Abstract

The Silk Road pattern (SRP) is a well-known teleconnection pattern along the upper-level westerly jet over the Eurasian continent during boreal summer. The SRP has experienced an interdecadal change around the late 1970s. The present study identified a new change of the SRP around the late 1990s, which is characterized by significant weakening and zonal phase shift of the major centers of the SRP during the recent decades. The recent reshaping of the SRP is attributed to an enhanced impact of precipitation anomalies over the northeastern Indian summer monsoon (ISM), which is associated with the leading mode change of the ISM precipitation anomalies around the late 1990s. The interdecadal weakening of the upper-level westerly jet over central and East Asia also favors the southward movement of the SRP during recent periods. The differences of the features, climate impact, and causes related to the recent SRP change from those related to the SRP change around the late 1970s are also contrasted in this study.

1. Introduction

The Silk Road pattern (SRP) is an atmospheric teleconnection pattern along the midlatitude westerly jet and is the leading mode of the upper-tropospheric meridional wind anomalies over Eurasia during boreal summer (Lu et al. 2002; Wu 2002; Kosaka et al. 2009; Enomoto et al. 2003). It has an equivalent-barotropic structure with significant circulation anomalies reaching the lower troposphere to exert a large influence on the summer climate over Eurasia (Lu et al. 2002; Ding and Wang 2005; Huang et al. 2011; Kosaka et al. 2011; Huang et al. 2013; Wang et al. 2013; Hong et al. 2017, 2018; Leung and Zhou 2018). Therefore, much work has been devoted to understanding the dynamics and mechanisms of the variability of the SRP.

The SRP illustrates largely the interannual teleconnection (hereafter, the term SRP refers to the interannual SRP unless otherwise specified) between the meridional wind anomalies over Eurasia (Fig. 1; Lu et al. 2002; Chen and Huang 2012; Sato and Takahashi 2006; Yasui and Watanabe 2010; Wang et al. 2017; Hong et al. 2018; Liu et al. 2020). It has two major modes (called SRP1 and SRP2 for brevity) that correspond to the first two leading modes, respectively, of the summer 200-hPa meridional wind (V200) anomalies over the domain 20°–60°N, 0°–150°E (Kosaka et al. 2012; Song et al. 2013). The SRP1 and SRP2 feature a common zonal wavenumber-5–6 structure with their zonal phases displacing by nearly a quarter wavelength (Kosaka et al. 2012). In many previous studies, SRP refers to the SRP1 mode that contributes largely to the total variances of the summer V200 anomalies over the above domain during the past 60 years, which is about twice the contribution of the SRP2 mode (Figs. 1 and 2). In the present study, SRP also refers to SRP1, which is defined as the first leading mode of the summer V200 anomalies over above domain.

Fig. 1.

The V200 anomalies (contours; units: m s−1) associated with (a),(b) EOF1 and (d),(e) EOF2 of the V200 anomalies over the domain 20°–60°N, 0°–150°E during 1965–2014 based on the (a),(d) NCEP1 and and (b),(e) JRA-55 datasets. Also shown are (c),(f) the principal components corresponding to EOF1 and EOF2, respectively. Shading denotes the V200 anomalies significant beyond the 95% confidence level. The green thick contours in (a), (b), (d), and (e) delineate the climatological U200 with values of 20 and 25 m s−1 during 1965–2014. The percentage in parentheses in the title of (a), (b), (d), and (e) indicates the variance explained by EOF1 to the total variance of the V200 over the domain 20°–60°N, 0°–150°E.

Fig. 1.

The V200 anomalies (contours; units: m s−1) associated with (a),(b) EOF1 and (d),(e) EOF2 of the V200 anomalies over the domain 20°–60°N, 0°–150°E during 1965–2014 based on the (a),(d) NCEP1 and and (b),(e) JRA-55 datasets. Also shown are (c),(f) the principal components corresponding to EOF1 and EOF2, respectively. Shading denotes the V200 anomalies significant beyond the 95% confidence level. The green thick contours in (a), (b), (d), and (e) delineate the climatological U200 with values of 20 and 25 m s−1 during 1965–2014. The percentage in parentheses in the title of (a), (b), (d), and (e) indicates the variance explained by EOF1 to the total variance of the V200 over the domain 20°–60°N, 0°–150°E.

Fig. 2.

(a) The evolution of the explanation percentages of the 15-yr sliding EOF1 (solid) and EOF2 (dashed) of the V200 anomalies over the domain 20°–60°N, 0°–150°E based on the NCEP1 and JRA-55 datasets. Also shown are spatial pattern correlations of the 15-yr sliding EOF1 with SRP1 (blue) and the s15-yr sliding EOF2 with SRP2 (red) over the domain 20°–60°N, 0°–150°E based on the (b) NCEP1 and (c) JRA-55 datasets.

Fig. 2.

(a) The evolution of the explanation percentages of the 15-yr sliding EOF1 (solid) and EOF2 (dashed) of the V200 anomalies over the domain 20°–60°N, 0°–150°E based on the NCEP1 and JRA-55 datasets. Also shown are spatial pattern correlations of the 15-yr sliding EOF1 with SRP1 (blue) and the s15-yr sliding EOF2 with SRP2 (red) over the domain 20°–60°N, 0°–150°E based on the (b) NCEP1 and (c) JRA-55 datasets.

Physically, the SRP is a quasi-stationary wave train along the summer Eurasian westerly jet. On the one hand, it is often regarded as an internal atmospheric mode that can be self-maintained by extracting kinetic and available potential energy from the basic flow (Enomoto et al. 2003; Sato and Takahashi 2006; Kosaka et al. 2009; Chen et al. 2013), since the jet can serve as an efficient Rossby waveguide (Hoskins and Ambrizzi 1993). On the other hand, it also can be modulated/triggered by the external forcings, such as the tropical Indian summer monsoon (ISM) heating (Lu et al. 2002; Wu 2002; Wu et al. 2003; Ding and Wang 2005), midlatitude heating (Sato and Takahashi 2006; Lin et al. 2017a,b; Liu et al. 2020), El Niño–Southern Oscillation (ENSO; Ding and Wang 2005; Chen and Huang 2012; Wang et al. 2012), and summer North Atlantic Oscillation (NAO; Liu et al. 2020). Accordingly, the SRP is considered as a crucial medium linking the ISM/North Alantic–European and the northern Asian summer climate variations (Lee et al. 2011; Ding et al. 2011; Kosaka et al. 2011; Huang et al. 2017; Liu and Huang 2019).

The interannual variations of the SRP are unstable and have experienced a notable interdecadal change around the late 1970s, characterized by intensified intensity and prolonged periodicity of the SRP after the late 1970s (Wu and Wang 2002; Wang et al. 2012; Liu et al. 2020, unpublished manuscript). This interdecadal change in the SRP is attributed to a weakened impact of the ISM precipitation due to the change of the ENSO properties (Wu and Wang 2002; Wang et al. 2012; Wu 2017) and an intensified impact of the summer NAO after the late 1970s, in particular on the 4–8-yr time scale (Liu et al. 2020), which is also associated with the abrupt climate regime shift in the late 1970s that involves changes in the global sea surface temperature (SST), ENSO properties, and atmospheric teleconnection patterns (Nitta and Yamada 1989; Trenberth 1990; Wang et al. 2012).

In recent years, another global-scale abrupt climate shift around the late 1990s has also been well recognized, which is accompanied by remarkable interdecadal changes in the global SST (Peterson and Schwing 2003; Deser et al. 2004; Sutton and Dong 2012; Dai 2013) and atmospheric teleconnection patterns as well (Zhu et al. 2011; Huang et al. 2013; Zhu et al. 2015; Si and Ding 2016; Hong et al. 2017; Wang et al. 2017). A natural question arises whether the SRP has been changed in association with the climate shift around the late 1990s. In the current study, we indicate a significant shift in the spatial structure of the SRP (the first leading mode of the summer V200 anomalies over the above domain) around the late 1990s, which is accompanied by significant weakening and southward/eastward movement of the SRP centers during the recent epoch after the late 1990s. Therefore, the main goal of the present study is to investigate the characteristics and possible causes for the SRP change around the late 1990s, as well as their differences from those related to the SRP change around the late 1970s.

The structure of the remainder is as follows. Section 2 describes the details of the data and method used in this study. Section 3 illustrates the changes in the SRP, associated circulation, and climate anomalies around the late 1990s. Section 4 proposes the possible causes of the SRP change; in particular, the roles of the summer precipitation anomalies over ISM and the basic-state change over Eurasia are highlighted in this study. Finally, section 5 provides the main conclusions.

2. Data and method

The datasets used in this study include 1) two monthly atmospheric reanalysis datasets from the National Centers for Environmental Prediction and National Center for Atmospheric Research (NCEP–NCAR), with a 2.5° × 2.5° horizontal resolution (Kalnay et al. 1996) and the Japanese 55-year Reanalysis (JRA-55) dataset, with a 1.25° × 1.25° resolution (Kobayashi et al. 2015); 2) the monthly precipitation from NOAA’s precipitation reconstruction (PREC) dataset (Chen et al. 2002), and the monthly precipitation and surface air temperature data from the Climatic Research Unit (CRU) high-resolution gridded datasets, version 4.03 (Harris et al. 2014); and 3) the monthly global SST from the Hadley Centre (Rayner 2003). The time span is 1961–2018 for all the datasets. This study focuses on the interannual variations. The linear trend and decadal variations of the summer (June–August) mean variables with periods longer than 9 years are removed using the Lanczos high-pass filter (Duchon 1979). Therefore, the variables during 1965–2014 are employed in this study.

Following Yasui and Watanabe (2010) and Kosaka et al. (2012), the empirical orthogonal function (EOF) analysis is used to obtain the first two leading modes (EOF1 and EOF2) of the summer V200 anomalies over the domain 20°–60°N, 0°–150°E. Typically, the EOF1 and EOF2 modes during 1965–2014 are regarded as SRP1 and SRP2, respectively. To investigate the interdecadal changes of the SRP, we first perform a 15-yr sliding EOF analysis on the summer V200 anomalies over the above domain, and detect the changes of the explanation percentages (i.e., explained variance) of the EOF1 and the spatial pattern correlations of the EOF1 with the SRP1 and SRP2 modes, which are defined as the timings of the SRP shifts. In this study, two timings are observed at 1979 and 1998. Therefore, the whole time period is divided into three subperiods: 1965–79 (P1), 1980–98 (P2), and 1999–2014 (P3). The respective SRP structures during three subperiods are adopted to quantify the interdecadal changes of the SRP. The significance of the distinction between the two EOFs is assessed by North’s method (North et al. 1982). The correlation and regression analyses are also used, and the two-tailed Student’s t test is used to estimate the significance. The effective degree of freedom is evaluated based on Bretherton et al. (1999). The significance is also assessed using the false discovery rate analysis method (Wilks 2016), which reveals a slight difference (figures not shown) from that based on the two-tailed Student’s t test, but does not change the results in the present study.

In addition, the wave-activity flux of Takaya and Nakamura (2001) is used to indicate the dynamic characteristic of the stationary Rossby wave trains related to the SRP. The formation of the wave activity flux is written as

 
W=12U¯[u¯(ψx2ψψxx)+υ¯(ψxψyψψxy)u¯(ψxψyψψxy)+υ¯(ψy2ψψyy)].

Here, U¯ is the magnitude of the horizontal vector wind (u, υ) and ψ is the streamfunction. Variables with overbar represent the climatological mean, and variables with a subscript and a prime denote their partial derivatives and anomalies, respectively.

3. Interdecadal change of the SRP around the late 1990s

a. Interdecadal change of the SRP around the late 1990s

To begin our analysis, the typical features of the two SRP modes during 1965–2014 are first revisited and employed as comparison, which are shown in Fig. 1. The spatial–temporal features of the two SRP modes exhibit a high similarity among the NCEP1 and JRA-55 datasets. It is observed that the two SRP modes reveal largely the interannual teleconnection relationship between the V200 anomalies over the Eurasian continent that have a clear wavelike structure and are characterized by alternate southerly and northerly anomalies along the Eurasian westerly jet (Fig. 1). The two SRP modes account for about 34.5% (35.5%) and 15.5% (15.4%) of the total variance of the V200 over Eurasia for the NCEP1 (JRA-55) dataset, respectively, which are well separated based on the method of North et al. (1982). Specifically, the SRP1 has five well-defined centers over northern Europe, eastern Europe, central Asia, North China, and Japan, with the downstream four centers within the westerly jet along about 42°N (Figs. 1a,b). The SRP2 also has five well-defined centers that are situated over central Europe, West Asia, central Asia, Northwest China, and East China (Figs. 1d,e). In contrast to the SRP1, the SRP2 centers are weaker and have narrower zonal phases, and are located much more eastward and southward. Meanwhile, the center over Japan disappears in the SRP2. It is consistent with Kosaka et al. (2012) that the zonal phases of the SRP1 and SRP2 displace about a quarter wavelength.

As mentioned in the section 2, to determine the timing of the interdecadal changes in the SRP (i.e., the SRP1 change), we perform a 15-yr sliding EOF analysis on the V200 anomalies over the above domain, and then investigate the change timings of the explanation percentages of the first two sliding EOFs and the spatial pattern correlations of the sliding EOF1 with the SRP1 and SPR2 modes. Figure 2a displays the evolutions of the explanation percentages of the sliding EOF1 and EOF2 modes in the NCEP1 and JRA-55 datasets. It consistently shows that prior to the late 1970s, the EOF1 mode explains more than 50% of the total variance of the V200 anomalies over the above domain, which decreases sharply around the late 1970s and has a value of about 36% during the period from the late 1970s to the late 1990s. The explanation percentage of the EOF1 mode also experiences a sharp decrease around the late 1990s with a value of 30%–35% during the recent period. As for the EOF2 mode, it accounts for below 20% of the total variance of the V200 anomalies over the above domain during almost the whole period, whose explanation percentage evolutions show a nearly reverse feature to that of the EOF1 mode and have two evident shifts around the late 1970s and late 1990s as well. These results indicate two interdecadal changes in the EOF1 mode (SRP) around the late 1970s and late 1990s, respectively.

Furthermore, the variations of the spatial pattern correlations of the sliding EOF1 mode with the SRP1 and SRP2 modes are explored in Figs. 2b and 2c. Both results of the NCEP1 and JRA-55 datasets demonstrate that the EOF1 mode shows a high pattern correlation with the SRP1 mode prior to the late 1990s, which has a value of beyond 0.85 despite a sharp decreasing around the late 1970s. This indicates the spatial structure of the EOF1 mode is more stable before the late 1990s, although the SRP has experienced an interdecadal change around the late 1970s but with slight change in the spatial structure (Wu 2002; Lin et al. 2017a; Liu et al. 2020; also refer to the following analysis). However, after the late 1990s, the pattern correlation of the sliding EOF1 mode with the SRP1 mode decreases sharply to about 0.75. Meanwhile, its pattern correlation with the SRP2 mode becomes much stable and has a value of 0.55 after the late 1990s. These results further indicate that in addition to the interdecadal change around the late 1970s, the SRP experienced a new interdecadal change around the late 1990s, which is characterized by significant SRP reshaping during recent period.

To clearly illustrate the SRP changes around the late 1990s, we divide the time periods into three subperiods: 1965–79 (P1), 1980–98 (P2), and 1999–2014 (P3), and contrast the SRP structures that are demonstrated as the EOF1 mode of the V200 anomalies over the domain 20°–60°N, 0°–150°E during the three subperiods. Also, the standardized principal component (PC) corresponding to the EOF1 mode during each subperiod is defined as the SRP index (SRPI) during this period (shown in Figs. 3g,h), and the regressed V200 anomalies against the SRPI during each period are considered as the SRP structure during this period.

Fig. 3.

As in Figs. 1a and 1b, but for the EOF1 during (a),(b) 1965–79, (c),(d) 1980–98, and (e),(f) 1999–2014. Also shown are the corresponding principal component during 1965–79 (solid black), 1980–98 (dashed blue), and 1999–2014 (dashed red) based on the (g) NCEP1 and (h) JRA-55 datasets. The green thick contours in (a)–(f) delineate the climatological U200 with values of 20 and 25 m s−1 during the corresponding period.

Fig. 3.

As in Figs. 1a and 1b, but for the EOF1 during (a),(b) 1965–79, (c),(d) 1980–98, and (e),(f) 1999–2014. Also shown are the corresponding principal component during 1965–79 (solid black), 1980–98 (dashed blue), and 1999–2014 (dashed red) based on the (g) NCEP1 and (h) JRA-55 datasets. The green thick contours in (a)–(f) delineate the climatological U200 with values of 20 and 25 m s−1 during the corresponding period.

Figure 3 displays the spatio-temporal structures of the SRP during the three subperiods. The EOF1 modes during the three subperiods account for about 50.8% (51.5%), 35.2% (35.4%), and 33.1% (34.3%) of the total variance of the V200 anomalies over the above domain based on the NCEP1 (JRA-55) dataset, respectively, which are well separated from the EOF2 mode during the corresponding period in both the NCEP1 and JRA-55 datasets based on North’s method. The spatial structures of the SRP during both P1 and P2 feature similar wavelike patterns with five well-defined centers over the midlatitude Eurasian continent, which are characterized by alternate southerly and northerly anomalies centered along about 42°N (Figs. 3a–d). The SRP during P1 and P2 resembles well the typical SRP1 structure during 1965–2014 (Figs. 1a,c), corresponding to the high pattern correlations between the sliding EOF1 mode with SRP1 prior to the late 1990s (Figs. 2b,c). In contrast to the SRP during P1, the centers of the SRP during P2 show little alternation but weaken clearly with the centers moving slightly eastward (Figs. 3c,d and 4a,b). In addition, the meridional scale of the SRP during P2 is narrower than that during P1 (Figs. 3c,d). These changes of the SRP around the late 1970s are in good agreement with the previous studies (Wu 2002; Wang et al. 2012; Liu et al. 2020).

Fig. 4.

The location of the maximum/minimum centers of the (a),(b) V200 in Figs. 3a and,3b and (c),(d) Z200 anomalies in Figs. 5c and,5d during 1965–79 (black), 1980–98 (blue), and 1999–2014 (red) based on the (a),(c) NCEP1 and (b),(d) JRA-55 datasets.

Fig. 4.

The location of the maximum/minimum centers of the (a),(b) V200 in Figs. 3a and,3b and (c),(d) Z200 anomalies in Figs. 5c and,5d during 1965–79 (black), 1980–98 (blue), and 1999–2014 (red) based on the (a),(c) NCEP1 and (b),(d) JRA-55 datasets.

During the recent epoch after the late 1990s, the SRP exhibits a different wavelike pattern from the SRP during P1 and P2 (Figs. 3e,f). It has six active centers along 36°N with southerly and northerly anomalies over central Europe, West Asia, central Asia, Northwest China, East China, and Japan, respectively. In contrast to the SRP during P1 and P2, the SRP during P3 is weaker over the European portion but stronger over the Asian portion. Also, the SRP centers during P3 shift much southward and eastward by in contrast to those during P1 and P2 (Figs. 3e,f and 4a,b). The SRP during P3 is to some extent similar to the SRP2 mode in Figs. 1c and,1d in particular over Europe and China, with the exception that the SRP during P3 has six centers with a strong negative center over Japan, as well as has stronger centers than the SRP2 mode over East Asia. Along with the above analyses, it can be concluded that the SRP experienced a new interdecadal change around the late 1990s, which is characterized by a clear shift in the intensity and location of the SRP centers during the recent period.

b. Changes in the associated circulation and climate anomalies

To shed light on the interdecadal change of the SRP around the late 1990s, the circulation and climate anomalies in association with the SRP during above three subperiods are contrasted in this section. As the associated anomalies based on the NCEP1 and JRA-55 datasets are highly consistent (refer to Fig. 5), only the results based on the JRA-55 dataset are provided in the following analysis because of its higher resolution.

Fig. 5.

As in Figs. 3a–f, but for the Z200 anomalies. The vectors denote the wave active flux (units: m2 s−2).

Fig. 5.

As in Figs. 3a–f, but for the Z200 anomalies. The vectors denote the wave active flux (units: m2 s−2).

Figures 5 and 6 display the upper and lower geopotential height (Z) anomalies regressed against the SRPI during above three subperiods. Corresponding to the SRP during P1 and P2, the Z anomalies bear a similar wave train structure with strong westerly wave-activity flux in tropospheric midlatitude Eurasia, featuring an equivalent barotropic pattern that tilts slightly westward with height except the region over the northwest Indian peninsula where the Z anomalies are baroclinic. The barotropic Z anomalies consist of two positive centers over central and eastern Asia, and two negative centers over eastern Europe and Northwest China, which are mainly located within the Eurasian westerly jet and have amplitude increasing with height (Figs. 5a–d and 6a–d). In contrast to the Z anomalies associated with the SRP during P1 (Figs. 5a,b and 6a,b), the Z anomaly centers related to the SRP during P2 have a smaller meridional scale (Figs. 5c,d and 6c,d) and shift slightly eastward (Figs. 4c,d); they are also accompanied by a weak high-latitude wave train and show a closer connection to the circulation anomalies over the North Atlantic (Figs. 5c,d and 6c,d). The weak high-latitude wave train may be triggered by the disturbance over the North Atlantic and Europe (Lin et al. 2017b; Liu et al. 2020). These results are consistent with previous findings related to the SRP change around the late 1970s (Wu 2002; Wang et al. 2012; Liu et al. 2020).

Fig. 6.

As in Fig. 5, but for the (left) 850- and (right) 500-hPa Z anomalies and associated wave activity flux (units: m2 s−2) during (a),(b), 1965–79, (c),(d) 1980–98, and (e),(f) 1999–2014 based on the JRA-55 dataset.

Fig. 6.

As in Fig. 5, but for the (left) 850- and (right) 500-hPa Z anomalies and associated wave activity flux (units: m2 s−2) during (a),(b), 1965–79, (c),(d) 1980–98, and (e),(f) 1999–2014 based on the JRA-55 dataset.

During P3, the associated Z anomalies have positive centers over West Asia and central and northeastern Asia and negative centers over central China and to the southeast of Japan, which also portray an equivalent barotropic wave train structure in the upper and lower troposphere that tilts westward with amplitude increasing with height (Figs. 5e,f and 6e,f). These features show notable differences from those related to the SRP during P1 and P2. On the one hand, the significant Z anomalies during P3 are mainly situated over the southern part of the Eurasian westerly jet, whose maximum centers over West Asia and central China move southward and eastward in contrast to their counterparts during P1 and P2, but the center over northeastern Asia shifts northward in contrast to that during P2 (Figs. 4c,d). On the other hand, the associated Z anomalies during P3 weaken evidently over central Asia but enhance over West Asia. In addition, the Z anomalies associated with the SRP during P3 are insignificant over Europe and the North Atlantic, which suggests a weak connection between the SRP and the upstream disturbance over Europe and the North Atlantic after the late 1990s. The changes in the intensity and location of the Z anomalies associated with SRP further support the interdecadal change of the SRP around the late 1990s.

Corresponding to the distinct changes of the SRP, its impact on the Eurasian summer surface air temperature and precipitation also experienced significant changes around the late 1990s. As shown in Figs. 7a and 7b, the responses of the temperature and precipitation to the SRP during P1 display a well-organized wave train structure over midlatitude Eurasia, which are characterized by above-normal temperature and below-normal precipitation over western Europe, central Asia/western Russia, eastern Russia, and the area downstream of the mei-yu belt and below-normal temperature (above-normal precipitation) over eastern Europe and Northwest China/South Asia (North China/Baikal). These features are also observed in the associated temperature and precipitation anomalies during P2 except that the anomalous centers move eastward, and the associated temperature (precipitation) anomalies weaken (enhance) over Europe but enhance (weaken) over East China and Japan (Figs. 7c,d). These anomalies to a large extent coincide with the circulation anomalies related to the SRP in terms of strength and geographic position over the extratropical Eurasia during P1 and P2, respectively (Figs. 3a–d, 5a–d, and 6a–d).

Fig. 7.

(left) CRU surface temperature (Ts; units: K) and (right) PREC precipitation anomalies (Pre; units: mm) regressed against the SRPI during (a),(b) 1965–79, (c),(d) 1980–98, and (e),(f) 1999–2014. Stippling denotes the anomalies significant at 95% confidence level.

Fig. 7.

(left) CRU surface temperature (Ts; units: K) and (right) PREC precipitation anomalies (Pre; units: mm) regressed against the SRPI during (a),(b) 1965–79, (c),(d) 1980–98, and (e),(f) 1999–2014. Stippling denotes the anomalies significant at 95% confidence level.

During P3, the responses of the temperature and precipitation to the SRP exhibit remarkable differences from those during P1 and P2. The associated temperature anomalies are significantly observed to the east of 60°E over the midlatitudes of central and East Asia (Fig. 7e), which features a zonal-uniform warm pattern and corresponds well to the above-normal Z anomalies related to the SRP during P3 (Figs. 5c and 6e,f). The associated precipitation anomalies are not as well organized as the temperature anomalies, which are accompanied by negative precipitation anomalies over the Mediterranean, northern Russia, and northeastern Asia and positive anomalies over Northeast China (Fig. 7f), which largely coincide with the circulation anomalies related to the SRP during P3 (Figs. 3e,f, 5e,f, and 6e,f). The differences of the associated temperature and precipitation anomalies during the above three subperiods indicate clear changes in the impacts of the SRP on the Eurasian summer climate around the late 1970s and late 1990s, due to the two interdecadal changes in the SRP.

4. Possible causes of the SRP change around the late 1990s

Previous studies have documented that the SRP variability can be triggered or modulated by the changes of tropical/extratropical precipitation (heating) variability (Lu et al. 2002; Wu 2002; Ding and Wang 2005), the ENSO variability (Ding and Wang 2005; Wang et al. 2012), the summer NAO (Liu et al. 2020), and the climate mean state (Hoskins and Ambrizzi 1993; Yun et al. 2011). The interdecadal changes of these factors and their impacts can lead to the interdecadal changes of the SRP. For instance, the SRP change around the late 1970s is a result of the weakened impact of the ISM precipitation anomalies and the intensified impact of the summer NAO since the late 1970s, which are associated with interdecadal changes of the ENSO properties, climate mean ISM precipitation (Wu 2002; Wang et al. 2012), and the summer NAO (Liu et al. 2020), respectively. Regarding the recent SRP change around the late 1990s, we find that it may be largely attributed to the enhanced impact of the ISM precipitation anomalies over the northeastern ISM (ISM_NE; Fig. 7f), which is associated with the leading mode change of the ISM precipitation anomalies around the late 1990s. Besides, the basic flow (200-hPa zonal wind) change over Eurasia also exerts a favorable background for the SRP reshaping during recent period. To verify that, the roles of the ISM precipitation anomalies and the basic flow changes are investigated in this section.

a. Changes in the ISM precipitation anomalies around the late 1990s

To begin with, the ISM precipitation anomalies associated with the SRP during the three subperiods defined above are analyzed. As shown in Fig. 7, the SRP during P1 is strongly connected to the above-normal precipitation anomalies over the ISM and the tropical Indian Ocean (TIO) during P1 (Fig. 7b), but has a weak connection to tropical heating during P2 (Fig. 7d). These features are consistent with Wang et al. (2012) and Liu et al. (2020), and they pointed out that the SRP variability is mainly controlled by the ISM/TIO heating (precipitation) anomalies prior to the late 1970s, and the weakened impact of the ISM/TIO heating anomalies on the SRP since the late 1970s due to the changes in the ENSO properties and ISM precipitation anomalies (Wu 2002; Wang et al. 2012) is considered as a major cause for the SRP change around the late 1970s.

As for the recent period, the SRP shows a significant linkage to the positive precipitation anomalies over the ISM_NE (Fig. 7f), which are obviously different from those during P1 and P2. In combination with the Z anomalies corresponding to the SRP during P3 (Figs. 5e,f and 6e,f), the positive and negative Z anomalies over the northwest and northeast Indian peninsula feature a Rossby wave type response to the above-normal ISM_NE precipitation anomalies (Gill 1980). This indicates that the SRP variation since the late 1990s may be contributed largely by the ISM_NE precipitation anomalies, which can be seen in the following analysis.

The above results indicate clear differences of the ISM precipitation anomalies associated with the SRP during the above three subperiods. Given that the changes in the ISM precipitation anomalies can lead to the changes in the SRP, it is hypothesized that the leading mode of the ISM precipitation anomalies may experience an interdecadal change around the late 1990s, which leads to the recent SRP reshaping. To confirm that, the changes of the EOF1 modes of the land precipitation anomalies over the ISM region (5°–35°N, 65°–95°E) during above three subperiods and their impacts on the SRP are investigated in the following analyses.

For easy explanation, the three EOF1 modes are named ISM_P1, ISM_P2, and ISM_P3, respectively, which are demonstrated by the ISM precipitation anomalies regressed against the corresponding PC during each subperiod. The three EOF1 modes respectively account for 31.7%, 27.6%, and 35% of the total variance of the ISM precipitation anomalies, which are well separated from the corresponding EOF2 mode during each subperiod based on North’s method (North et al. 1982). As shown in Figs. 8a and 8b, the ISM_P1 and ISM_P2 exhibit a similar zonal dipolar structure with positive anomalies and negative anomalies over the west and east flanks of 85°E, respectively, except that the explanation percentage and the positive centers of ISM_P2 are weaker than those of ISM_P1 (Fig. 8b). This indicate a reduced ISM precipitation variability around the late 1970s, consistent with Wu (2002) and Wang et al. (2012). With respect to the ISM_P3, it reveals a nearly monosign pattern with significantly above-normal precipitation anomalies mainly over the ISM_NE (Fig. 8c), showing clearly different from the ISM_P1 and ISM_P2. These differences between the three EOF1 modes indicate two interdecadal changes in leading mode of the ISM precipitation anomalies around the late 1970s and the late 1990s. The two timings are also verified using the 15-yr sliding EOF analysis on the ISM precipitation anomalies (figure not shown).

Fig. 8.

As in Fig. 3, but for EOF1 of the CRU precipitation anomalies (contours; units: mm) over the domain 5°–35°N, 65°–95°E during (a) 1965–79, (b) 1980–98, and (c) 1999–2014. (d) The change ratio (%) of the standard deviation of the precipitation anomalies during 1999–2014 relative to the standard deviation during 1965–2014.

Fig. 8.

As in Fig. 3, but for EOF1 of the CRU precipitation anomalies (contours; units: mm) over the domain 5°–35°N, 65°–95°E during (a) 1965–79, (b) 1980–98, and (c) 1999–2014. (d) The change ratio (%) of the standard deviation of the precipitation anomalies during 1999–2014 relative to the standard deviation during 1965–2014.

In contrast to Fig. 7, it is noticed that ISM_P1 and ISM_P3 show similar patterns to the ISM precipitation anomalies related to the SRP during P1 and P3, respectively, whereas ISM_P2 differs largely from those during P2. The correlation coefficients between the SRP index and the principal component of ISM_P1, ISM_P2, and ISM_P3 are 0.55, 0.15, and 0.6, respectively, which are significant beyond 95%, 45%, and 99% confidence level, respectively. This further suggests the close connections of the SRP to the ISM precipitation anomalies during P1 and P3.

b. Contribution of the ISM precipitation change to the recent SRP reshaping

To shed light on the impacts of the ISM precipitation anomalies on the SRP changes, the circulation anomalies related to the above three ISM EOF1 mode are investigated. Figure 9 displays the V200 and 200-hPa Z (Z200) anomalies associated with the three EOF1 modes of the ISM precipitation anomalies. Corresponding to the ISM_P1, the V200 and Z200 anomalies over the Eurasia, to a large extent, feature similar wave train patterns to the V200 and Z200 anomalies related to the SRP during P1, but the anomalies associated with the ISM_P1 are a little weaker and have lower significant confidence level in particular for the V200 anomalies (Figs. 9a,b). This implies that the ISM_P1 seems exerts a weak impact on the SRP during P1, which explains only about 30% of the SRP variations during P1. This relative weak linkage between them may lie on the one hand that in addition to the ISM precipitation anomalies, the TIO precipitation anomalies also have a strong influence on the SRP variations during P1 (Fig. 7b), and on the other hand in that the ISM precipitation anomalies related to the SRP during P1 are significant over the western ISM (ISM_W), slightly different from those related to the ISM_P1. This means that the western ISM precipitation anomalies may explain a larger percentage of the SRP variations during P1. To reveal that, we define a regional precipitation index over ISM_W (ISM_WI; 15°–35°N, 70°–85°E) and re-examine its correlation with the SRP index and association with the V200 and Z200 anomalies during P1. The correlation of the ISM_WI with the SRP index increases to 0.70 during P1. And the associated V200 and Z200 anomalies are more coherent with those related to the SRP index during P1 (Figs. 10a,b).

Fig. 9.

(left) V200 and (right) Z200 anomalies regressed against (a),(b) ISM_P1 (1965–79), (c),(d) ISM_P2 (1980–98), and (e),(f) ISM_P3(1999–2014). Shading denotes the anomalies significant beyond the 95% confidence level. Vectors are the associated wave activity flux (units: m2 s−2).

Fig. 9.

(left) V200 and (right) Z200 anomalies regressed against (a),(b) ISM_P1 (1965–79), (c),(d) ISM_P2 (1980–98), and (e),(f) ISM_P3(1999–2014). Shading denotes the anomalies significant beyond the 95% confidence level. Vectors are the associated wave activity flux (units: m2 s−2).

Fig. 10.

As in Fig. 9, but for (a),(b) ISM_WI during 1965–79 and (c),(d) ISM_NEI during 1999–2014.

Fig. 10.

As in Fig. 9, but for (a),(b) ISM_WI during 1965–79 and (c),(d) ISM_NEI during 1999–2014.

The associated V200 and Z200 anomalies associated with the ISM_P2 exhibit weak wave train patterns over mid- and high-latitude Eurasia and have significant centers mainly over the North Atlantic and western Europe, which differ largely from the circulation anomalies related to the SRP during P2 (Figs. 9c,d). The weak connection between the SRP and ISM heating during P2 is coherent with above analyses and consistent with previous studies (e.g., Wu 2002; Wang et al. 2012; Lin et al. 2017a,b; Liu et al. 2020). As documented by Liu et al. (2020), the SRP variations during P2 is controlled largely by the summer NAO in particular on the 4–8-yr time scale, which is associated with the interdecadal change of the summer NAO around the late 1970s.

The V200 and Z200 anomalies associated with the ISM_P3 are shown in Figs. 9e and 9f, which resemble highly the circulation anomalies related to the SRP during P3 (Figs. 3c and 5e,f), with similar wave train structures and similar geographically fixed centers along the southern part of the Eurasian westerly jet (Figs. 9e,f), but the anomalies associated with the ISM_P3 are a little weaker especially around the Mediterranean Sea and over northeast Asia. This is because the precipitation anomalies related to the SRP index during P3 are significantly over the ISM_NE, which spread a large domain than those related to the ISM_P3. This means that the ISM_P3 may underestimate the role of the ISM_NE precipitation anomalies on the SRP variations during P3. Likewise, we define a regional precipitation index over ISM_NE (ISM_NEI; 20°–35°N, 75°–95°E) and re-examine its relationship with the SRP index and the V200 and Z200 anomalies during P3. The coefficient between the ISM_NEI and the SRP index during P3 is about 0.85, significant beyond the 99% confidence level. This means that the ISM_NE precipitation anomalies explain about 72% of the SRP variations during P3, which is also supported by the associated V200 and Z200 anomalies (Figs. 10c,d) that are more coherent with those related to the SRP index during P3 (Figs. 3e,f and 5e,f). Meanwhile, the positive Z200 anomalies over northeastern Asia are still weaker and insignificant, which may be partially associated with the high-latitude disturbance over Eurasia (Figs. 5e,f and 6e,f).

Regarding how the ISM precipitation anomalies excite the SRP, two processes have been documented in many previous studies (Lu et al. 2002; Ding and Wang 2005; Chen and Huang 2012). The first one is the “monsoon-desert” mechanism (Rodwell and Hoskins 1996), which means that the ISM heating anomalies could affect the central Asia circulation through the Rossby wave response and further excite the SRP. As indicated above, both the positive Z200 anomalies over central Asia associated with the ISM_WI (Fig. 10b)/SRP during P1 (Figs. 5a,b) and ISM_NEI (Fig. 10d)/SRP during P3 (Figs. 5e,f) feature Rossby wave responses to the heating forcing induced by the precipitation anomalies over the ISM_W and ISM_NE.

Another process is the Rossby wave source (RWS) mechanism, which is associated with the wind divergence and the advection of the vorticity by its divergent flow (Sardeshmukh and Hoskins 1988). As Chen and Huang (2012) documented that the advection of vorticity by the divergent wind induced by the ISM heating serves as an effective RWS for the extratropical teleconnection patterns (Sardeshmukh and Hoskins 1988; Chen and Huang 2012), through which the ISM precipitation anomalies excite the SRP. To verify this mechanism, the wind divergence, velocity potential at 850 and 200 hPa, 500-hPa vertical motion, and 200-hPa Rossby wave source anomalies related to the ISM_P1 and ISM_P3 are explored. Here, the ISM_WI and ISM_NEI are employed in the analysis. The RWS is defined as S = −∇ ⋅ vx(f + ζ), where vx is the divergent wind velocity (Sardeshmukh and Hoskins 1988). It can be further divided into two parts: 1) the advection term due to the vorticity advection by the divergent flow [−vx ⋅ ∇(f + ζ)] and 2) the divergence term related to the wind divergence [−(f + ζ)∇ ⋅ vx].

By comparing with the study of Chen and Huang (2012), it is noticed that the SRP during P1 actually resembles well the SRP in July (Fig. 11a in their study) that also has a similar relationship to the ISM and TIO precipitation anomalies. As Chen and Huang (2012) documented, the ISM/TIO heating anomalies exit the SRP by inducing divergent flow at the upper troposphere and the advection of the vorticity by the divergent flow act as effective Rossby wave sources for the SRP (Figs. 12a and 13a in their study). Similar analysis is conducted to reveal the role of the ISM_W heating anomalies on the SRP variation during P1. As shown in Fig. 11a herein, the positive precipitation (heating) anomalies over ISM_W (Fig. 7b) are associated with significant ascent over ISM and the eastern part of West Asia, which leads to significant divergence at the upper troposphere that is westward titled in contrast to the ISM_W heating center (Figs. 11b,c). Correspondingly, a significant anticyclonic vorticity source (RWS) is observed over central Asia and the northern ISM (Fig. 11d). Being located to the southwest of the positive center of the SRP during P1 (Figs. 5a,b and 10b), this RWS is strongest in the upper reaches of the SRP wave train (Fig. 11d), indicating its dominant contribution to the excitation of the SRP during P1. An inspection of the two parts of the RWS suggests that the RWS is mainly contributed by the divergence term (Fig. 11e), which results from the anomalous divergence induced by the ISM_W heating, while the advection of the vorticity by divergent flow exerts a weak contribution to the RWS (Fig. 11f), which is different from Chen and Huang (2012).

Fig. 11.

Anomalies of the (a) 500-hPa vertical velocity, (b) 200-hPa wind divergence, (c) 200-hPa velocity potential (VP) and divergent wind, (d) 200-hPa Rossby wave sources (RWSs), (e) RWS divergence term, and (f) RWS advection term regressed against the ISM_WI during P1 (1965–79). Shading denotes the anomalies significant beyond the 95% confidence level.

Fig. 11.

Anomalies of the (a) 500-hPa vertical velocity, (b) 200-hPa wind divergence, (c) 200-hPa velocity potential (VP) and divergent wind, (d) 200-hPa Rossby wave sources (RWSs), (e) RWS divergence term, and (f) RWS advection term regressed against the ISM_WI during P1 (1965–79). Shading denotes the anomalies significant beyond the 95% confidence level.

With respect to the ISM_NE heating, it triggers strong ascent with two centers over the northern and eastern ISM, respectively (Fig. 12a). In addition, another weaker ascent center is observed to the south of central Asia (Fig. 12a). The anomalous ascent results in significant upper-level divergence over the northern ISM and central Asia (Figs. 12b,c). As for the RWS, a significant anticyclonic vorticity source is observed with three maximum centers over the northern and eastern ISM and the south of central Asia (Fig. 12d), respectively, which is also contributed largely by the divergent term associated with the upper-level divergence induced by the ISM_NE heating (Fig. 12e). This RWS is strongest over the southwest of the positive center (upper reaches) of the SRP during P3 (Figs. 5e,f and 10d), indicating its dominant contribution to the excitation of the SRP during P3. In contrast to those related to the ISM_W heating (Fig. 11), the ascent, divergence, and RWS anomalies are weaker, and their maximum centers shift eastward. These differences may be responsible for the distinct structure and decreased intensity of the SRP during P3.

Fig. 12.

As in Fig. 11, but for the ISM_NEI during P3 (1999–2014).

Fig. 12.

As in Fig. 11, but for the ISM_NEI during P3 (1999–2014).

Furthermore, the role of ISM_NE precipitation anomalies with regard to the SRP is verified with a barotropic model to imposed idealized divergence and convergence anomalies. The barotropic model formulation follows Sardeshmukh and Hoskins (1988):

 
ηt+Vψη=(Vχη)+damping,

where η is the absolute vorticity, and Vψ (Vχ) is the rotational (divergent) component of the wind. The first term on the right-hand side of the formulation is the so-called Rossby wave source. The damping term includes two components: 1) linear dissipation with a 60-day e-folding time scale and 2) biharmonic diffusion. The barotropic model is a spectral model with truncation at rhomboidal wavenumber 40.

We perform two kinds of experiments with an integration for 40 days: one with the climatological summer mean divergence during P3 as the control run (CTL) and the other with the same climatological summer mean divergence plus prescribed divergence anomalies as the forcing run (FCR). The divergence anomaly in the FCR is prescribed over the ISM with a maximum intensity of 4 × 10−6 s−1 at the western center (W: 32°N, 65°E) and 6 × 10−6 s−1 at the central center (C: 28°N, 78°E) and eastern center (E: 25°N, 95°E), which is selected according to the three significant divergence centers over the ISM in Fig. 12b. Another FCR experiment is also conducted with prescribed divergence over both C and W centers (referred to as C&E). In addition, a set of FCR experiment with prescribed convergence over W, C, E, and C&E are conducted for comparison. The differences of the Z200 anomalies between the FCR and CTL averaged over the model days 31–40 are considered as the response of the Z200 to specific forcing.

By comparing the responses of Z200 to the individual forcing (shown in Fig. S1 in the online supplemental material), we find that the anomalous divergence/convergence over C and E exerts an essential role in the SRP formation. Here, the responses of Z200 to the C&E forcing are provided. As shown in Figs. 13a and 13b, the Z200 anomalies corresponding to the C&E divergence and convergence feature nearly a symmetrical structure over the subtropical Eurasia, which are characterized by a wavelike pattern over Eurasia with positive (negative) anomalies over western Mediterranean Sea, west-central Asia, and Northeast Asia, and negative (positive) anomalies over East Asia. The differences between the Z200 responses to the C&E divergence and convergence are similar to those to the C&E divergence (Fig. 13c), which resemble Z200 anomalies associated with ISM_NE/ISM_P3 and SRP during P3 although the anomalous centers shift westward and the lower reaches of the SRP over northern East Asia are not reproduced in the simulation. The model results to some extent support the above findings that the ISM_NE precipitation anomalies play an essential role in the recent SRP reshaping.

Fig. 13.

Response of the Z200 anomalies to the (a) divergence and (b) convergence forcing (shading) over C&E under the climatological mean basic state during P3 and (c) their differences [(a) minus (b)] in a barotropic model.

Fig. 13.

Response of the Z200 anomalies to the (a) divergence and (b) convergence forcing (shading) over C&E under the climatological mean basic state during P3 and (c) their differences [(a) minus (b)] in a barotropic model.

A natural question arises of why the leading mode of the ISM precipitation anomalies changed around the late 1970s and late 1990s. Wang et al. (2012) attributed the reduced ISM precipitation variability to the weakened ISM–ENSO connection since the late 1970s that is associated with the change of ENSO properties around the late 1970s. Regarding the recent change of the ISM precipitation mode around the late 1990s, one plausible reason is the change in the distribution of ISM precipitation variability. As shown in Fig. 8d, the standard deviation of the ISM precipitation anomalies increases about 20%–60% over most of the ISM_NE during the recent period. The enhanced deviation favors the shift of the leading mode of the ISM precipitation from ISM_P2 to ISM_P3 around the late 1990s.

c. The role of background changes

As mentioned above, the SRP is a quasi-stationary wave train along the summer Eurasian westerly jet that can serve as an efficient Rossby waveguide (Hoskins and Ambrizzi 1993). The changes in the westerly jet over Eurasia can modulate the atmospheric teleconnections (Yun et al. 2011; Huang et al. 2013; Lin et al. 2017b). Figure 14a displays the interdecadal changes of the 200-hPa zonal wind (U200) around the late 1990s, which is defined as the differences between the climatology mean of the U200 during P3 and P2. The changes in the U200 over Eurasia are characterized by significant easterly anomalies in the southern flank and the exit of the East Asian westerly jet and westerly anomalies in the northern flank (Fig. 14a), which are nearly opposite to the changes around the late 1970s (Fig. 3 in Wang et al. 2012).

Fig. 14.

(a) The climatological mean zonal wind at 200 hPa during 1999–2014 (red contours) and 1979–98 (blue contours) and their differences (shading). (b) The climatological mean Rossby wavenumber at 200 hPa during 1979–98 (blue contours) and the differences between the Rossby wavenumber averaged during 1999–2014 and 1979–98 (shading). Stippling denotes the anomalies significant at the 95% confidence level. (c) The difference between the Z200 responses to the same divergence (shading) over C&E under climatological mean basic state during P3 and P2.

Fig. 14.

(a) The climatological mean zonal wind at 200 hPa during 1999–2014 (red contours) and 1979–98 (blue contours) and their differences (shading). (b) The climatological mean Rossby wavenumber at 200 hPa during 1979–98 (blue contours) and the differences between the Rossby wavenumber averaged during 1999–2014 and 1979–98 (shading). Stippling denotes the anomalies significant at the 95% confidence level. (c) The difference between the Z200 responses to the same divergence (shading) over C&E under climatological mean basic state during P3 and P2.

In theory, the stationary Rossby wavenumber (Ks) of the SRP is determined by the basic state (Hoskins and Ambrizzi 1993), which is defined as Ks=β(2U¯/y2)/U¯, where β is meridional gradient of the Coriolis parameter, and U¯ is the climatology of zonal westerly wind. Figure 14b shows the wavenumber at 200 hPa for the basic flow during P2 and its change since the late 1990s. During P2, a maximum wavenumber with wavenumbers 5–7 corresponding the SRP wave train appears around the westerly jet that serves as a waveguide of the SRP. Since the late 1990s, the wavenumber is significantly increased over the southern flank of the Asian westerly jet where the wavenumber is larger than 6 (Fig. 14b), indicating a southward shift of the dominant wave train center over subtropical and midlatitude Eurasia. These changes in the wavenumber over Eurasia favor the southward shift of the Rossby wave propagation, and lead to in the southward movement of the SRP centers over East Asia during the recent period.

To verify the role of the basic-state change in the recent SRP reshaping, a similar set of FCR experiments with same C&E divergence but with climatological mean conditions during P2 is conducted using the barotropic model (Fig. S2). Figure 14c is the difference between the Z200 responses to the basic state during P3 and P2. It is noticed that, corresponding to same C&E divergence, the basic-state change results in a wave train to the south of 40°N over Eurasia, which has positive centers over West Asia/ISM and northeastern Asia and a negative center over East Asia, resembling the SRP during P3. The SRP is the dominant wave train pattern over Eurasia, which has active centers along about 42°N (Fig. 4) during P1 and P2. The model results indicate that the basic-state change favors the southward shift of the SRP wave train during P3 in particular during its positive phase.

5. Conclusions

The SRP is a well-known atmospheric teleconnection within the Eurasian summer westerly jet and usually defined as the first EOF mode of the V200 anomalies over Eurasia. In general, it accounts for about 36% of the total variance of the V200 anomalies over Eurasia during 1965–2014 and reveals a largely interannual variability. The SRP features a wavelike pattern with five well-defined centers over northern Europe, eastern Europe, central Asia, North China, and Japan having the downstream four centers within the westerly jet along about 42°N (in terms of V200 anomalies). Previous studies have indicated an interdecadal change in the intensity and periodicity of the SRP around the late 1970s (Wu 2002; Wang et al. 2012; Liu et al. 2020). Through further analyses, the present study indicates a new change of the SRP around the late 1990s. The features and possible causes of the recent SRP change and their differences from those related to the SRP change around the late 1970s are investigated in this study. The main results are summarized as follows.

Prior to the late 1970s (P1: 1965–78), the SRP accounts for about 52% of the total variance of the V200 anomalies over Eurasia and resembles the typical SRP pattern during the whole period but with stronger centers. It shows a strong linkage to the ISM precipitation anomalies and ENSO. During the period from the late 1970s to the late 1990s (P2: 1980–98), the SRP becomes weak (in terms of V200 anomalies) and its explanation percentage of the SRP decreases to about 35%. It features a similar pattern to the typical SRP during the whole period and P1 but has a weaker center over Japan. By contrast, the SRP shows a strong connection to the circulation anomalies over the North Atlantic but the connection to ISM/ENSO weakens during P2. The SRP variations during this period are dominated by the summer NAO, in particular on the 4–8-yr time scale (Liu et al. 2020). This interdecadal change of the SRP has been documented by previous studies and is attributed to the weakened impact of the ISM precipitation anomalies due to the reduced ISM variability (Wu 2002; Wang et al. 2012) and the intensified impact of the summer NAO associated with the interdecadal change of the summer NAO (Liu et al. 2020).

As for the recent period (P3: 1999–2014), the SRP exhibits a different wavelike pattern from the SRP during P1 and P2, which has six active centers along the 36°N with southerly and northerly anomalies over central Europe, West Asia, central Asia, Northwest China, East China, and Japan (in terms of V200 anomalies). In contrast to the SRP during P1 and P2, the SRP centers during P3 shift southward and eastward, and weaken significantly over West Asia and central Asia. The SRP center over Japan during P3 is weaker than that during P1 but stronger than that during P2. The SRP during this period is closely related to the ISM_NE precipitation anomalies.

We find that the interdecadal changes of the SRP show close associations with the interdecadal changes in leading mode of the ISM precipitation anomalies. The recent SRP change is attributed to the leading mode change of the ISM precipitation anomalies around the late 1990s. During the recent period (P3), the EOF1 of the ISM precipitation anomalies exhibits a nearly monosign pattern with significantly above-normal precipitation anomalies over ISM_NE, which is significantly distinct from the zonal dipolar patterns during P1 and P2. The recent change of the ISM precipitation anomalies results from the interdecadal enhancement of ISM precipitation variability over ISM_NE, which induces upper-level circulation anomalies over south of central Asia through the “monsoon-desert” mechanism that acts as effective Rossby wave sources and triggers the SRP during P3. The role of the ISM_NE precipitation anomalies in the SRP variation during the recent period is verified by the results of simulation with a barotropic model.

The basic-state changes over Eurasia around the late 1990s may also contribute to the recent SRP reshaping. The weakened upper westerly jet in its southern flank and the exit over East Asia during recent period results in the Rossby wavenumber increasing over the southern flank of the Asian westerly jet, which favors the southward shift of the Rossby wave and corresponds to the southward movement of the SRP centers over East Asia during the recent period.

Corresponding to the two interdecadal changes, the impact of the SRP on the summer climate over Eurasia also showed clear shifts around the late 1970s and late 1990s, which coincide well with the associated circulation changes. The present study has implications for understanding the variability and prediction of the SRP, which is helpful for seasonal forecasting of the summer climate over Eurasia. In addition, on the interdecadal time scale, the SRP phase also reveals significant oscillations around the late 1970s and the mid- to late 1990s (e.g., Hong et al. 2017; Wang et al. 2017); their interaction with the changes of interannual SRP still needs further investigation.

Acknowledgments

We thank the three anonymous reviewers for their constructive comments that led to the improvements of the manuscripts. This study is jointly supported by the National Key Research and Development Program (Grant 2016YFA0600603) and the National Natural Science Foundation of China (Grants 41605058, 41530425, and 41831175).

REFERENCES

REFERENCES
Bretherton
,
C. S.
,
M.
Widmann
,
V. P.
Dymnikov
,
J. M.
Wallace
, and
I.
Bladé
,
1999
:
The effective number of spatial degrees of freedom of a time-varying field
.
J. Climate
,
12
,
1990
2009
, https://doi.org/10.1175/1520-0442(1999)012<1990:TENOSD>2.0.CO;2.
Chen
,
G.
, and
R.
Huang
,
2012
:
Excitation mechanisms of the teleconnection patterns affecting the July precipitation in Northwest China
.
J. Climate
,
25
,
7834
7851
, https://doi.org/10.1175/JCLI-D-11-00684.1.
Chen
,
G.
,
R.
Huang
, and
L.
Zhou
,
2013
:
Baroclinic instability of the Silk Road pattern induced by thermal damping
.
J. Atmos. Sci.
,
70
,
2875
2893
, https://doi.org/10.1175/JAS-D-12-0326.1.
Chen
,
M.
,
P.
Xie
,
J. E.
Janowiak
, and
P. A.
Arkin
,
2002
:
Global land precipitation: A 50-yr monthly analysis based on gauge observations
.
J. Hydrometeor.
,
3
,
249
266
, https://doi.org/10.1175/1525-7541(2002)003<0249:GLPAYM>2.0.CO;2.
Dai
,
A.
,
2013
:
The influence of the inter-decadal Pacific oscillation on US precipitation during 1923–2010
.
Climate Dyn.
,
41
,
633
646
, https://doi.org/10.1007/s00382-012-1446-5.
Deser
,
C.
,
A. S.
Phillips
,
J. W.
Hurrell
,
2004
:
Pacific interdecadal climate variability: Linkages between the tropics and the North Pacific during boreal winter since 1900
.
J. Climate
,
17
,
3109
3124
, https://doi.org/10.1175/1520-0442(2004)017<3109:PICVLB>2.0.CO;2.
Ding
,
Q.
, and
B.
Wang
,
2005
:
Circumglobal teleconnection in the Northern Hemisphere summer
.
J. Climate
,
18
,
3483
3505
, https://doi.org/10.1175/JCLI3473.1.
Ding
,
Q.
,
B.
Wang
,
J. M.
Wallace
, and
G.
Branstator
,
2011
:
Tropical–extratropical teleconnections in boreal summer: Observed interannual variability
.
J. Climate
,
24
,
1878
1896
, https://doi.org/10.1175/2011JCLI3621.1.
Duchon
,
C. E.
,
1979
:
Lanczos filtering in one and two dimensions
.
J. Appl. Meteor.
,
18
,
1016
1022
, https://doi.org/10.1175/1520-0450(1979)018<1016:LFIOAT>2.0.CO;2.
Enomoto
,
T.
,
B. J.
Hoskins
, and
Y.
Matsuda
,
2003
:
The formation mechanism of the Bonin high in August
.
Quart. J. Roy. Meteor. Soc.
,
129
,
157
178
, https://doi.org/10.1256/qj.01.211.
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.
Harris
,
I.
,
P. D.
Jones
,
T. J.
Osborn
, and
D. H.
Lister
,
2014
:
Updated high-resolution grids of monthly climatic observations—The CRU TS3.10 dataset
.
Int. J. Climatol.
,
34
,
623
642
, https://doi.org/10.1002/joc.3711.
Hong
,
X.
,
R.
Lu
, and
S.
Li
,
2017
:
Amplified summer warming in Europe–West Asia and northeast Asia after the mid-1990s
.
Environ. Res. Lett.
,
12
,
094007
, https://doi.org/10.1088/1748-9326/aa7909.
Hong
,
X.
,
S.-H.
Xue
,
R.-Y.
Lu
, and
Y.-Y.
Liu
,
2018
:
Comparison between the interannual and decadal components of the Silk Road pattern
.
Atmos. Oceanic Sci. Lett.
,
11
,
270
274
, https://doi.org/10.1080/16742834.2018.1439661.
Hoskins
,
B. J.
, and
T.
Ambrizzi
,
1993
:
Rossby wave propagation on a realistic longitudinally varying flow
.
J. Atmos. Sci.
,
50
,
1661
1671
, https://doi.org/10.1175/1520-0469(1993)050<1661:RWPOAR>2.0.CO;2.
Huang
,
G.
,
Y.
Liu
, and
H.
Ronghui
,
2011
:
The interannual variability of summer rainfall in the arid and semiarid regions of Northern China and its association with the Northern Hemisphere circumglobal teleconnection
.
Adv. Atmos. Sci.
,
28
,
257
268
, https://doi.org/10.1007/s00376-010-9225-x.
Huang
,
R.
,
Y.
Liu
, and
T.
Feng
,
2013
:
Interdecadal change of summer precipitation over Eastern China around the late-1990s and associated circulation anomalies, internal dynamical causes
.
Chin. Sci. Bull.
,
58
,
1339
1349
, https://doi.org/10.1007/s11434-012-5545-9.
Huang
,
R.
,
Y.
Liu
,
Z.
Du
,
J.
Chen
, and
J.
Huangfu
,
2017
:
Differences and links between the East Asian and South Asian summer monsoon systems: Characteristics and variability
.
Adv. Atmos. Sci.
,
34
,
1204
1218
, https://doi.org/10.1007/s00376-017-7008-3.
Kobayashi
,
S.
, and et al
,
2015
:
The JRA-55 reanalysis: General specifications and basic characteristics
.
J. Meteor. Soc. Japan
,
93
,
5
48
, https://doi.org/10.2151/jmsj.2015-001.
Kalnay
,
E.
, and et al
,
1996
:
The NCEP/NCAR 40-Year Reanalysis Project
.
Bull. Amer. Meteor. Soc.
,
77
,
437
471
, https://doi.org/10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.
Kosaka
,
Y.
,
H.
Nakamura
,
M.
Watanabe
, and
M.
Kimoto
,
2009
:
Analysis on the dynamics of a wave-like teleconnection pattern along the summertime Asian jet based on a reanalysis dataset and climate model simulations
.
J. Meteor. Soc. Japan
,
87
,
561
580
, https://doi.org/10.2151/jmsj.87.561.
Kosaka
,
Y.
,
S.-P.
Xie
, and
H.
Nakamura
,
2011
:
Dynamics of interannual variability in summer precipitation over East Asia
.
J. Climate
,
24
,
5435
5453
, https://doi.org/10.1175/2011JCLI4099.1.
Kosaka
,
Y.
,
J. S.
Chowdary
,
S.-P.
Xie
,
Y.-M.
Min
, and
J.-Y.
Lee
,
2012
:
Limitations of seasonal predictability for summer climate over East Asia and the northwestern Pacific
.
J. Climate
,
25
,
7574
7589
, https://doi.org/10.1175/JCLI-D-12-00009.1.
Lee
,
J. Y.
,
B.
Wang
,
Q.
Ding
,
K. J.
Ha
,
J. B.
Ahn
,
A.
Kumar
,
B.
Stern
, and
O.
Alves
,
2011
:
How predictable is the Northern Hemisphere summer upper-tropospheric circulation?
Climate Dyn.
,
37
,
1189
1203
, https://doi.org/10.1007/s00382-010-0909-9.
Leung
,
M. Y. T.
, and
W.
Zhou
,
2018
:
Circumglobal teleconnection and eddy control of variation in summer precipitation over Northwest China
.
Climate Dyn.
,
51
,
1351
1362
, https://doi.org/10.1007/s00382-017-3958-5.
Lin
,
Z.
,
F.
Liu
,
B.
Wang
,
R.
Lu
, and
X.
Qu
,
2017a
:
Southern European rainfall reshapes the early-summer circumglobal teleconnection after the late 1970s
.
Climate Dyn.
,
48
,
3855
3868
, https://doi.org/10.1007/s00382-016-3306-1.
Lin
,
Z.
,
R.
Lu
, and
R.
Wu
,
2017b
:
Weakened impact of the Indian early summer monsoon on North China rainfall around the late 1970s: Role of basic-state change
.
J. Climate
,
30
,
7991
8005
, https://doi.org/10.1175/JCLI-D-17-0036.1.
Liu
,
Y.
, and
R.
Huang
,
2019
:
Linkages between the South and East Asian monsoon water vapor transport during boreal summer
.
J. Climate
,
32
,
4509
4524
, https://doi.org/10.1175/JCLI-D-18-0498.1.
Lu
,
R. Y.
,
J. H.
Oh
, and
B. J.
Kim
,
2002
:
A teleconnection pattern in upper-level meridional wind over the North African and Eurasian continent in summer
.
Tellus
,
54A
,
44
55
, https://doi.org/10.1034/J.1600-0870.2002.00248.X.
Nitta
,
T.
, and
S.
Yamada
,
1989
:
Recent warming of tropical sea surface temperature and its relationship to the Northern Hemisphere circulation
.
J. Meteor. Soc. Japan
,
67
,
375
383
, https://doi.org/10.2151/JMSJ1965.67.3_375.
North
,
G. R.
,
T. L.
Bell
,
R. F.
Cahalan
, and
F. J.
Moeng
,
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<0699:SEITEO>2.0.CO;2.
Peterson
,
W. T.
, and
F. B.
Schwing
,
2003
:
A new climate regime in northeast Pacific ecosystems
.
Geophys. Res. Lett.
,
30
,
1896
, https://doi.org/10.1029/2003GL017528.
Rayner
,
N. A.
,
2003
:
Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century
.
J. Geophys. Res.
,
108
,
4407
, https://doi.org/10.1029/2002JD002670.
Rodwell
,
M. J.
, and
B. J.
Hoskins
,
1996
:
Monsoons and the dynamics of deserts
.
Quart. J. Roy. Meteor. Soc.
,
122
,
1385
1404
, https://doi.org/10.1002/qj.49712253408.
Sardeshmukh
,
P. D.
, and
B. J.
Hoskins
,
1988
:
The generation of global rotational flow by steady idealized tropical divergence
.
J. Atmos. Sci.
,
45
,
1228
1251
, https://doi.org/10.1175/1520-0469(1988)045<1228:TGOGRF>2.0.CO;2.
Sato
,
N.
, and
M.
Takahashi
,
2006
:
Dynamical processes related to the appearance of quasi-stationary waves on the subtropical jet in the midsummer Northern Hemisphere
.
J. Climate
,
19
,
1531
1544
, https://doi.org/10.1175/JCLI3697.1.
Si
,
D.
, and
Y.
Ding
,
2016
:
Oceanic forcings of the interdecadal variability in East Asian summer rainfall
.
J. Climate
,
29
,
7633
7649
, https://doi.org/10.1175/JCLI-D-15-0792.1.
Song
,
F.
,
T.
Zhou
, and
L.
Wang
,
2013
:
Two modes of the Silk Road pattern and their interannual variability simulated by LASG/IAP AGCM SAMIL2.0
.
Adv. Atmos. Sci.
,
30
,
908
921
, https://doi.org/10.1007/s00376-012-2145-1.
Sutton
,
R. T.
, and
B.
Dong
,
2012
:
Atlantic Ocean influence on a shift in European climate in the 1990s
.
Nat. Geosci.
,
5
,
788
792
, https://doi.org/10.1038/ngeo1595.
Takaya
,
K.
, and
H.
Nakamura
,
2001
:
A formulation of a phase-independent wave-activity flux for 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.
Trenberth
,
K. E.
,
1990
:
Recent observed interdecadal climate changes in the Northern Hemisphere
.
Bull. Amer. Meteor. Soc.
,
71
,
988
993
, https://doi.org/10.1175/1520-0477(1990)071<0988:ROICCI>2.0.CO;2.
Wang
,
H.
,
B.
Wang
,
F.
Huang
,
Q.
Ding
, and
J. Y.
Lee
,
2012
:
Interdecadal change of the boreal summer circumglobal teleconnection (1958–2010)
.
Geophys. Res. Lett.
,
39
,
L12704
, https://doi.org/10.1029/2012GL052371.
Wang
,
L.
,
P.
Xu
,
W.
Chen
, and
Y.
Liu
,
2017
:
Interdecadal variations of the Silk Road pattern
.
J. Climate
,
30
,
9915
9932
, https://doi.org/10.1175/JCLI-D-17-0340.1.
Wang
,
W. W.
,
W.
Zhou
,
X.
Wang
,
S. K.
Fong
, and
K. C.
Leung
,
2013
:
Summer high temperature extremes in southeast China associated with the East Asian jet stream and circumglobal teleconnection
.
J. Geophys. Res. Atmos.
,
118
,
8306
8319
, https://doi.org/10.1002/jgrd.50633.
Wilks
,
D. S.
,
2016
:
“The stippling shows statistically significant grid points”: How research results are routinely overstated and overinterpreted, and what to do about it
.
Bull. Amer. Meteor. Soc.
,
97
,
2263
2273
, https://doi.org/10.1175/BAMS-D-15-00267.1.
Wu
,
R.
,
2002
:
A mid-latitude Asian circulation anomaly pattern in boreal summer and its connection with the Indian and East Asian summer monsoons
.
Int. J. Climatol.
,
22
,
1879
1895
, https://doi.org/10.1002/joc.845.
Wu
,
R.
,
2017
:
Relationship between Indian and East Asian summer rainfall variations
.
Adv. Atmos. Sci.
,
34
,
4
15
, https://doi.org/10.1007/s00376-016-6216-6.
Wu
,
R.
, and
B.
Wang
,
2002
:
A contrast of the East Asian summer monsoon–ENSO relationship between 1962–77 and 1978–93
.
J. Climate
,
15
,
3266
3279
, https://doi.org/10.1175/1520-0442(2002)015<3266:ACOTEA>2.0.CO;2.
Wu
,
R.
,
Z.
Hu
, and
B. P.
Kirtman
,
2003
:
Evolution of ENSO-related rainfall anomalies in East Asia
.
J. Climate
,
16
,
3742
3758
, https://doi.org/10.1175/1520-0442(2003)016<3742:EOERAI>2.0.CO;2.
Yasui
,
S.
, and
M.
Watanabe
,
2010
:
Forcing processes of the summertime circumglobal teleconnection pattern in a dry AGCM
.
J. Climate
,
23
,
2093
2114
, https://doi.org/10.1175/2009JCLI3323.1.
Yun
,
K.-S.
,
S.-Y.
Kim
,
K.-J.
Ha
, and
M.
Watanabe
,
2011
:
Effects of subseasonal basic state changes on Rossby wave propagation during northern summer
.
J. Geophys. Res.
,
116
,
D24102
, https://doi.org/10.1029/2011JD016258.
Zhu
,
Y. L.
,
H. J.
Wang
,
W.
Zhou
, and
J. H.
Ma
,
2011
:
Recent changes in the summer precipitation pattern in East China and the background circulation
.
Climate Dyn.
,
36
,
1463
1473
, https://doi.org/10.1007/s00382-010-0852-9.
Zhu
,
Y. L.
,
H.
Wang
,
J.
Ma
,
T.
Wang
, and
J.
Sun
,
2015
:
Contribution of the phase transition of Pacific decadal oscillation to the late 1990s’ shift in East China summer rainfall
.
J. Geophys. Res. Atmos.
,
120
,
8817
8827
, https://doi.org/10.1002/2015JD023545.
For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).