Abstract

Interdecadal variations in the relationship between the winter North Atlantic Oscillation (NAO) and surface air temperature in China are investigated using observational and reanalysis data. Focus is on south-central China, in which temperature variability is strongly related to the NAO. It is revealed that the relationship shows clear interdecadal variations in midwinter during 1951–2015. A relatively weak in-phase relationship occurs before the early 1970s (P1), but a significant out-of-phase relationship dominates in the last two decades of the twentieth century (P2), though it is clearly weaker from the late 1990s onward. Observational evidence shows that such interdecadal variations are related mainly to variations in the spatial pattern and amplitude of the NAO. The northern center of the NAO shifted eastward over the second half of the twentieth century. In addition, the amplitude of the center strengthened from P1 to P2, resulting in a perturbation in the atmospheric circulation response pattern over Eurasian mid-to-high latitudes. During P2, the eastward shift and amplitude intensification of the NAO favored a north–south dipole structure in circulation anomalies over the Asian continent, which tended to produce cold temperature anomalies in south-central China during the positive NAO phase and warm anomalies during the negative phase. However, in the past two decades the northern center of the NAO has weakened and retreated westward. This was concurrent with a weakening relationship between the NAO and temperature anomalies in south-central China and northern Eurasia, indicating weaker downstream impacts of the NAO in midwinter.

1. Introduction

The North Atlantic Oscillation (NAO) is the dominant mode of atmospheric low-frequency variability over the North Atlantic–Northern Hemisphere throughout the year (van Loon and Rogers 1978; Wallace and Gutzler 1981; Barnston and Livezey 1987; Hurrell 1995). In the positive (negative) NAO phase, both the Azores high and the Icelandic low are intensified (weakened). Changes in the polarity of the NAO are generally accompanied by profound changes in the weather and climate in the Northern Hemisphere on multiple time scales (e.g., Hurrell et al. 2003; Scaife et al. 2008; Li et al. 2013).

Observational evidence shows that the position of the NAO action centers shows clear interdecadal variations in winter. The northern action center, for example, shifted eastward over the last two decades of the twentieth century compared with the preceding two decades (e.g., Hilmer and Jung 2000; Lu and Greatbatch 2002; Jung et al. 2003; Moore et al. 2013; Raible et al. 2014). Such shifts in the NAO action centers are accompanied by significant changes in the relationship between the NAO and arctic sea ice export, North Atlantic cyclone activity, and surface air temperature and precipitation over Europe in winter (Hilmer and Jung 2000; Jung and Hilmer 2001; Jung et al. 2003; Polyakova et al. 2006; Beranová and Huth 2008; Vicente-Serrano and López-Moreno 2008). For example, the covariability of the winter NAO and surface air temperature anomalies over Eastern Europe strengthened considerably over the last two decades of the twentieth century because of NAO-related intensified zonal flow anomalies over this region (Jung et al. 2003).

Some studies suggest that the NAO tends to meander eastward in its positive phase compared with the negative phase, and therefore the eastward shift of the NAO is likely related to the trend toward an increased frequency in the positive NAO phase in the last quarter of the twentieth century (Peterson et al. 2003; Cassou et al. 2004; Johnson et al. 2008; Woollings et al. 2010). Observational results demonstrate that on the interdecadal time scale the winter NAO index reached its maximum around the mid-1990s and then experienced a downward trend over the past two decades (Zuo et al. 2012; Nakamura et al. 2015). Questions arise as to whether the eastward shift of the winter NAO action center will continue and how this will affect NAO-related climate regimes in the Northern Hemisphere.

The impacts of the NAO on European winter weather and climate are well established (e.g., see Hurrell et al. 2003 for a review). Atmospheric teleconnections may also allow the NAO to exert a significant downstream impact on the East Asian winter climate. Wu and Huang (1999) find that the negative (positive) NAO phase favors a stronger (weaker) East Asian winter monsoon and thus cold (warm) temperature anomalies over East Asia. Hong et al. (2008) report a close decadal relationship between the NAO and cold surge frequencies in northern Taiwan. Tan et al. (2010) demonstrate that the NAO has an out-of-phase relationship with temperature anomalies in central and southern China in January, whereas Xu et al. (2012) reveal an in-phase relationship between the NAO and precipitation anomalies over southwestern China in winter. Some studies suggest that the downstream influences of the NAO on the East Asian winter climate are related to NAO-induced upper-level convergence anomalies over the Mediterranean–Sahara region, which theoretically act to excite a Rossby wave train emanating toward East Asia along the subtropical jet stream (Branstator 2002; Watanabe 2004; Hong et al. 2008).

Recently, Zuo et al. (2015) report differential impacts of the Arctic Oscillation (AO), which has a spatial pattern in sea level pressure anomalies over the North Atlantic sector highly resembling the NAO pattern, on the temperature anomalies in southern China between early and mid-to-late winter during 1979–2011. This is also true for the relationship between the NAO and temperature anomalies in southern China during the same period (Fig. 1); that is, a weak in-phase relationship occurs in December, while a significant (weak) out-of-phase relationship occurs in January (February). Such a clear contrast in the relationships between early and mid-to-late winter are primarily attributed to the large intraseasonal zonal migrations of the AO–NAO Azores center (Zuo et al. 2015). This implies that the east–west movements of the NAO action centers may modulate the downstream impact of the NAO on the East Asian winter climate. In other words, observed interdecadal variations in the zonal positions of the NAO action centers are likely accompanied by a varying relationship between the winter NAO and temperature anomalies in China, though more work is needed to further define this relationship.

Fig. 1.

Correlations between the NAO index and surface air temperature anomalies over China in (a) DJF, (b) December, (c) January, and (d) February during 1979–2011. The inset in the lower right of each panel is the South China Sea map. Black (white) dots denote significance at the 90% (95%) confidence level. [The red box in (c) covers the region used to define the CSCT index in Fig. 2.]

Fig. 1.

Correlations between the NAO index and surface air temperature anomalies over China in (a) DJF, (b) December, (c) January, and (d) February during 1979–2011. The inset in the lower right of each panel is the South China Sea map. Black (white) dots denote significance at the 90% (95%) confidence level. [The red box in (c) covers the region used to define the CSCT index in Fig. 2.]

In this study, we investigate interdecadal variations in the relationship between the winter NAO and temperature anomalies in China and further explore the possible mechanisms for such interdecadal variations. We particularly focus on recent interdecadal changes in the winter NAO pattern and its relationship with temperature anomalies over Eurasia since the late 1990s, which may help with predictions of future changes. The remainder of the paper is organized as follows. Section 2 introduces the data and methods used in the study, and section 3 presents the relationship between the winter NAO and temperature anomalies in China as well as its interdecadal variations since the 1950s. Section 4 explores the causes of the interdecadal variations in the relationship between the midwinter NAO and temperature anomalies in south-central China, and section 5 focuses on the recent interdecadal changes in the midwinter NAO pattern and associated changes in Eurasian temperature anomalies. Section 6 provides a summary and discussion.

2. Data and methods

This study employs monthly global atmospheric circulation data, including geopotential height, air temperature, zonal and meridional wind components, and sea level pressure (SLP), from the National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis (Kalnay et al. 1996). These data have a horizontal resolution of 2.5° latitude × 2.5° longitude and are available from 1948 to the present. In addition, we utilize gauge-based observations of monthly surface air temperature at 160 stations in China provided by the National Climate Center of the China Meteorological Administration since 1951 (National Climate Center 2015). Linear trends in the temperature time series were removed prior to analysis.

The monthly NAO index is obtained from the National Oceanic and Atmospheric Administration (NOAA) Climate Prediction Center (CPC; NOAA/Climate Prediction Center 2015). This index is constructed using a rotated principal component analysis (Barnston and Livezey 1987) with monthly mean standardized 500-hPa height anomalies over the extratropical Northern Hemisphere (20°–90°N). To focus on the interannual relationship between the NAO and temperature anomalies, the decadal component of the NAO index is simply eliminated by subtracting the 9-yr moving mean from the raw index. Following Zuo et al. (2015), the intensity of the Middle East jet stream (MEJS) is calculated from the regionally averaged zonal wind at 300 hPa over subtropical western Asia (25°–35°N, 50°–80°E; see box in Fig. 4e). The statistical significance of regression and correlation coefficients is assessed using a two-tailed Student’s t test, and the difference between two correlation coefficients is assessed using the Fisher z transformation (Fisher 1915). The effective number of degrees of freedom Ne is defined following Davis (1976).

3. Variations in the relationship between the winter NAO and temperature in China

Figure 1 shows the spatial distribution of correlations between the NAO index and temperature anomalies in China for the December, January, and February total mean (DJF) and December, January, and February, respectively, from 1979 to 2011. Correlation patterns in DJF resemble those in January and February, with negative correlations observed in south-central China and positive correlations in northern China (Figs. 1a,c,d). Note that correlations in south-central China are much higher for January compared with DJF and February. However, in December, correlations between the NAO index and temperature anomalies are weakly positive in China, except for in northwestern China and part of southwestern China (Fig. 1b). The difference in correlation coefficients between early and mid-to-late winter is statistically significant over most of south-central China (figures not shown). These results indicate a contrast in the relationship between the NAO and temperature anomalies in south-central China between early and mid-to-late winter, which is consistent with results associated with the AO (see Fig. 1 in Zuo et al. 2015). Such a sharp contrast leads to weak correlations between the DJF mean NAO index and temperature anomalies over most of China. Therefore, to better investigate interdecadal variations in the NAO–temperature relationship and the associated causes of this variability, this study focuses on January, when the NAO has the highest correlation with temperature anomalies in south-central China during 1979–2011.

Figure 2 shows sliding correlations between the NAO index and regionally averaged temperature anomalies over south-central China (denoted CSCT; 22°–40°N, 100°–115°E see box in Fig. 1c), with a 21-yr moving window in January during 1951–2015. The NAO–CSCT correlation is positive before the early 1970s but negative afterward, and the value of negative correlation reaches a maximum around the late 1980s. Additionally, the NAO–CSCT relationship weakens evidently after the late 1990s. These results reveal clear interdecadal variations in the relationship between the midwinter NAO and temperature anomalies in south-central China in recent decades.

Fig. 2.

Sliding correlations between the NAO and CSCT indices displaced at the central year of the 21-yr window in January during 1951–2015. Dotted (dashed) lines indicate significance at the 90% (95%) confidence level.

Fig. 2.

Sliding correlations between the NAO and CSCT indices displaced at the central year of the 21-yr window in January during 1951–2015. Dotted (dashed) lines indicate significance at the 90% (95%) confidence level.

Note that, in Fig. 2, interdecadal variations in NAO–CSCT correlations are primarily characterized by a sharp contrast between 1952–72 (referred to as P1) and 1978–98 (referred to as P2). The correlation coefficient R is about 0.31 in P1 but −0.61 in P2, and Ne has a value of 20 for both periods. Therefore, the correlation in P2 is statistically significant at the 95% confidence level, whereas the one in P1 is close to the threshold of R = 0.36 for the 90% confidence level. The difference in correlation coefficients between the two periods is significant at the 95% confidence level, confirming a sharp contrast in NAO–CSCT correlations between P1 and P2. As shown in Figs. 3a,b, there is a clear contrast in spatial correlation patterns between the NAO index and temperature anomalies in China for January during P1 and P2. Significant positive correlations are observed in southern China during P1 (Fig. 3a); however, the correlations become negative and stronger in south-central China during P2 (Fig. 3b). Additionally, for 2000–15 (referred to as P3), correlations between the NAO and temperature anomalies become fairly weak over most of China (Fig. 3c). These results further validate our conclusion regarding the varying relationship between the midwinter NAO and temperature anomalies in south-central China during 1951–2015.

Fig. 3.

Correlations between the NAO index and surface air temperature anomalies over China in January for the periods of (a) 1952–72, (b) 1978–98, and (c) 2000–15. Black and white dots are as in Fig. 1.

Fig. 3.

Correlations between the NAO index and surface air temperature anomalies over China in January for the periods of (a) 1952–72, (b) 1978–98, and (c) 2000–15. Black and white dots are as in Fig. 1.

4. Plausible causes of the varying relationship

To detect the plausible causes of the varying relationship between the midwinter NAO and temperature anomalies in south-central China, we examine atmospheric circulation anomalies during P1 and P2, when there were large variations in the NAO–CSCT relationship. We also investigate P3 to advance our knowledge of the recent weakening of the NAO–CSCT relationship.

a. General characteristics of atmospheric circulation anomalies

Figure 4 displays geopotential height anomalies at 500 hPa (H500) and zonal wind anomalies at 300 hPa regressed against the January CSCT index and multiplied by −1 for 1951–2015 and the three subperiods, P1, P2, and P3. The common feature in CSCT-related H500 anomalies is a north–south dipole over mid- and high latitudes in Asia, although there are some differences in the amplitude and center positions of the dipole among these periods (Figs. 4a–d). When the CSCT index is lower than normal, the H500 anomalies are above normal over Siberia and below normal to the south, which favors the southward invasion of continental polar air masses into south-central China. Moreover, westerly anomalies at 300 hPa tend to intensify over the Middle East for all the time periods (Figs. 4e–h), indicating an intensified MEJS. Therefore, an intensified MEJS and the dipole-like geopotential height anomalies over the Asian mid-to-high latitudes produce favorable conditions for the formation of midwinter cold temperature anomalies in south-central China and with the reverse also true.

Fig. 4.

Regressions of geopotential height anomalies at 500 hPa (gpm) against the CSCT index multiplied by −1 in January for the periods (a) 1951–2015, (b) 1952–72, (c) 1978–98, and (d) 2000–15. Orange and green shading denotes significance at the 90% confidence level. [The boxes labeled A and B in (a) are used to define the AD index in Fig. 6b.] (e)–(h) As in (a)–(d), but for the zonal wind anomalies at 300 hPa (m s−1), where the box in (e) is used to define the MEJS index in Fig. 6a. Purple contours in (e)–(h) denote the climatic mean of the 300-hPa zonal wind with a value of 30 m s−1, and the red line refers to the longitude of 65°E.

Fig. 4.

Regressions of geopotential height anomalies at 500 hPa (gpm) against the CSCT index multiplied by −1 in January for the periods (a) 1951–2015, (b) 1952–72, (c) 1978–98, and (d) 2000–15. Orange and green shading denotes significance at the 90% confidence level. [The boxes labeled A and B in (a) are used to define the AD index in Fig. 6b.] (e)–(h) As in (a)–(d), but for the zonal wind anomalies at 300 hPa (m s−1), where the box in (e) is used to define the MEJS index in Fig. 6a. Purple contours in (e)–(h) denote the climatic mean of the 300-hPa zonal wind with a value of 30 m s−1, and the red line refers to the longitude of 65°E.

However, the patterns of CSCT-related H500 anomalies over the North Atlantic are obviously different for P1, P2, and P3 (Figs. 4b–d). Does this mean that the NAO pattern, as the contributor to winter temperature anomalies in China, has changed among the different periods? In Fig. 5, although the H500 anomalies regressed against the NAO index are generally characterized by a north–south dipole over the North Atlantic in January for all three subperiods, the center positions and amplitude of the NAO dipole structure as well as the pattern of downstream anomalies over Eurasia show evident differences among P1, P2, and P3. This implies that the varying relationship between the midwinter NAO and temperature anomalies in south-central China is likely related to interdecadal variations in the patterns of NAO-associated circulation anomalies over the Atlantic–Eurasian region. It should be noted that correlations do not necessarily imply causality. Such a varying relationship between the NAO and temperature anomalies in south-central China could be tied to a third independent factor affecting both the NAO and China climate, such as ENSO (Folland et al. 2012; Zhang et al. 2015), arctic sea ice (Nakamura et al. 2015; Zuo et al. 2016), and Eurasian snow cover (Cohen et al. 2012), which needs to be investigated in further studies.

Fig. 5.

As in Fig. 4, but for the patterns regressed against the NAO index. Horizontal red line refers to the latitude of 65°N, and vertical red line refers to the longitude of 40°W in (a)–(d) and 65°E in (e)–(h).

Fig. 5.

As in Fig. 4, but for the patterns regressed against the NAO index. Horizontal red line refers to the latitude of 65°N, and vertical red line refers to the longitude of 40°W in (a)–(d) and 65°E in (e)–(h).

b. Roles of the MEJS and Asian dipole

Zuo et al. (2015) demonstrate that the positive (negative) phase of the AO–NAO potentially yields the intensified (weakened) MEJS, which plays a role in linking the AO–NAO and temperature anomalies in southern China during winter. As indicated in Figs. 5f–h, NAO-related 300-hPa zonal wind anomaly patterns over the Middle East in January are similar for P1, P2, and P3. In particular, significant westerly anomalies are observed over the Middle East in the positive NAO phase, indicating an intensified MEJS. According to the previous study, the intensified MEJS should favor cold temperature anomalies in south-central China (see Fig. 4 in Zuo et al. 2015), but during P1, warm temperature anomalies are actually observed over this region in the positive NAO phase (Fig. 3a). Moreover, examination of 21-yr sliding correlations for January shows that the NAO index has a nearly stable in-phase relationship with the MEJS index during 1951–2015, aside from a slight weakening of the correlation occurring after the mid-to-late 1990s (Fig. 6a). Therefore, these results suggest that interdecadal variations in the MEJS appear to contribute little to the varying relationship between the NAO and temperature anomalies in south-central China in midwinter. An alternative pathway likely exists for the linkage between the NAO and temperature anomalies in south-central China, which requires further examination.

Fig. 6.

As in Fig. 2, but for the correlations between the NAO and (a) MEJS and (b) AD indices.

Fig. 6.

As in Fig. 2, but for the correlations between the NAO and (a) MEJS and (b) AD indices.

It can be seen in Figs. 5b,c that the NAO-associated H500 anomalies show obviously different patterns in P1 and P2 over Eurasian mid-to-high latitudes in January. During P1, a wave train–like pattern is observed downstream of the northern center of the NAO, concurrent with above (below) normal H500 anomalies over the Barents and Kara Seas (northeastern Asia) in the positive NAO phase. Also, above-normal H500 anomalies occur over eastern China and Japan, which yields a slight weakening of the East Asian trough. Consequently, southerly surface wind anomalies are observed over south-central China (not shown), which is consistent with warm temperature anomalies over this region (Fig. 3a). During P2, the NAO tends to be accompanied by a north–south dipole in H500 anomalies over Eurasian mid-to-high latitudes (Fig. 5c). This dipole is similar to the pattern of CSCT-related H500 anomalies over this region, but with an opposite sign. It therefore yields a significant and out-of-phase relationship between the NAO and temperature anomalies in south-central China during P2.

These results reveal that atmospheric circulation anomalies associated with the midwinter NAO exhibit different patterns over Eurasian mid-to-high latitudes between P1 and P2, leading to an interdecadal change in the relationship between the NAO and temperature anomalies in south-central China. To further confirm this finding, we examine interdecadal variations in the relationship between the midwinter NAO and circulation anomaly patterns over Eurasia in past decades. Focus is on the aforementioned north–south dipole in the H500 anomalies over Asia, owing to its important role in controlling temperature variations in south-central China.

An Asian dipole (AD) index was constructed from the difference between regionally averaged normalized H500 anomalies between Siberia (57.5°–72.5°N, 65°–115°E; see box A in Fig. 4a) and the southeast Asian continent (30°–40°N, 70°–110°E; see box B in Fig. 4a). Then, 21-yr sliding correlations between the NAO and AD indices are calculated for January during 1951–2015 (see Fig. 6b). It is found that the NAO index is closely correlated with the AD index, but correlations are characterized by clear interdecadal variations in past decades. A comparison of Fig. 6b with Fig. 2 indicates that interdecadal variations in the NAO–AD correlations are clearly out of phase with the NAO–CSCT correlations. The former features an upward trend before the late 1980s, which is concurrent with a downward trend in the latter; however, both show weakened correlations since the late 1990s.

Therefore, these results suggest that the varying relationship between the midwinter NAO and temperature anomalies in south-central China is primarily attributed to interdecadal variations in NAO-related circulation anomaly patterns over Eurasian mid-to-high latitudes in recent decades. Then, the question arises as to why NAO-related circulation patterns feature clear interdecadal changes.

c. Interdecadal variations in NAO-related circulation anomalies

Figures 5b,c shows that the NAO pattern in H500 anomalies during P2 is marked by a larger zonal extent and stronger amplitude relative to that during P1. Additionally, the northern center of the NAO is mainly located on the west coast of Greenland during P1 but shifts eastward into Iceland during P2. These features are also clearly observed in the NAO-related SLP anomalies (not shown). This indicates a significant eastward shift and stronger amplitude of the northern center of the NAO from P1 to P2, in agreement with previous studies (e.g., Jung et al. 2003).

During P2, the northern center of the NAO appears to favor one branch of anomalous wave activity flux (Takaya and Nakamura 2001) that propagates eastward into northern Eurasia (Fig. 7b), concurrent with a wave train oriented along the subpolar waveguide (Hoskins and Ambrizzi 1993, their Fig. 13). Also, there exists another branch of wave trains propagating from the North Atlantic southeastward into Europe and the Arabian Gulf (Fig. 7b), which are preferred propagation regions for Rossby waves (Hoskins and Ambrizzi 1993). The northern wave train is accompanied by above-normal geopotential height anomalies over northern Asia, and the southern wave train is concurrent with above-normal geopotential height anomalies over the southern Arabian Peninsula and northern Arabian Sea in the positive phase of the NAO (Fig. 7b). Consequently, below-normal geopotential height anomalies occur over northwestern China, forming a north–south dipole in circulation anomalies over the Asian continent that well resembles the CSCT-related circulation anomaly pattern.

Fig. 7.

Regressions of geopotential height anomalies (shading; gpm) and associated wave activity flux (vectors; m2 s−2) at 300 hPa against the NAO index in January for the periods (a) 1952–72, (b) 1978–98, and (c) 2000–15. Regions inside the gray (white) contours denote significance at the 90% (95%) confidence level. Horizontal red line refers to the latitude of 65°N, and vertical red line refers to the longitude of 20°W.

Fig. 7.

Regressions of geopotential height anomalies (shading; gpm) and associated wave activity flux (vectors; m2 s−2) at 300 hPa against the NAO index in January for the periods (a) 1952–72, (b) 1978–98, and (c) 2000–15. Regions inside the gray (white) contours denote significance at the 90% (95%) confidence level. Horizontal red line refers to the latitude of 65°N, and vertical red line refers to the longitude of 20°W.

During P1, however, the northern center of the NAO tends to be accompanied by wave activity flux anomalies propagating northeastward into the Barents and Kara Seas and then southeastward to northeastern Asia, resulting in a wave train across Eurasian high latitudes (Fig. 7a). It can be estimated roughly from Figs. 7a,b that the northern center of the NAO and the associated downstream pattern over northern Eurasia both shift eastward by about 30° in longitude from P1 to P2. During P1, a north–south dipole in geopotential height anomalies occurs over the Barents and Kara Seas as well as the Caspian Sea, which is located far away from East Asia and therefore consistent with a relatively weak NAO–CSCT relationship.

These results indicate that interdecadal variations in the pattern of NAO-related circulation anomalies over Eurasian mid-to-high latitudes are likely the result of interdecadal variations in the zonal positions and intensity of the northern center of the NAO in midwinter. Figure 8a shows a time–longitude section of the NAO-related H500 anomalies averaged over the northern center (60°–70°N) for January, providing more evidence for the aforementioned relationship. The northern center of the NAO shifts eastward significantly from the 1950s to the mid-to-late 1990s, concurrent with an enhancement in the amplitude of the center (Fig. 8a). After the late 1990s, however, the northern center of the NAO retreats westward and its amplitude is significantly weakened relative to the preceding two decades. Corresponding interdecadal variations are also observed in downstream circulation anomalies over Eurasian mid-to-high latitudes (Figs. 8a,b), consistent with interdecadal changes in the NAO–AD relationship (Fig. 6b). Therefore, these results confirm that interdecadal variations in the spatial structure and intensity of the NAO are associated with significantly different circulation anomaly patterns over Eurasian mid-to-high latitudes.

Fig. 8.

Evolutions of 500-hPa geopotential height anomalies (gpm) averaged over (a) 60°–70°N and (b) 35°–40°N and regressed against the NAO index with a 21-yr moving window in January during 1951–2015. Gray and white contours are as in Fig. 7, and green line roughly refers to the maximum (minimum) of the regression coefficients in (a).

Fig. 8.

Evolutions of 500-hPa geopotential height anomalies (gpm) averaged over (a) 60°–70°N and (b) 35°–40°N and regressed against the NAO index with a 21-yr moving window in January during 1951–2015. Gray and white contours are as in Fig. 7, and green line roughly refers to the maximum (minimum) of the regression coefficients in (a).

In brief, the zonal position and amplitude of the northern center of the midwinter NAO are characterized by clear interdecadal variations during 1951–2015, leading to significant changes in the circulation anomalies over Eurasian mid-to-high latitudes, which play an important role in controlling temperature variations in south-central China. As a result, the relationship between the NAO and south-central China temperature anomalies also shows interdecadal variations in past decades. More evidence validating the dependence of the NAO–CSCT relationship upon the zonal position and amplitude of the NAO pattern is provided in the following subsection.

d. Dependence on the position and amplitude of the NAO pattern

Figure 9 examines the dependence of the NAO–CSCT relationship on the zonal position and amplitude of the northern center of the NAO in January. Here, the NAO pattern is represented using the H500 anomalies regressed against the NAO index for each 21-yr moving window, and its amplitude is described as the maximum absolute value of regression coefficient at the center.

Fig. 9.

Scatterplot of (a) zonal position (°E) and (b) amplitude (10 gpm) of the northern center of the NAO vs sliding correlation coefficients between the NAO and CSCT indices with a 21-yr moving window in January during 1951–2015. Solid red lines denote linear fitting of the dots, R is the correlation, and N is the total number of samples for fitting.

Fig. 9.

Scatterplot of (a) zonal position (°E) and (b) amplitude (10 gpm) of the northern center of the NAO vs sliding correlation coefficients between the NAO and CSCT indices with a 21-yr moving window in January during 1951–2015. Solid red lines denote linear fitting of the dots, R is the correlation, and N is the total number of samples for fitting.

It is indicated that the 21-yr sliding NAO–CSCT correlation is highly correlated with the zonal position of the northern center of the NAO, with a correlation coefficient of −0.87 for 45 samples during 1951–2015 (Fig. 9a), which is significant at the 99% confidence level (Ne = 6). Moreover, it is noteworthy that the NAO index tends to have an out-of-phase relationship with the CSCT index when the northern center of the NAO is located east of about 40°W. A closely out-of-phase relationship also exists between the 21-yr sliding NAO–CSCT correlation and the amplitude of the northern center of the NAO (Fig. 9b). However, such a relationship is somewhat asymmetric; that is, the correlation coefficient between these two is only −0.22 for all 45 samples (Ne = 22) but as high as −0.83 when the sliding NAO–CSCT correlation is negative (33 samples and Ne = 5). The former is below significance at the 90% confidence level, while the latter exceeds significance at the 95% confidence level. The difference between the two correlation coefficients is significant at the 90% confidence level.

Since the NAO–CSCT relationship is highly dependent on both the zonal position and amplitude of the northern center of the NAO, we illustrate such a dependence in the scatterplot in Fig. 10. It is shown that an asymmetric relationship exists between the zonal position and amplitude of the northern center of the NAO in January during 1951–2015. When the center is located east of about 40°W, the amplitude is highly correlated with the zonal position of the center, with a high correlation coefficient of about 0.77 (29 samples). In this case, the NAO index tends to be more out of phase with the CSCT index as its northern center shifts farther eastward and its amplitude becomes stronger. When the northern center of the NAO is located west of about 40°W, however, there is no statistically significant relationship between the amplitude and zonal position of the center.

Fig. 10.

Scatterplot of zonal position (x axis; °E) vs amplitude (y axis; 10 gpm) of the northern center of the NAO with a 21-yr moving window in January during 1951–2015. Shaded dots indicate the 21-yr sliding correlation coefficient between the NAO and CSCT indices. Solid red line denotes linear fitting of the dots, R is the correlation, and N is the total number of samples for fitting. Fitting is only applied to the data with a zonal position eastward of 320°E.

Fig. 10.

Scatterplot of zonal position (x axis; °E) vs amplitude (y axis; 10 gpm) of the northern center of the NAO with a 21-yr moving window in January during 1951–2015. Shaded dots indicate the 21-yr sliding correlation coefficient between the NAO and CSCT indices. Solid red line denotes linear fitting of the dots, R is the correlation, and N is the total number of samples for fitting. Fitting is only applied to the data with a zonal position eastward of 320°E.

We also examine interdecadal variations in the zonal position and amplitude of the southern center of the NAO and its impacts on the NAO–CSCT relationship in midwinter. The 21-yr sliding NAO–CSCT correlations appear to be in phase with the zonal position of the southern center in January during 1951–2015 (Fig. 11a) but relatively weaker than that of the northern center. In addition, the relationship between the 21-yr sliding NAO–CSCT correlation and amplitude of the southern center of the NAO is fairly weak in January during 1951–2015 (Fig. 11b).

Fig. 11.

As in Fig. 9, but for the southern center of the NAO.

Fig. 11.

As in Fig. 9, but for the southern center of the NAO.

Model results from Peterson et al. (2003) reveal that the spatial pattern of the NAO depends nonlinearly on the NAO index, the pattern being shifted to the east (west) for high (low) NAO index. As seen in Fig. 12a, the correlation coefficient between the 21-yr running-averaged NAO index and zonal position of the northern center of the NAO is as high as 0.89 (Ne = 5), which is significant at the 99% confidence level. This further supports the results shown in Peterson et al. (2003). Since our results have shown that the NAO–CSCT relationship is highly dependent on both the zonal position and amplitude of the northern center of the NAO, it would be expected to see a link between the NAO–CSCT relationship and the NAO index. To support this, Fig. 12b shows the scatterplot of 21-yr running-averaged NAO index versus sliding NAO–CSCT correlation coefficient in January during 1951–2015. It is indicated that the sliding NAO–CSCT correlation is highly related to the NAO index, with a correlation coefficient of −0.85 that is significant at the 99% confidence level (Ne = 6).

Fig. 12.

Scatterplot of 21-yr running average of the NAO index vs (a) zonal position (°E) of the northern center of the NAO and (b) 21-yr sliding NAO–CSCT correlation coefficients in January during 1951–2015. Solid red lines denote linear fitting of the dots, R is the correlation, and N is the total number of samples for fitting.

Fig. 12.

Scatterplot of 21-yr running average of the NAO index vs (a) zonal position (°E) of the northern center of the NAO and (b) 21-yr sliding NAO–CSCT correlation coefficients in January during 1951–2015. Solid red lines denote linear fitting of the dots, R is the correlation, and N is the total number of samples for fitting.

Therefore, these results confirm that the relationship between the midwinter NAO and temperature anomalies in south-central China strongly depends on the zonal position and amplitude of the northern center of the NAO, supporting our conclusion regarding the plausible cause of the varying relationship between the NAO and south-central China temperature anomalies in past decades. Moreover, because of the dependence of the spatial pattern of the NAO on the NAO index (Peterson et al. 2003), a significant and linear relationship is observed between NAO–CSCT correlation and the NAO index in midwinter.

e. Westward retreat of the NAO pattern since the late 1990s

As indicated in Figs. 2 and 3, the out-of-phase relationship between the midwinter NAO and temperature anomalies in south-central China weakens from P2 to P3. This is concurrent with the westward retreat and weaker amplitude of the northern center of the NAO from P2 to P3 (Figs. 5c,d and Fig. 8a). It is also concurrent with significant changes in the pattern of NAO-related H500 anomalies over Eurasian mid-to-high latitudes (Figs. 8a,b). During P3, the negative (positive) H500 anomalies over the Asian mid- (high) latitudes shift westward compared to those during P2 in the positive phase of the NAO (Figs. 7b,c). This indicates that the NAO-associated dipole-like H500 anomalies over the Asian mid-to-high latitudes move far away from East Asia, which does not favor the southward invasion of high-latitude cold air masses into China during P3. Consequently, the relationship between the NAO and temperature anomalies becomes fairly weak over most of China during P3. Consistent evidence, obtained from Figs. 2 and 6b, shows that the correlations of the NAO index with both the CSCT and AD indices weaken starting in the late 1990s.

Moreover, the NAO-associated zonal flow anomalies across Europe and northern Asia are clearly weakened from P2 to P3, which yields a weakened relationship between the NAO and temperature anomalies over these regions in January (Figs. 13b,c). This indicates that the recent westward retreat and weaker amplitude of the northern center of the NAO are also accompanied by significant changes in the surface flow and air temperature anomalies over Europe and northern Asia in midwinter.

Fig. 13.

As in Fig. 3, but using surface air temperature anomalies (shading) derived from the NCEP–NCAR reanalysis. Vectors indicate regressions of surface wind anomalies with significance at the 90% confidence level (m s−1). Gray and white contours are as in Fig. 7.

Fig. 13.

As in Fig. 3, but using surface air temperature anomalies (shading) derived from the NCEP–NCAR reanalysis. Vectors indicate regressions of surface wind anomalies with significance at the 90% confidence level (m s−1). Gray and white contours are as in Fig. 7.

Previous studies demonstrate that the NAO tends to shift eastward in its positive phase relative to the negative phase (e.g., Peterson et al. 2003; Cassou et al. 2004). They suggested that the eastward shift of the NAO action centers is a consequence of the upward trend in the winter NAO index in the last quarter of the twentieth century. As seen in Fig. 14, the January NAO index appears to experience a downward trend toward its neutral phase in the past two decades, which is consistent with the recent westward retreat of the northern center of the NAO. This supports our conclusions regarding interdecadal variations in the relationship of the NAO with the atmospheric circulation and surface air temperature anomalies over Eurasia in midwinter.

Fig. 14.

Time series of the NAO index (gray line) for January during 1951–2015. Black line denotes the 9-yr running average.

Fig. 14.

Time series of the NAO index (gray line) for January during 1951–2015. Black line denotes the 9-yr running average.

5. Summary and discussion

The NAO, as the dominant mode of atmospheric low-frequency variability over the North Atlantic, profoundly influences weather and climate over surrounding continents. The NAO also has a close relationship with East Asian climate variations and particularly with surface air temperature anomalies in south-central China in winter. However, the present study reveals that the relationship between the midwinter NAO and temperature anomalies in south-central China features a clear interdecadal variation during 1951–2015. There is an in-phase relationship before the early 1970s but out-of-phase relationship afterward. A particularly sharp contrast in the relationship has been found between the time periods 1952–72 (P1) and 1978–98 (P2). Also of note is that the relationship between the midwinter NAO and temperature anomalies in south-central China have weakened since the late 1990s, compared with the preceding two decades.

Observational evidence, derived from the NCEP–NCAR reanalysis dataset, suggests that a changing relationship between the NAO and temperature anomalies in south-central China is primarily attributed to interdecadal variations in the zonal position and amplitude of the northern center of the NAO in midwinter. The northern center of the NAO shifts farther eastward and its amplitude becomes stronger from P1 to P2, leading to distinctly different NAO-associated circulation anomaly patterns over Eurasian mid-to-high latitudes. The northern center of the NAO is located to the west of Iceland during P1, accompanying a wave train propagating eastward into the Barents and Kara Seas and eastern Asia. In this case, the East Asian trough appears to be weakened in the positive NAO phase, leading to southerly surface flow anomalies and thus warm temperature anomalies over south-central China, which yields an in-phase relationship between the NAO and temperature anomalies over this region. During P2, however, the northern center of the NAO shifts eastward into Iceland, accompanying a meridional dipole of geopotential height anomalies over Asia that is favorable for the cold (warm) temperature anomalies over south-central China in the positive (negative) phase of the NAO. Consequently, an out-of-phase relationship is observed between the NAO and temperature anomalies in south-central China during P2.

Further, we showed that the relationship between the midwinter NAO and temperature anomalies in south-central China strongly depends on the zonal position and amplitude of the northern center of the NAO. The 21-yr sliding correlation between the NAO and regionally averaged temperature anomalies over south-central China is linearly related to the zonal position of the northern center of the NAO during 1951–2015. However, the relationship between the 21-yr sliding NAO–temperature correlation and the amplitude of the northern center of the NAO is somewhat asymmetric, which is likely attributed to the asymmetry of the relationship between the zonal position and amplitude of the center. When the northern center of the NAO in 500-hPa geopotential height anomalies is located east of about 40°W, a closely in-phase relationship exists between the zonal position and amplitude of the center. In this case, the NAO tends to have a stronger out-of-phase relationship with south-central China temperature anomalies as the northern center of the former shifts farther eastward and its amplitude becomes stronger. In contrast, when the northern center of the NAO is located west of about 40°W, there is no statistically significant relationship between the amplitude and zonal position of the center. In this situation, the amplitude of the northern center of the NAO has a relatively weak relationship with sliding correlations between the NAO and south-central China temperature anomalies, indicating the dominant role of the zonal position in controlling the NAO’s impact on the temperature anomalies over south-central China in midwinter. Additionally, because of the dependence of the spatial pattern of the NAO on the NAO index (Peterson et al. 2003; Fig. 12a), correlation between the midwinter NAO and temperature anomalies in south-central China is linearly tied to the NAO index.

Moreover, we found that the northern center of the NAO clearly retreats westward beginning in the late 1990s, concurrent with a weakening of both its amplitude and relationship with the south-central China temperature anomalies in midwinter. Also, the recent westward retreat and weaker amplitude of the northern center of the NAO are accompanied by significant changes in surface flow and air temperature anomalies over Europe and northern Asia. It is well known that the wintertime NAO index trends toward its positive extreme phase from the 1970s to the mid-1990s and then enters its neutral phase in recent decades. Previous studies suggest that the NAO tends to move farther east in its positive phase than its negative phase (Peterson et al. 2003; Cassou et al. 2004; Johnson et al. 2008; Woollings et al. 2010). In terms of this notion, interdecadal variations in the zonal position of the NAO pattern are highly consistent with those in the NAO index during winter in past decades.

We reexamine our results with the monthly products of atmospheric circulation and air temperature from the European Center for Medium-Range Weather Forecasts (ECMWF), including ERA-40 for the period of 1958–98 (Uppala et al. 2005) and ERA-Interim for the period of 1979–2015 (Dee et al. 2011). It is found that the results derived from the ECMWF reanalysis are quite consistent with those derived from the NCEP–NCAR reanalysis (not shown). Therefore, our conclusions regarding the interdecadal variations in the NAO pattern and associated changes in atmospheric circulation anomalies and surface air temperature over Eurasia are not dependent on the reanalysis datasets we used.

Though the NAO is an internal mode of atmospheric low-frequency variability (e.g., Ren et al. 2009, 2012; Tan et al. 2014), it may be modulated by external forcing (Bader et al. 2011). Some model studies suggested that the upward trend in the winter NAO index over the second half of the twentieth century is closely tied to the local and/or remote forcing of ocean surface conditions (Rodwell et al. 1999; Hoerling et al. 2004). But some other studies argued that most models generally show much weaker NAO changes in response to observed ocean forcing and do not reproduce the magnitude of the observed NAO trend (Schneider et al. 2003; Scaife et al. 2009). Many previous studies have demonstrated that below-normal sea ice in the Arctic (Budikova 2009; Vihma 2014) and above-normal snow cover across Eurasia (Cohen et al. 2007) in late autumn can lead to a negative NAO phase in the following winter, suggesting that the recent arctic sea ice decline and Eurasian snow-cover increase in late autumn might have contributed to the NAO pattern shift (Allen and Zender 2011; Cohen et al. 2012; Nakamura et al. 2015). Additionally, several numerical studies suggest that increases in greenhouse gas concentrations in the atmosphere might have played a role in the northeastward shift of the NAO action centers (Ulbrich and Christoph 1999; Hu and Wu 2004; Dong et al. 2011) and give rise to higher frequency of occurrence of the negative NAO phase (Manzini et al. 2014). But it seems that the continuation of global changes cannot explain the recent westward retreat of the winter NAO pattern, which is supported by Semenov et al. (2008), who concluded that the upward trend in the NAO index over the last half of the twentieth century may be internally generated within the coupled atmosphere–ocean system without considering external forcing. Therefore, corresponding changes in other meteorological factors as well as plausible causes for the recent westward retreat of the NAO pattern still require investigation.

Comparing the results found herein with those from Song et al. (2014) and Zuo et al. (2015), there exist two different pathways for the linkage between the NAO and temperature anomalies in south-central China during winter. One is related to a wave train spanning the Arabian Sea, which is concurrent with an intensified (weakened) MEJS in the positive (negative) phase of the NAO. This wave train is excited by the upper-level divergence anomalies over the Mediterranean Sea induced by near-surface Ekman pumping in association with the southern center of the NAO (Zuo et al. 2015). The other is linked to the northern center of the NAO and the associated downstream circulation anomalies, which could significantly influence the cold air activity moving from Siberia into south-central China. The northern pathway features interdecadal variations due to changes in the zonal position of the northern center of the NAO and its amplitude in past decades.

Acknowledgments

This work was jointly supported by the 973 Program of China under Grant 2013CB430203, the China National Science Foundation under Grants 41205058 and 41375062, and the Meteorological special program under Grant GYHY201406022 and also partly supported by the UK–China Research and Innovation Partnership Fund through the Met Office Climate Science for Service Partnership (CSSP) China as part of the Newton Fund. The authors are grateful to three anonymous reviewers for their insightful comments to improve the quality of the paper.

REFERENCES

REFERENCES
Allen
,
R. J.
, and
C. S.
Zender
,
2011
:
Forcing of the Arctic Oscillation by Eurasian snow cover
.
J. Climate
,
24
,
6528
6539
, doi:.
Bader
,
J.
,
M. D. S.
Mesquita
,
K. I.
Hodges
,
N.
Keenlyside
,
S.
Østerhus
, and
M.
Miles
,
2011
:
A review on Northern Hemisphere sea-ice, storminess and the North Atlantic Oscillation: Observations and projected changes
.
Atmos. Res.
,
101
,
809
834
, doi:.
Barnston
,
A. G.
, and
R. E.
Livezey
,
1987
:
Classification, seasonality and persistence of low-frequency atmospheric circulation patterns
.
Mon. Wea. Rev.
,
115
,
1083
1126
, doi:.
Beranová
,
R.
, and
R.
Huth
,
2008
:
Time variations of the effects of circulation variability modes on European temperature and precipitation in winter
.
Int. J. Climatol.
,
28
,
139
158
, doi:.
Branstator
,
G.
,
2002
:
Circumglobal teleconnections, the jet stream waveguide, and the North Atlantic Oscillation
.
J. Climate
,
15
,
1893
1910
, doi:.
Budikova
,
D.
,
2009
:
Role of arctic sea ice in global atmospheric circulation: A review
.
Global Planet. Change
,
68
,
149
163
, doi:.
Cassou
,
C.
,
L.
Terray
,
J. W.
Hurrell
, and
C.
Deser
,
2004
:
North Atlantic winter climate regimes: Spatial asymmetry, stationarity with time, and oceanic forcing
.
J. Climate
,
17
,
1055
1068
, doi:.
Cohen
,
J.
,
M.
Barlow
,
P.
Kushner
, and
K.
Saito
,
2007
:
Stratosphere–troposphere coupling and links with Eurasian land surface variability
.
J. Climate
,
20
,
5335
5343
, doi:.
Cohen
,
J.
,
J. C.
Furtado
,
M. A.
Barlow
,
V. A.
Alexeev
, and
J. E.
Cherry
,
2012
:
Arctic warming, increasing snow cover and widespread boreal winter cooling
.
Environ. Res. Lett.
,
7
,
014007
, doi:.
Davis
,
R. E.
,
1976
:
Predictability of sea surface temperature and sea level pressure anomalies over the North Pacific Ocean
.
J. Phys. Oceanogr.
,
6
,
249
266
, doi:.
Dee
,
D. P.
, and Coauthors
,
2011
:
The ERA-Interim reanalysis: Configuration and performance of the data assimilation system
.
Quart. J. Roy. Meteor. Soc.
,
137
,
553
597
, doi:.
Dong
,
B.
,
R. T.
Sutton
, and
T.
Woollings
,
2011
:
Changes of interannual NAO variability in response to greenhouse gases forcing
.
Climate Dyn.
,
37
,
1621
1641
, doi:.
Fisher
,
R. A.
,
1915
:
Frequency distribution of the values of the correlation coefficient in samples of an indefinitely large population
.
Biometrika
,
10
,
507
521
.
Folland
,
C. K.
,
A. A.
Scaife
,
J.
Lindesay
, and
D. B.
Stephenson
,
2012
:
How potentially predictable is northern European winter climate a season ahead?
Int. J. Climatol.
,
32
,
801
818
, doi:.
Hilmer
,
M.
, and
T.
Jung
,
2000
:
Evidence for a recent change in the link between the North Atlantic Oscillation and arctic sea ice export
.
Geophys. Res. Lett.
,
27
,
989
992
, doi:.
Hoerling
,
M. P.
,
J. W.
Hurrell
,
T.
Xu
,
G. T.
Bates
, and
A. S.
Phillips
,
2004
:
Twentieth century North Atlantic climate change. Part II: Understanding the effect of Indian Ocean warming
.
Climate Dyn.
,
23
,
391
405
, doi:.
Hong
,
C.-C.
,
H.-H.
Hsu
,
H.-H.
Chia
, and
C.-Y.
Wu
,
2008
:
Decadal relationship between the North Atlantic Oscillation and cold surge frequency in Taiwan
.
Geophys. Res. Lett.
,
35
,
L24707
, doi:.
Hoskins
,
B. J.
, and
T.
Ambrizzi
,
1993
:
Rossby wave propagation on a realistic longitudinally varying flow
.
J. Atmos. Sci.
,
50
,
1661
1671
, doi:.
Hu
,
Z.-Z.
, and
Z.
Wu
,
2004
:
The intensification and shift of the annual North Atlantic Oscillation in a global warming scenario simulation
.
Tellus
,
56A
,
112
124
, doi:.
Hurrell
,
J. W.
,
1995
:
Decadal trends in the North Atlantic Oscillation: Regional temperatures and precipitation
.
Science
,
269
,
676
679
, doi:.
Hurrell
,
J. W.
,
Y.
Kushnir
,
M.
Visbeck
, and
G.
Ottersen
,
2003
: An overview of the North Atlantic Oscillation. The North Atlantic Oscillation: Climatic Significance and Environmental Impact, Geophys. Monogr., Vol. 134, Amer. Geophys. Union, 1–35.
Johnson
,
N. C.
,
S. B.
Feldstein
, and
B.
Trembley
,
2008
:
The continuum of Northern Hemisphere teleconnection patterns and a description of the NAO shift with the use of self-organizing maps
.
J. Climate
,
21
,
6354
6371
, doi:.
Jung
,
T.
, and
M.
Hilmer
,
2001
:
The link between the North Atlantic Oscillation and arctic sea ice export through Fram Strait
.
J. Climate
,
14
,
3932
3943
, doi:.
Jung
,
T.
,
M.
Hilmer
,
E.
Ruprecht
,
S.
Kleppek
,
S. K.
Gulev
, and
O.
Zolina
,
2003
:
Characteristics of the recent eastward shift of interannual NAO variability
.
J. Climate
,
16
,
3371
3382
, doi:.
Kalnay
,
E.
, and Coauthors
,
1996
:
The NCEP/NCAR 40-Year Reanalysis Project
.
Bull. Amer. Meteor. Soc.
,
77
,
437
472
, doi:.
Li
,
J.
,
C.
Sun
, and
F.-F.
Jin
,
2013
:
NAO implicated as a predictor of Northern Hemisphere mean temperature multidecadal variability
.
Geophys. Res. Lett.
,
40
,
5497
5502
, doi:.
Lu
,
J.
, and
R. J.
Greatbatch
,
2002
:
The changing relationship between the NAO and Northern Hemisphere climate variability
.
Geophys. Res. Lett.
,
29
,
1148
, doi:.
Manzini
,
E.
, and Coauthors
,
2014
:
Northern winter climate change: Assessment of uncertainty in CMIP5 projections related to stratosphere-troposphere coupling
.
J. Geophys. Res. Atmos.
,
119
,
7979
7998
, doi:.
Moore
,
G. W. K.
,
I. A.
Renfrew
, and
R. S.
Pickart
,
2013
:
Multidecadal mobility of the North Atlantic Oscillation
.
J. Climate
,
26
,
2453
2466
, doi:.
Nakamura
,
T.
,
K.
Yamazaki
,
K.
Iwamoto
,
M.
Honda
,
Y.
Miyoshi
,
Y.
Ogawa
, and
J.
Ukita
,
2015
:
A negative phase shift of the winter AO/NAO due to the recent arctic sea-ice reduction in late autumn
.
J. Geophys. Res. Atmos.
,
120
,
3209
3227
, doi:.
National Climate Center
,
2015
: Temperature data of 160 stations over China. National Climate Center, accessed April 2015. [Available online at http://ncc.cma.gov.cn/Website/index.php?ChannelID=43&WCHID=5.]
NOAA/Climate Prediction Center
,
2015
: Monthly mean North Atlantic Oscillation index since January 1950. NOAA/Climate Prediction Center, accessed April 2015. [Available online at ftp://ftp.cpc.ncep.noaa.gov/cwlinks/.]
Peterson
,
K. A.
,
J.
Lu
, and
R. J.
Greatbatch
,
2003
:
Evidence of nonlinear dynamics in the eastward shift of the NAO
.
Geophys. Res. Lett.
,
30
,
1030
, doi:.
Polyakova
,
E. I.
,
A. G.
Journel
,
I. V.
Polyakov
, and
U. S.
Bhatt
,
2006
:
Changing relationship between the North Atlantic Oscillation and key North Atlantic climate parameters
.
Geophys. Res. Lett.
,
33
,
L03711
, doi:.
Raible
,
C. C.
,
F.
Lehner
,
J. F. G.
Rouco
, and
L. F.
Donado
,
2014
:
Changing correlation structures of the Northern Hemisphere atmospheric circulation from 1000 to 2100 AD
.
Climate Past
,
10
,
537
550
, doi:.
Ren
,
H.-L.
,
F.-F.
Jin
,
J.-S.
Kug
,
J.-X.
Zhao
, and
J.
Park
,
2009
:
A kinematic mechanism for positive feedback between synoptic eddies and NAO
.
Geophys. Res. Lett.
,
36
,
L11709
, doi:.
Ren
,
H.-L.
,
F.-F.
Jin
, and
L.
Gao
,
2012
:
Anatomy of synoptic eddy–NAO interaction through eddy-structure decomposition
.
J. Atmos. Sci.
,
69
,
2171
2191
, doi:.
Rodwell
,
M. J.
,
D. P.
Rowell
, and
C. K.
Folland
,
1999
:
Oceanic forcing of the wintertime North Atlantic Oscillation and European climate
.
Nature
,
398
,
320
323
, doi:.
Scaife
,
A. A.
,
C. K.
Folland
,
L. V.
Alexander
,
A.
Moberg
, and
J. R.
Knight
,
2008
:
European climate extremes and the North Atlantic Oscillation
.
J. Climate
,
21
,
72
83
, doi:.
Scaife
,
A. A.
, and Coauthors
,
2009
:
The CLIVAR C20C project: Selected twentieth century climate events
.
Climate Dyn.
,
33
,
603
614
, doi:.
Schneider
,
E. K.
,
L.
Bengtsson
, and
Z.-Z.
Hu
,
2003
:
Forcing of Northern Hemisphere climate trends
.
J. Atmos. Sci.
,
60
,
1504
1521
, doi:.
Semenov
,
V. A.
,
M.
Latif
,
J. H.
Jungclaus
, and
W.
Park
,
2008
:
Is the observed NAO variability during the instrumental record unusual?
Geophys. Res. Lett.
,
35
,
L11701
, doi:.
Song
,
J.
,
C.
Li
, and
W.
Zhou
,
2014
:
High and low latitude types of the downstream influences of the North Atlantic Oscillation
.
Climate Dyn.
,
42
,
1097
1111
, doi:.
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
, doi:.
Tan
,
G.
,
H.
Chen
,
Z.
Sun
, and
W.
Deng
,
2010
:
Linkage of the cold event in January 2008 over China to the North Atlantic Oscillation and stratospheric circulation anomalies (in Chinese)
.
Chin. J. Atmos. Sci.
,
34
,
175
183
.
Tan
,
G.
,
F.-F.
Jin
,
H.-L.
Ren
, and
Z.-B.
Sun
,
2014
:
The role of eddy feedback in the excitation of NAO
.
Meteor. Appl.
,
21
,
768
776
, doi:.
Ulbrich
,
U.
, and
M.
Christoph
,
1999
:
A shift of the NAO and increasing storm track activity over Europe due to anthropogenic greenhouse gas forcing
.
Climate Dyn.
,
15
,
551
559
, doi:.
Uppala
,
S. M.
, and Coauthors
,
2005
:
The ERA-40 Re-Analysis
.
Quart. J. Roy. Meteor. Soc.
,
131
,
2961
3012
, doi:.
van Loon
,
H.
, and
J. C.
Rogers
,
1978
:
The seesaw in winter temperatures between Greenland and Northern Europe. Part I: General description
.
Mon. Wea. Rev.
,
106
,
296
310
, doi:.
Vicente-Serrano
,
S. M.
, and
J. I.
López-Moreno
,
2008
:
Nonstationary influence of the North Atlantic Oscillation on European precipitation
.
J. Geophys. Res.
,
113
,
D20120
, doi:.
Vihma
,
T.
,
2014
:
Effects of arctic sea ice decline on weather and climate: A review
.
Surv. Geophys.
,
35
,
1175
1214
, doi:.
Wallace
,
J. M.
, and
D. S.
Gutzler
,
1981
:
Teleconnections in the geopotential height field during the Northern Hemisphere winter
.
Mon. Wea. Rev.
,
109
,
784
812
, doi:.
Watanabe
,
M.
,
2004
:
Asian jet waveguide and a downstream extension of the North Atlantic Oscillation
.
J. Climate
,
17
,
4674
4691
, doi:.
Woollings
,
T.
,
A.
Hannachi
,
B.
Hoskins
, and
A.
Turner
,
2010
:
A regime view of the North Atlantic Oscillation and its response to anthropogenic forcing
.
J. Climate
,
23
,
1291
1307
, doi:.
Wu
,
B.
, and
R.
Huang
,
1999
:
Effect of the extremes in the North Atlantic Oscillation on East Asian winter monsoon (in Chinese)
.
Chin. J. Atmos. Sci.
,
23
,
641
651
.
Xu
,
H.
,
J.
Li
,
J.
Feng
, and
J.
Mao
,
2012
:
The asymmetric relationship between the winter NAO and the precipitation in southwest China (in Chinese)
.
Acta Meteor. Sin.
,
70
,
1276
1291
.
Zhang
,
W.
,
L.
Wang
,
B.
Xiang
,
L.
Qi
, and
J.
He
,
2015
:
Impacts of two types of La Niña on the NAO during boreal winter
.
Climate Dyn.
,
44
,
1351
1366
, doi:.
Zuo
,
J.
,
W.
Li
,
H.-L.
Ren
, and
L.
Chen
,
2012
:
Change of the relationship between spring NAO and East Asian summer monsoon and its possible mechanism
.
Chin. J. Geophys.
,
55
,
23
34
, doi:.
Zuo
,
J.
,
H.-L.
Ren
, and
W.
Li
,
2015
:
Contrasting impacts of the Arctic Oscillation on surface air temperature anomalies in southern China between early and middle-to-late winter
.
J. Climate
,
28
,
4015
4026
, doi:.
Zuo
,
J.
,
H.-L.
Ren
,
B.
Wu
, and
W.
Li
,
2016
:
Predictability of winter temperature in China from previous autumn Arctic sea ice
.
Climate Dyn.
, doi:, in press.