The variability of global or regional air temperature is closely related to atmospheric circulation anomalies. The most dominant feature of atmospheric circulation is the occurrence of atmospheric teleconnection patterns (also referred to as oscillation or seesaw patterns), which are made up of chains of lobes whose centers are connected by lines that have prominent meridional components (Branstator, 2002). Teleconnection patterns in the atmosphere have been noted by atmospheric scientists since the early 20th century (Walker and Bliss, 1932), who extensively documented the Southern Oscillation on a global scale. During the 20th century, many teleconnection patterns were identified, such as the North Atlantic Oscillation (Walker and Bliss, 1932), North Pacific Oscillation (Rogers, 1981), eastern Atlantic pattern, western Atlantic pattern, western Pacific pattern, Pacific–North America pattern, Eurasian pattern (Wallace and Gutzler, 1981), and Arctic Oscillation (AO) (Thompson and Wallace, 1998). The relationship between these teleconnection patterns and global climate has been extensively studied (Branstator, 2002; Linkin and Nigam, 2008; Gu et al., 2009).

At the beginning of 21st century, a new teleconnection pattern along the summertime Asian jet was identified, referred to as the Silk Road Pattern (SRP) (Lu et al., 2002; Orsolini et al., 2015). The SRP can be indicated by the correlation of 200-hPa meridional velocity at the base point (42.5°N, 105°E) with that at every grid point in the Northern Hemisphere, which shows an obvious wave-like structure extending from Northwest Africa to East Asia (Lu et al., 2002). It has been revealed that the SRP plays an important role in the variability of extratropical climate. Lu et al. (2002) suggested that the SRP shows a significant correlation with summer precipitation over East Asia and India, suggesting its possible contribution to the linkage between the East Asian summer monsoon and the Indian summer monsoon (Ding and Wang, 2005). The stationary Rossby wave associated with the SRP propagates along the Asian jet and accumulates in the jet exit region, resulting in the formation of the Bonin high, which can affect the path of tropical cyclones and further change the moisture supply to the Meiyu (also called Baiu) frontal zone (Enomoto et al., 2003; Enomoto, 2004). A number of studies have documented that the tripole pattern of the East Asian summer rainfall anomaly, which is characterized by a zonally elongated and meridionally banded structure with signs changing alternately from 20° to 50°N along the East Asian coast, is closely associated with the SRP (Hsu and Lin, 2007; Wang and He, 2015).

Almost all previous studies have focused on the concurrent relationship between summertime climate variables and the SRP. Possibly due to the chaos of the atmosphere, very few scientists have investigated the cross-seasonal influence of the summer SRP. However, some studies have investigated the cross-seasonal influence of other teleconnection patterns, such as the AO. It has been suggested that the anomalous signal of wintertime AO could be maintained by the snow cover on land (Ogi et al., 2004) or sea surface temperature (SST) (Kryjov, 2002), and further affects climate variables in the following months. Choi et al. (2016) even revealed a significant connection between November sea level pressure (SLP) at mid and high latitudes with the preceding wintertime AO. The implication is that the cross-seasonal influence of atmospheric teleconnection patterns is possible, and this motivates us to explore the possible connection between the July SPR and succeeding temperature over the Northern Hemisphere. Considering the highly developed economy and dense population of East Asia, we concentrate on this region in the present study.

This study focuses on filling the research gaps left or highlighted by previous studies. We calculate the correlation between the July SRP and the succeeding (August to the following February) temperature and find that a close connection exists between the July SRP and following January temperature over East Asia. The most intriguing peculiarity of this phenomenon is the gap of six months. It implies that the anomalous signal of the July SRP is maintained by persistent surface anomalies such as SST.

Following this introduction, Section 2 describes the datasets and methods used in the study. The relationships between the July SRP, following January temperature, and atmospheric circulation are studied in Section 3. The possible mechanisms involved in the lagged influence of the July SRP are discussed in Section 4. Conclusions are given in Section 5.

This study is mainly based on the monthly SLP, zonal wind, meridional wind, surface air temperature (SAT), 850-hPa air temperature, and geopotential height at 200 and 500 hPa obtained from ECMWF’s 40-yr Reanalysis (ERA-40) (Uppala et al., 2005). The monthly mean SST, with 1° × 1° horizontal resolution, is from the Hadley Centre (Rayner et al., 2003). In order to confirm the cross-seasonal influence of the July SRP on the January temperature in East Asia, we use another observational high-resolution gridded near-surface temperature dataset from the Climate Research Unit (CRU_TS_3.21), with a resolution 0.5° × 0.5° (Mitchell and Jones, 2005). The SRP index (SRPI) is defined as the first leading principal component of the empirical orthogonal function (EOF) for the 200-hPa meridional wind velocity anomalies in July over the region (30°–60°N, 30°–130°E) (Chen et al., 2013). The first EOF (EOF1) mode explains 33.1% of the total variance. An East Asian temperature index (T_EA) is defined as the area-weighted mean January SAT in the domain (30°–45°N, 110°–150°E) (He and Wang, 2013). To emphasize the interannual variability, we remove the linear trend for the aforementioned data. All datasets cover the period 1958–2002.

To display the spatial distribution of the SRP, Fig. 1 illustrates the regression map of 200-hPa meridional wind anomalies during July 1958–2002 with regard to the simultaneous SRPI. It is clear that the SRP-related meridional wind anomalies at 200 hPa are wave-like and zonally organized, with an alternating appearance of negative and positive anomalies over central Europe, western Asia, central Asia, and central China (Fig. 1, black contours). The associated wave activity flux, which is calculated by using the method of Takaya and Nakamura (2001), i.e., the regressed 200-hPa geopotential height anomalies onto the SRPI, is also shown in Fig. 1 (vectors). The wave activity flux can indicate the propagation of the stationary Rossby wave (Plumb, 1985; He and Wang, 2016). According to the divergence of the wave activity flux, the SRP originates from the entrance of the East Asian jet stream over the Mediterranean and Caspian seas. Meanwhile, it is indicated that the SRP seems to be trapped by the Asian jet stream and propagates from Europe eastward to East Asia along its core region (Fig. 1, red contours), which is regarded as a typical wave guide in boreal summer (Hoskins and Ambrizzi, 1993).

Figure 1 Regression map of July 200-hPa meridional wind anomalies (black contours; m s–1) and wave activity flux (vectors; m2 s–2) associated with the simultaneous July SRPI during 1958–2002. Light and dark shadings denote anomalies significant at the 90% and 95% confidence level, respectively, based on the two-tailed Student’s t test. Red contours indicate the climatology of 200-hPa zonal wind (m s–1) for July 1958–2002.

Figure options

Motivated by the cross-seasonal influence of the November AO (Choi et al., 2016), we calculate the regression map of monthly (August to the following February) mean air temperature anomalies at 850 hPa with regard to the July SRPI. Interestingly, a close connection between air temperature and the July SRPI is only found in the following January. As depicted by Fig. 2a, following a positive anomalous SRPI in July, the January 850-hPa air temperature shows significant negative anomalies over East Asia and significant positive anomalies over western and eastern Siberia (Fig. 2a). This feature can also be observed in the SAT field (Fig. 2b). To confirm such a cross-seasonal connection, we repeat the above analysis using the observational temperature dataset (i.e., CRU_TS_3.21). The spatial distribution of January near-surface temperature anomalies (Fig. 2c) related to the preceding July SPRI is similar to the pattern derived from ERA-40. Overall, Fig. 2 suggests that the connection between the July SRP and following January temperature over East Asia is robust. The interannual variability of the July SPRI during 1958–2001 and January T_EA is displayed in Fig. 3, which shows an obvious out-of-phase variation with a significant negative correlation coefficient of –0.44 (significant at the 95% confidence level). Moreover, it is indicated that roughly two-thirds (29 out of 44) of the years feature the July SPRI and following January T_EA being out-of-phase.

Figure 2 (a) Regression map of January 850-hPa temperature anomalies (contours; m s–1) during 1959–2002 associated with the preceding July SRPI. (b, c) As in (a), but for the ERA-40 SAT and CRU near-surface temperature anomalies, respectively. The light (dark) shading denotes anomalies significant at the 90% (95%) confidence level, based on the two-tailed Student’s t test.

Figure options

Figure 3 Normalized detrended time series of the July SRPI (bars) during 1958–2001 and following January TEA (line). The dots indicate years when the SRPI and TEA are out-of-phase. The correlation coefficient between the SRPI and TEA is given in the top right-hand corner.

Figure options

To better understand the impact of the July SRP on following January temperature over East Asia, we examine the regression map of January atmospheric circulation anomalies corresponding to the July SPRI. Since the atmospheric anomalies associated with positive and negative phases of the SRP largely mirror one another, the following description and discussion focus mainly on the positive phase.

Figure 4 illustrates the January atmospheric circulation anomalies including the SLP, 850-hPa wind, 500-hPa geopotential height, and 300-hPa zonal wind anomalies, which are regressed onto the July SRPI. Corresponding to a positive anomalous July SPRI, in the following January, the SLP shows significant positive anomalies over Siberia and negative ones over North Pacific (Fig. 4a, contours), indicating a strengthening of the Siberian high and Aleutian low. Consistent with the change in SLP, a dominant anomalous surface anticyclone appears over Siberia and an anomalous cyclone occupies North Pacific, leading to significant northerly wind anomalies along coastal East Asia (Fig. 4a, vectors). At 500 hPa (Fig. 4b), significant negative anomalies extend from East Asia to North Pacific where the East Asian trough is usually located. This implies that the January East Asian trough is reinforced when the preceding July SRPI is higher than normal. Besides, there are significant negative anomalies over Europe and significant positive ones over western and central Siberia (Fig. 4b). The spatial distribution of January 500-hPa height anomalies related to the preceding July SRP suggests a Rossby wave pattern propagating eastward throughout Eurasia, which will be discussed in the following analysis. In the upper troposphere, the East Asian jet stream exhibits significant positive anomalies in its core region and negative ones in the north (Fig. 4c), which would contribute to a strengthening of the meridional shear of the East Asian jet stream (Jhun and Lee, 2004; Li and Yang, 2010; Li and Wang, 2012). All the above responses of January atmospheric circulation to the preceding July SRP are consistent with the characteristics of a stronger-than-normal East Asian winter monsoon (EAWM) (He and Wang, 2012a; Li and Wang, 2012, 2013; Wang and He, 2012; He et al., 2013). To offer more evidence in support of this view, we present the linear relationship between the SRPI and four kinds of EAWM indices (Fig. 5). The correlation between the July SRPI and the Siberian high, west–east SLP differences, East Asian trough, and meridional shear of the East Asian jet stream indices are 0.37, 0.31, –0.31, and 0.47, respectively (significant at the 95% confidence level). It is evident that a significant negative relationship exists between the intensity of the July SRPI and the East Asian monsoon in January. In this sense, the January temperature over East Asia generally decreases when the preceding July SRPI is higher than normal, and vice versa.

Figure 4 Regression maps of (a) SLP (contours; hPa) and 850-hPa wind (vectors; m s–1), (b) 500-hPa geopotential height (contours; gpm), and (c) 300-hPa zonal wind (m s–1) anomalies during January 1959–2002 associated with the preceding July SRPI. The light (dark) shading denotes anomalies significant at the 90% (95%) confidence level, based on the two-tailed Student’s t test. The thick black isolines in (a) indicate the climatological position of the Siberian high and Aleutian low.

Figure options

Figure 5 Scatter plots of the July SRPI versus (a) the Siberian high (SH) (Chuan and Wu, 2015), (b) the SLP difference (SLP_Diff) between the domains of 30°–55°N, 100°–120°E and 30°–55°N, 150°–170°E (Chan and Li, 2004), (c) the East Asian trough (EAT) (He and Wang, 2012b), and (d) the meridional shear of the East Asian jet stream (EAJS) (Jhun and Lee, 2004) indices in January 1959–2002, with slopes and correlations.

Figure options

To further investigate the connection between January atmospheric circulation anomalies and the preceding July SRP on the global scale, in Fig. 6, we present the differences in January 500- and 200-hPa geopotential height anomalies (contours), and wave activity flux (vectors) between high and low July SRPI years. After a higher-than-normal SRPI, the 500-hPa height shows significant negative anomalies over North Atlantic, Europe, East Asia, North Pacific, and North America. At the same time, significant positive anomalies appear over west coast of America, and western and eastern Siberia (Fig. 6a, contours). The spatial distribution of 500-hPa height anomalies clearly shows a zonally wave-like pattern on the global scale. By inspecting the wave activity flux, we can see that the Rossby wave propagates eastward with an obvious wave source region in North Pacific (Fig. 6a, vectors). A similar feature is observed in the upper-level troposphere, indicating a quasi-barotropic structure of January atmospheric anomalies related to the preceding July SRP (Fig. 6b).

Figure 6 Composite maps of (a) 500-hPa and (b) 200-hPa geopotential height anomalies (gpm) and wave activity flux (vectors; m2 s–2) in the following January, between high (1973, 1976, 1977, 1979, 1983, 1984, 1985, 1995, and 1997) and low (1959, 1970, 1972, 1991, 1999, and 2000) July SRPI years. Shaded regions indicate height anomalies significant at the 95% confidence level, based on the Student’s t test.

Figure options

Numerous previous studies have documented that stationary Rossby waves in the atmosphere are forced by external forcing, such as the orography and thermal forcing arising from the distribution of land and oceans (Charney and Eliassen, 1949; Hoskins and Karoly, 1981). The dominant wave source over North Pacific (Fig. 6) encourages us to seek out the possible thermal effect of North Pacific SST. Figure 7 illustrates the evolu-tion of SST and 850-hPa wind anomalies from July to the following January associated with the anomalous July SRP. Concurrent with the positive anomalous SRPI, the SST in the central North Pacific shows significant negative anomalies, which are surrounded by significant positive anomalies along the west coast of America (Fig. 7g, contours). Such an anomalous SST pattern might be triggered by the local air–sea interaction (Zhang et al., 2005; Nan et al., 2009; Zhao et al., 2009, 2011, 2016). Climatologically, there is an anticyclone over North Pacific (figure omitted). Southwesterlies prevail in the subtropical North Pacific (around 40°N), which promotes warmer water from the lower latitudes flowing northward. However, during positive anomalous SRPI years, there is an anomalous cyclone over North Pacific (Fig. 7g, vectors), meaning the northward flow of warmer water is weakened, leading to negative SST anomalies in the central subtropical North Pacific. The anomalous southwesterly flow to the west coast of America would favor local positive SST anomalies in situ. This SST anomaly pattern persists well into September (Figs. 7e–g). From October to November, the negative SST anomalies in the central North Pacific move eastward and weaken, confined to east of 160°W (Figs. 7c, d). Interestingly, a re-emergence of the SST anomaly pattern observed in July takes place in December, though with smaller magnitude (Fig. 7b). The SST anomaly in December keeps developing and reaches its maximum in the following January (Fig. 7a, contours). Such SST anomalies in January would strengthen the cyclone over North Pacific (Fig. 7a, vectors), according to thermal wind balance theory. We should bear in mind that the January SST anomalies might be a response to contemporaneous wind anomalies to some extent. It is difficult to identify unequivocally whether the atmosphere is the response or forcing. Nevertheless, dynamical analysis and model simulations have suggested the potential effect of such SST anomalies in January on the simultaneous atmospheric circulation (He and Wang, 2013). The evolution of the North Pacific SST anomalies from July to the following January might be explained by the mechanism of SST anomaly re-emergence given by Namias and Born (1970) and Alexander and Deser (1995), in which it is indicated that ocean temperature anomalies preserved in summer could reappear at the surface in the following winter.

Figure 7 Composite maps of SST (°C) and 850-hPa wind anomalies (vectors) in (a) the following January and the simultaneous, (b) December, (c) November, (d) October, (e) September, (f) August, and (g) July, between high and low July SRPI years. Regions enclosed by white contours indicate SST anomalies significant at the 90% confidence level. Wind anomalies significant at the 90% confidence level are colored blue. The black boxes indicate the key regions of SST anomalies used to define the SST indices.

Figure options

To investigate the possible influence of North Pacific SST anomalies on the atmospheric circulation, we define three SST indices as the area-averaged negative SST anomalies in the domain of 20°–40°N, 140°E–150°W, and positive SST anomalies in the domains of 50°–60°N, 180°–140°W and 15°–30°N, 140°–120°W, referred to as SST1, SST2, and SST3, respectively. Figure 8 shows the 200-hPa meridional wind anomalies and wave activity flux associated with the North Pacific SST anomalies in January. Corresponding to anomalous SST1 and SST2, the alternate appearance of significant positive and nega-tive meridional wind anomalies encircles the midlatitudes (Figs. 8a, b, contours), indicating a global-scale Rossby wave. The wave activity flux depicts the eastward wave propagation well (Figs. 8a, b, vectors). In contrast, the effect of SST3 on the formation of the Rossby wave is not dominant (Fig. 8c). By inspecting the vectors in Figs. 8a, b, we can see a clear wave source over North Pacific, suggesting a potential thermal effect of SST1 and SST2. This is further supported by Fig. 8d, which illustrates well the eastward propagating Rossby wave with an obvious wave source over North Pacific.

Figure 8 Regression maps of 200-hPa meridional wind anomalies (m s–1) and wave activity flux (vectors; m2 s–2) in January with respect to the simultaneous (a) SST1, (b) SST2, (c) SST3, and (d) (SST1+SST2)/2. The light (dark) shading indicates wind anomalies significant at the 90% (95%) confidence level, based on the Student’s t test.

Figure options

Whether the wave source is located in North Pacific is an important indicator for the effect of North Pacific SST anomalies. Therefore, in the next part of the study, we detect the location of the wave source. Considering the theory proposed by Sardeshmukh and Hoskins (1988) that the advection of vorticity by divergent wind can be regarded as a Rossby wave source, we begin by checking the anomalous divergence and convergence associated with the North Pacific SST anomalies. Figure 9a displays the SLP, 1000-hPa divergent wind component, and velocity potential anomalies related to SST1 and SST2. The most notable feature in Fig. 9a is the apparent low SLP anomalies over North Pacific (shaded), which might be induced by the local air–sea interaction (He and Wang, 2013). Naturally, there are significant convergence anomalies over North Pacific (Fig. 9a, vectors), which are consistent with positive velocity potential ano-malies (Fig. 9a, contours). Corresponding to the near-surface anomalies, there are significant negative velocity potential (Fig. 9b, contours) and divergence (Fig. 9b, vectors) anomalies in the upper-tropospheric levels. Consequently, significant wave source anomalies emerge over North Pacific (Fig. 9b, shaded).

Figure 9 Composite anomalies of (a) 1000-hPa velocity potential (contours; 105m2 s–1), divergent wind component (vectors; m s–1), and sea level pressure (shaded; hPa), between high and low (SST1+SST2)/2 years. (b) As in (a), but for the anomalies of 200-hPa velocity potential (contours; 105 m2 s–1), divergent wind component (vectors; m s–1), and Rossby wave source (shaded; 10–10 s–2). The dotted region indicates values significant at the 90% confidence level, based on the two-tailed Student’s t test.

Figure options

In summary, the anomalous July SRP might induce North Pacific SST anomalies through local air–sea interaction in situ, which persist into September, weaken in October and November, and re-emerge in December before intensifying in the following January. The January North Pacific SST anomalies then cause significant low SLP and convergence anomalies in situ, which further leads to significant divergence anomalies in the upper-level troposphere. Due to the advection of vorticity by divergent wind, a stationary Rossby wave originates from North Pacific and propagates eastward, leading to temperature anomalies over East Asia through its influence on the large-scale atmospheric circulation.

Possibly due to the chaotic nature of the atmosphere, very few studies have attempted to explore the cross-seasonal influence of atmospheric teleconnection patterns. Two recent studies conducted by Kim and Ahn (2012) and Choi et al. (2016), which discussed the persistent influence of the AO, implied that a cross-seasonal influence of atmospheric teleconnection pattern is possible.

The present study investigates the non-simultaneous correlation between the July SRP and the monthly mean temperature from August to the following February. The results derived from both ERA-40 and CRU indicate that a significant correlation of temperature is only observed in the following January, when it is dominant in East Asia (Fig. 2). Specifically, following a higher-than-normal SRPI in July, the East Asian January temperature shows significant negative anomalies. The correlation coefficient between the July SRPI and the following January T_EA is –0.44 (significant at the 99% confidence level) during the period 1958/59–2001/02 (Fig. 3). This implies that the preceding July SRP and January T_EA show a frequently out-of-phase occurrence, which is supported by the fact that 29 out of 44 years feature a July SRPI and following T_EA that are out-of-phase (Fig. 3). Linear regression analysis reveals that, following a higher-than-normal SRPI in July, significant positive SLP ano-malies appear over Siberia in the following January, together with significant negative SLP anomalies over North Pacific (Fig. 4a). This means that both the Siberian high and Aleutian low intensify. Meanwhile, the 500-hPa geopotential height shows significant negative anoma-lies over East Asia and positive ones in the Arctic and western Siberia (Fig. 4b), implying a strengthening of the Urals high and East Asian trough. At 300 hPa, the zonal wind displays positive and negative anomaly bands over and to the north of Japan (Fig. 4c). These atmospheric circulation anomalies favor northeasterly wind anoma-lies along the western flank of the Aleutian low and northwesterly wind anomalies along the eastern flank of the Siberian high (Fig. 4a, vectors), leading to more cold air reaching East Asia and causing a lowering of tempera-ture (Fig. 2).

Considering the chaotic nature of atmospheric motion, we speculate that the North Pacific SST anomalies might play a critical role in storing the anomalous signal of July SRP. An anomalous SRP could induce negative SST anomalies in the central North Pacific, surrounded by positive SST anomalies, which could persist into September (Figs. 7e–g). Therein, SST anomalies become weakened in October and November, and with smaller extent (Figs. 7c, d). However, the SST anomaly pattern observed in July re-emerges in December and develops well into the following January (Figs. 7a, b). Such an SST anomaly pattern could induce a surface anomalous cyclone over North Pacific and lead to dominant convergence anomalies (He and Wang, 2013). Correspondingly, significant divergence anomalies appear in the upper-level troposphere (Fig. 9). As the advection of vorticity by divergent wind can be regarded as a Rossby wave, a stationary Rossby wave originates from North Pacific (Fig. 8). This Rossby wave propagates eastward to East Asia, leading to temperature anomalies through its influence on the large-scale atmospheric circulation.

This paper discusses the climatic influence of the SRP, which is depicted by the EOF1 of the 200-hPa meridional wind. However, a more recent study revealed that the second mode (EOF2), of which the centers of action are shifted slightly compared to EOF1, also shows some influence on the East Asian climate (Orsolini et al., 2015). Therefore, the influence of EOF2 might be worthy of further investigation in future work.