estimated SCSSMI to the SCSSMI account for 28% and 30% by the ENSO-dependent and ENSO-independent SST components, respectively, and their correlation coefficients with the SCSSMI are 0.52 and 0.55, significant at the 95% confidence level. All this evidence suggests that the ENSO-independent SSTs are also important forcing factors for the interannual variation of the SCSSM.

Furthermore, according to the above-estimated regression coefficients of -0.34, -0.22, 0.19, and -0.11, which correspond to the preceding spring ENSO-independent SST in the TAO and the summer ENSO-mdependent SSTs in the MC, CEP, and TAO regions, respectively, we may also evaluate the individual relative contribution of the ENSO-independent SST in three key regions to the interannual variability of the SCSSMI: the spring ENSO-independent SST in the TAO has greatest contribution to the SCSSMI, followed by the summer counterparts in the MC, CEP and TAO.

5 MECHANISMS FOR IMPACTS OF ENSO-INDEPENDENT SSTS ON SCSSM INTENSITY

According to the significance of correlation coefficients in Table 1, we will discuss the processes and mechanisms of the impact of ENSO-independent SSTs in three key regions of the TAO, CEP, and MC on the SCSSM. 5.1 Spring and summer ENSO-independent SST in TAO

Firstly, based on the normalized SCSSMI, the typical strong and weak SCSSM years are discriminated with the threshold criteria of ±1. There are twelve strong SCSSM years (SESYs): 1967, 1972, 1982, 1985, 1990, 1994, 1997, 2006, 2009, 2012, 2018, 2019; eight weak SCSSM years (WESYs): 1983, 1988, 1995, 1998, 2003, 2010, 2017, 2020. The seasonal evolution of the composite SST and ENSO-independent component in the TAO for the extreme SCSSM years are shown in Fig. 4. In SESYs, the TAO SST presents obvious negative anomalies in spring and summer, especially with maximum amplitude in the former season. The results in WESYs are the opposite of those in SESYs. However, compared to the raw SST, the ENSO-independent component anomalies are weakened to some extent and have comparable amplitudes in the spring and summer seasons. It should be pointed out that the ENSO-independent component anomalies in WESYs are not significant and have smaller amplitudes than those in SESYs, which may imply that the negative anomalies of the ENSO-independent SST in spring and summer in the TAO have greater impacts on the SCSSM than the positive anomalies.

independent SSTA in the TAO. As depicted in Fig. 5al, the warm SSTA in spring in the TAO can trigger an anomalous cyclonic circulation in the lower troposphere to the northwest of the TAO region via the Matsuno-Gill-type Rossby wave atmospheric response (Gill 1980; Matsuno 1966). This cyclonic circulation extends across the tropical Atlantic and eastern Pacific. Additionally, anomalous northerly winds over the Northeastern Pacific strengthen the prevailing local northeasterly trade winds and lead to subsequent cooler SST in the equatorial eastern Pacific through a wind evaporation feedback mechanism. Meanwhile, this cold SSTA pattern also favors the formation of an anomalous anticyclone in the central North Pacific. The easterly wind anomalies to the south of this anticyclone are unfavorable for the onset of the SCSSM.

In summer, the CEP region exhibits significant cold SSTA (Fig. 5a2), indicating the onset of the La Nina event. The anticyclone over the Northwestern Pacific further intensifies and shifts westward, which weakens the SCSSM intensity (Xu et al. 2019). The divergent wind field also presents notable divergence in the lower troposphere over the CEP, leading to convergence in the lower troposphere over the MC region, as well as anomalous ascending motion in the southern SCS and anomalous descending motion in the northern part (Fig. 5b2). Conversely, the upper-level circulation shows the opposite pattern (Fig. 5c2), which is essentially a continuation and development of the springtime anomalous circulation (Figs. 5b 1 and 5c 1). The zonal-vertical circulation averaged over the tropics in spring also indicates significant anomalous upward motion over the TAO region, accompanied by anomalous downward and upward motions over the CEP and MC regions, respectively (Fig. 5dl). The anomalous upward motion over the MC induces anomalous downward motion in the central and northern parts of the SCS via anomalous meridional circulation (Fig. 5el). These anomalous zonal and meridional vertical circulations can persist and intensify into summer (Figs. 5d2 and 5e2).

Some studies suggested that the tropical Atlantic SSTA can cause atmospheric circulation anomalies in the Northwestern Pacific via westward teleconnections or influence East Asian climate anomalies via eastward-propagating Kelvin waves (Ham et al. 2013; Rong et al. 2010). Therefore, we further explore how the spring ENSO-independent SST signal in the TAO affects the SCSSM through an eastward propagation pathway.

The regression of spring tropospheric temperatures against the spring TAO ENSO-independent SST (Fig. 5fl) shows that there is significant warming in the troposphere extending from the TAO eastward to the tropical western Pacific. This suggests that the atmospheric Kelvin waves induced by the warm SST in the TAO region may propagate eastward through the TIO to the MC and the tropical western Pacific. These Kelvin waves can induce an anomalous low-level anticyclone over the Northwestern

Pacific via the mechanism of Kelvin-wave-induced Ekman divergence (Xie et al. 2009; Yang et al. 2007; Zhang et al. 2021). By summer, the warming in the troposphere from the TAO eastward to the tropical western Pacific is weaker compared to spring (Fig. 5f2). The eastward propagation signal of the atmospheric Kelvin waves induced by the TAO warm SST significantly weakens over the TIO. Consequently, the influence of eastward-propagating Kelvin waves on the SCSSM circulation over the

Northwestern Pacific diminishes.

Since the spring ENSO-independent SSTA in the TAO can significantly persist into summer, with a high correlation coefficient of 0.66, the SST and circulation anomaly features associated with the summer ENSO-independent SSTA in the TAO (Figs. 6al-6dl) are very similar to those summer counterparts regressed against the spring ENSO-independent SSTA in the TAO (Figs. 5a2, 5b2, 5c2, and 5f2). Therefore, the results presented in Fig. 5 actually exhibit the continuous and relaying impact process of the spring and summer TAO SST on the western Pacific circulation.

It is important to note that the spring SSTA in the TAO can trigger the occurrence of ENSO events through the westward pathway (Ham et al. 2013), making the summer ENSO signal in the CEP region an important relaying factor in the influence of the spring and summer ENSO-independent SST signals in the TAO on the SCSSM through the westward pathway. Specifically, the spring TAO ENSO-independent SST exhibits a significant negative correlation with summer Nino3.4 SST, with a correlation coefficient of-0.44. If the effect of the summer CEP SST signal is removed, the SST and circulation anomalies in the CEP and MC regions clearly weaken (Figs. 6a2-6d2), especially the anomalous anticyclone near the Philippines and the secondary circulation in the MC and the equatorial Pacific. Interestingly, compared to Fig. 6al, the positive SSTA in the TIO becomes more pronounced when the summer CEP SST signal is removed (Fig. 6a2). A more significant warming is also present in the troposphere from the TAO to the TIO (Figs. 6dl and 6d2) compared to the results with the concurrent CEP SST signal included. In other words, during summer, the process by which the TAO ENSO-independent SSTA induces Kelvin waves to affect the western Pacific circulation through the eastward pathway is partially counteracted by the concurrent CEP SST signal.

Figure 7 exhibits the longitudinal-time cross-sections of SST and circulation anomalies, illustrating how the spring TAO SST signal affects the circulation over the SCS via both westward and eastward pathways. For the westward pathway, the warm SSTA in the TAO during spring can initially trigger local convection indicated by negative vertical ^-velocity anomalies (Fig. 7a). This results in an anomalous cyclonic circulation in the lower troposphere spanning the tropical Atlantic and eastern Pacific via the Matsuno-Gill response, with negative 850-hPa streamfunction anomalies (Gill 1980; Matsuno 1966). Subsequently, the northerly winds on the western side of this anomalous cyclonic circulation enhance the prevailing northeasterly trade winds in the eastern Pacific, leading to an SST cooling. The SST cooling in the eastern Pacific then induces an anomalous anticyclone to its northwest. The anticyclone strengthens gradually via the wind-evaporation feedback mechanism and continues to develop westward. By summer, the anomalous anticyclone develops into a strong anticyclone over the Northwestern Pacific, and weakens the SCSSM intensity (Xu et al. 2019).

For the eastward pathway, the spring warm SSTA in the TAO stimulates eastward-propagating Kelvin waves (Fig. 7b). These waves cause tropospheric warming over the TAO and TIO regions and further propagate eastward to the MC and the SCS through convective coupling process. The tropospheric warming leads to an anomalous low-pressure center at lower levels around the MC, which generates easterly wind anomalies to the east of the MC. The easterly wind anomalies result in lower-level convergence and anomalous ascending motion over the MC, and further enhance the equatorial easterly wind anomalies. By summer, the easterly wind anomalies weaken the SCSSM intensity (Ham et al. 2013). 5.2 Summer ENSO-independent SSTs in MC and CEP

As shown in Fig. 3c3 and Table 1, during summer, the zonal tripole pattern with alternative SSTA in the MC, CEP, and TAO regions is closely associated with the interannual variation of the SCSSM intensity. This indicates a synergistic effect of the SSTA in above three key regions on the variability of the SCSSM intensity. To quantify the effect of the zonal tripole pattern, we define a composite SST index based on the SSTA in three key regions as follows: where the ENSO-independent SSTA indices for three key regions are standardized, the weights are determined with the regression coefficients of the multiple linear regression equation for the summer SCSSMI and the SST in three key regions. The 7SST index defined in Eq. 1 has a correlation coefficient of -0.51 with the SCSSMI, which is more significant than the individual correlations of each key region with the SCSSMI (Table 1).

The SSTA and anomalous atmospheric circulation regressed against the summer 7SST index are shown in Fig. 8. The tropical SSTA is characterized by a quadrupole pattern, with negative SSTA in the TIO and CEP regions and positive SSTA in the MC and TAO regions (Fig. 8a). This pattern is very similar to that shown in Fig. 3c3, except with opposite signs, and the significance of the TIO SSTA is somewhat more pronounced.

Under the influence of these tropical SSTA, anomalous westerlies prevail in the lower troposphere over the TIO and TAO regions, while significant easterly anomalies are observed from the MC to the CEP regions. Moreover, an anomalous anticyclone is present over the northern SCS and the subtropical Northwestern Pacific, indicating a weakening of the SCSSM intensity.

From the perspective of underlying mechanisms, positive SSTA in the MC region, along with negative SSTA in the TIO and CEP regions, lead to anomalous easterlies and westerlies converging in the lower troposphere over the MC region, resulting in intense ascending motion (Fig. 8b) as well as upper-level divergence (Fig. 8c). The anomalous meridional circulation pattern shows that strong upward motion and upper-level divergence over the MC induce anomalous downward motion in the central and northern SCS through meridional circulation (Fig. 8e), which suppresses convection activities in the central and northern SCS and creates an anomalous anticyclone at lower levels, thereby weakening the SCSSM. This indicates that the SSTA in the MC is a significant forcing factor influencing the SCSSM via meridional circulation (Chung et al. 2011; Sui et al. 2007; Wu and Zhou 2008; Zhang et al. 2021).

Additionally, the cold SSTA in the CEP region can induce an anomalous anticyclone to its northwest via the Matsuno-Gill response (Gill 1980; Matsuno 1966), strengthening the anticyclone in the Northwestern Pacific and leading to a weakening of the SCSSM. Meanwhile, the cold SSTA in the CEP region favor the development of local anomalous downward motion and regulate vertical motion over the MC via the anomalous zonal-vertical circulation (Figs. 8b, 8c, and 8d), indirectly affecting the SCSSM through meridional-vertical circulation.

The anomalous ascending motion forced by the warm SSTA in the TAO can, on one hand, be coupled with the anomalous descending motion over the CEP region via the tropical zonal-vertical circulation (Fig. 8d). On the other hand, it can synergistically influence the SCSSSM together with the SSTA in other key regions through the westward and eastward pathways discussed in Section 5.1, which will not be repeated here.

6 SUMMARY AND DISCUSSION

This study has identified the leading mode of the interannual variability of the SCSSM intensity during the summer season from June to September, and examined the seasonal and spatial characteristics of tropical SST components dependent on and independent of preceding winter ENSO, in terms of correlation with the interannual variability in the SCSSM intensity. Moreover, the study has also explored the physical processes and mechanisms by which the ENSO-independent SSTA in key tropical regions affect the SCSSM.

The leading pattern for the strong SCSSM cases on interannual time scale is characterized by an anomalous cyclonic circulation in the northern SCS, and an anomalous anticyclonic circulation in the southern SCS, accompanied by strong anomalous westerlies in the central SCS. The weak SCSSM cases are opposite to the strong cases.

In spring, the interannual variability of the ENSO-dependent SST component predominates in the tropical eastern Indian Ocean, the western part of MC, and the central Pacific, compared to the SST component independent of the preceding winter ENSO. In contrast, the interannual variability of summer tropical SST is predominantly driven by the ENSO-independent SST component.

The interannual variability of the SCSSM intensity is closely associated with the ENSO-dependent SST during the preceding winter and spring. The tropical ENSO-independent SST during spring and summer in the MC, CEP, and TAO regions also play a comparably significant role in the interannual variation of the SCSSM intensity.

The ENSO-independent SSTA in the TAO during spring and summer exhibit significant persistence. They can influence the SCSSM through westward propagation of teleconnection, as well as through eastward-propagating Kelvin waves.

During summer, the ENSO-independent SSTA in the MC, CEP, and TAO regions contribute jointly to the interannual variability of the SCSSM. The MC SSTA affects local convection and generates anomalous meridional-vertical circulation to impact the SCSSM intensity. The CEP SSTA directly influences the SCSSM via the Matsuno-Gill response mechanism and indirectly affects it via meridional circulation by modulating vertical motions over the MC through zonal-vertical circulation. The TAO SSTA impacts the SCSSM through both westward and eastward pathways, as well as by influencing zonal circulation patterns in the tropical and subtropical North Pacific.

Although the second mode of the SCSSM only accounts for 8.8% of the total variance (Figs, lc and Id) and has not been discussed in detail in previous sections, a brief discussion on the impact of SST on it is helpful for comprehensively understanding the influence of SST on the diverse variation of the SCSSM. The regression patterns of SST against the second principal component (PC2) of the EOF analysis on the SCSSM (not shown) indicate that PC2 is not related to the preceding winter SST in the CEP, while it has significant positive correlation with the summer ENSO-independent SST in the CEP and TIO, and negative correlation with the counterpart in the MC. These results suggest that the preceding ENSO events have no obvious effect on the subsequent second mode of the SCSSM, while the summer positive (negative) ENSO-independent SSTA in the CEP and TIO, jointed with the negative (positive) MC SSTA, can lead to an anomalous anticyclone (cyclone) over the SCS. This phenomenon may be related to developing ENSO events. The related physical processes and mechanisms remain to be studied in the future.

In this study, to acquire the ENSO-independent SST component, the ENSO-dependent component of SST is removed using a linear regression. While this is a common approach in many previous studies, the linear regression cannot fully separate the ENSO-independent component of SST from the ENSO-dependent SST component. Thus, the results of this study should be applied with the limitations of linear regression in mind.

The results obtained in our study offer valuable insights into the factors influencing the interannual variability of the SCSSM intensity. Although this study has revealed the importance of the tropical ENSO-independent SST during spring and summer in affecting the SCSSM variability and explored the associated mechanisms, there are still some issues remained to be explored in the future. For instance, which of the two pathways through which the TAO SST affects the SCSSM is more significant? What is the relative importance of the SSTA in the MC, CEP, and TAO regions for the SCSSM interannual variability? Furthermore, is there an interaction between the ENSO-dependent and ENSO-independent SST components, and how do they influence each other and the SCSSM?

References

REFERENCES

Alexander, M., and J. Scott, 2002: The influence of ENSO on air - sea interaction in the Atlantic. Geophys. Res. Lett., 29, 46-41^16-44, https://doi.org/10.1029/2001GL014347.

Chen, W., P. Hu, and J. Huangfu, 2022: Multi-scale climate variations and mechanisms of the onset and withdrawal of the South China Sea summer monsoon. Sci. China Earth Sci., 65, 1030-1046, https://doi.org/10.1007/sll430-021-9902-5.

Chikamoto, Y, Z. F. Johnson, S. Y S. Wang, M. J. McPhaden, and T. Mochizuki, 2020: El Nino-Southern Oscillation Evolution Modulated by Atlantic Forcing. J. Geophys. Res., 125, https://doi.org/10.1029/2020jc016318.

Choi, Y.-W., and J.-B. Ahn, 2019: Possible mechanisms for the coupling between late spring sea surface temperature anomalies over tropical Atlantic and East Asian summer monsoon. Climate Dyn., 53, 6995-7009, https://doi.org/ 10.1007/s00382-019-04970-3.

Chung, P.-H., C.-H. Sui, and T. Li, 2011: Interannual relationships between the tropical sea surface temperature and summertime subtropical anticyclone over the western North Pacific. J. Climate, 16, https://doi.org/10.1029/2010JD015554.

Ding, H., N. S. Keenlyside, and M. Latif, 2011: Impact of the Equatorial Atlantic on the El Nino Southern Oscillation. Climate Dyn., 38, 1965-1972, https://doi.org/10.1007/ s00382-011-1097-y.

Enfield, D. B., and D. A. Mayer, 1997: Tropical Atlantic sea surface temperature variability and its relation to El Nino -Southern Oscillation. J. Geophys. Res, 102, 929-945, https:// doi.org/10.1029/96jc03296.

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.

Gu, D. J., T. Li, and Z. P. Ji, 2010: Connection of the South China sea summer monsoon to maritime continent convection and ENSO. J. Trop. Meteor, 16, 1-9, https://doi.Org/10.3969/j. issn.1006-8775.2010.01.001.

Ham, Y.-G., J.-S. Kug, J.-Y Park, and F.-F. Jin, 2013: Sea surface temperature in the north tropical Atlantic as a trigger for El Nino/Southern Oscillation events. Nat. Geosci., 6, 112-116, https://doi.org/10.1038/ngeol686.

He, J., and Z. Zhu, 2015: The relation of South China Sea monsoon onset with the subsequent rainfall over the subtropical East Asia. Int. J. Climatol, 35, 4547^1556, https://doi.org/10.1002/joc.4305.

Hersbach, H., and Coauthors, 2023: ERA5 monthly averaged data on pressure levels from 1940 to present. Copernicus Climate Change Service Climate Data Store, accessed July 4, 2023, https://doi.org/10.24381/cds.6860a573.

Hu, P., W. Chen, R. Huang, and D. Nath, 2018: On the weakening relationship between the South China Sea summer monsoon onset and cross - equatorial flow after the late 1990s. Int. J. Climatol, 38, 3202-3208, https://doi. org/10.1002/joc.5472.

Hu, P., W. Chen, S. Chen, Y Liu, and R. Huang, 2020a: Extremely early summer monsoon onset in the South China Sea in 2019 following an El Nino event. Mon. Wea. Rev., 148, 1877-1890, https://doi.Org/10.1175/mwr-d-19-0317.l.

Hu, P., W. Chen, S. Chen, Y Liu, L. Wang, and R. Huang, 2020b: Impact of the March Arctic Oscillation on the South China Sea summer monsoon onset. Int. J. Climatol, 41, E3239-

E3248, https://doi.org/10.1002/joc.6920.

Huang, R, L. Gu, L. Zhou, and S. Wu, 2006: Impact of the thermal state of the tropical western Pacific on onset date and process of the South China Sea summer monsoon. Adv. Atmos. Sci., 23, 909-924, https://doi.org/10.1007/s00376-006-0909-1.

Jiang, L., and T. Li, 2021: Impacts of tropical North Atlantic and equatorial Atlantic SST anomalies on ENSO. J. Climate, 34, 5635-5655, https://doi.Org/10.1175/jcli-d-20-0835.l.

Jiang, N., and C. Zhu, 2020: Tropical Pacific cold tongue mode triggered by enhanced warm pool convection due to global warming. Environ. Res. Lett., 15, https://doi.org/10.1088/ 1748-9326/ab7d5e.

Jiang, N., 2021: Seasonal Forecast of South China Sea summer monsoon onset disturbed by cold tongue La Nina in the past decade. Adv. Atmos. Sci., 38, 147-155, https://doi.org/ 10.1007/s00376-020-0090-y.

Jiang, X., Z. Wang, and Z. Li, 2018: Signature of the South China Sea summer monsoon onset on spring-to-summer transition of rainfall in the middle and lower reaches of the Yangtze River basin. Climate. Dyn., 51, 3785-3796, https://doi.org/ 10.1007/s00382-018-4110-x.

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.

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<3C0917: RSSTVD>3E2.0.CO;2.

Lau, K.-M., and S. Yang, 1997: Climatology and interannual variability of the southeast asian summer monsoon. Adv. Atmos. Sci., 14, 550-562, https://doi.org/10.1007/s00376-997-0016-y.

Luo, M., Y Leung, H. F. Graf, M. Herzog, and W. Zhang, 2015: Interannual variability of the onset of the South China Sea summer monsoon. Int. J. Climatol, 36, 550-562, https://doi. org/10.1002/joc.4364.

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.

Matsuno, T., 1966: Quasi-Geostrophic Motions in the Equatorial Area. J. Meteorol. Soc. Japan, 44, 25^13, https://doi.org/ 10.2151/jmsjl965.44.1_25.

Mestas-Nunez, A. M., and D. B. Enfield, 2001: Eastern Equatorial Pacific SST Variability: ENSO and Non-ENSO components and their climatic associations. J. Climate, 14, 391-402, https://doi.Org/l0.1175/1520-0442(2001) 014<3C0391:EEPSVE>3E2.0.CO;2.

North, G. R, L. B. Thomas, F. C. RObert, and J. M. Fanthune, 1982: Sampling errors in the estimation of empirical orthogonal functions. Mon. Wea. Rev, 110, 699-706, https://doi.org/10.1175/1520-0493(1982)110<3C0699:SEI-TEO>3E2.0.CO;2.

Rayner, N.A., P. Brohan, D. E. Parker, C. K. Folland, J. J. Kennedy, M. Vanicek, T. J. Ansell, and S. F. B. Tett, 2006: Improved analyses of changes and uncertainties in sea surface temperature measured in situ since the mid-nineteenth century: The HadSST2 dataset. J. Climate, 19, 446-469, https://doi.Org/10.1175/JCLI3637.l.

Ren, B., and R. Huang, 2002: 10-25-day intraseasonal variations of convection and circulation associated with thermal state of the western Pacific warm pool during boreal summer. Adv. Atmos. Sci., 19, 321-336, https://doi.org/10.1007/s00376-002-0025-9.

Rong, X., R. Zhang, and T. Li, 2010: Impacts of Atlantic sea surface temperature anomalies on Indo-East Asian summer monsoon-ENSO relationship. Chin. Sci. Bull., 55, 2458-2468, https://doi.org/10.1007/sll434-010-3098-3.

Sui, C.-H, P.-H. Chung, and T. Li, 2007: Interannual and interdecadal variability of the summertime western North Pacific subtropical high. Geophys. Res. Lett., 34, https://doi. org/10.1029/2006GL029204. Wang, B., 1994: Climatic regimes of tropical convection and rainfall. J. Climate, 7, 1109-1118, https://doi.org/10.1175/1520-442(1994)007<3C1109:CROTCA>3E2.0.CO;2.

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<3C1517:PEATHD>3E2.0.CO;2. Wang, B., R. Wu, and K.-M. Lau, 2001: Interannual variability of the Asian summer monsoon: contrasts between the Indian and the Western North Pacific-East Asian monsoons. J. Climate, 14, 4073^1090, https://doi.org/10.1175/1520-0442 (2001)014<3C4073:IVOTAS>3E2.0.CO;2.

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. Ocean., 47, 15-37, https://doi.org/10.1016/j.dynatmoce.2008.09.004. Wu, B., and T. Zhou, 2008: Oceanic origin of the interannual and interdecadal variability of the summertime western Pacific subtropical high. Geophys. Res. Lett., 35, https://doi.org/10.1029/2008GL034584. Wu, R., and Z. He, 2019: Northern tropical Atlantic warming in El Nino decaying spring: impacts of El Nino amplitude. Geophys. Res. Lett, 46, 1407-14081, https://doi.org/10.1029/2019gl085840.

Xie, M., and C. Wang, 2020: Decadal variability of the anticyclone in the Western North Pacific. J. Climate, 33, 9031-9043, https://doi.org/10.1175/JCLI-D-20-0008.1.

Xie, M., C. Wang, and S. Chen, 2022: The role of the maritime continent SST anomalies in maintaining the Pacific-Japan pattern on decadal time scales. J. Climate, 35, 1079-1095, https://doi.Org/10.1175/JCLI-D-21-0555.l.

Xie, S.-P, 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 Nino. J. Climate, 22, 730-747, https://doi.org/10.1175/ 2008jcli2544.1.

Xu, P., L. Wang, W. Chen, J. Feng, and Y. Liu, 2019: Structural changes in the Pacific-Japan pattern in the late 1990s. J. Climate, 32, 607-621, https://doi.org/10.1175/jcli-d-18-0123.1.

Xu, Q., and Z. Guan, 2017: Interannual variability of summertime outgoing longwave radiation over the Maritime Continent in relation to East Asian summer monsoon anomalies. J. Meteor. Res., 31, 665-677, https://doi.org/ 10.1007/sl3351-017-6178-3.

Xu, J., L. Huo, C. Song, Y Liu, J. Chen, and Q. Tao, 2019: Influence of northern Atlanic SST anomaly on South China Sea summer monsoon and its mechanism. Trans. Atmos. Sci., 42, 293-302, https://doi.Org/10.13878/j.cnki. dqkxxb.20180831001.(in Chinese)

Yang, J., Q. Liu, S. P. Xie, Z. Liu, and L. Wu, 2007: Impact of the Indian Ocean SST basin mode on the Asian summer monsoon. Geophys. Res. Lett., 34, https://doi.org/10.1029/ 2006gl028571.

Yuan, Y, W. Zhou, J. C. L. Chan, and C. Li, 2008: Impacts of the basin - wide Indian Ocean SSTA on the South China Sea summer monsoon onset. Int. J. Climatol, 28, 1579-1587, https://doi.org/10.1002/joc. 1671.

Zhang, Z., M. Jian, S. Gao, and Y Qiao, 2021: Interdecadal change in the relationship between interannual variation of the South China Sea monsoon trough and tropical Indo-Pacific sea surface temperature. Int. J. Climatol, 41, E2379-E2394, https://doi.org/10.1002/joc.6853.

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.

Citation: Li, H. C, M. Q. Jian, and S. Gao, 2025: Impact of Tropical Sea Surface Temperature Independent of the Preceding Winter ENSO on the Interannual Variability of South China Sea Summer Monsoon Intensity. J. Trop. Meteor, 31, 165-177, https://doi.org/ 10.3724/j.1006-8775.2025.010.