Abstract

Using NCEP–NCAR reanalysis and Japanese 25-yr Reanalysis (JRA-25) winter daily (1 December–28 February) data for the period 1979–2012, this paper reveals the leading pattern of winter daily 850-hPa wind variability over northern Eurasia from a dynamic perspective. The results show that the leading pattern accounts for 18% of the total anomalous kinetic energy and consists of two subpatterns: the dipole and the tripole wind patterns. The dipole wind pattern does not exhibit any apparent trend. The tripole wind pattern, however, has displayed significant trends since the late 1980s. The negative phase of the tripole wind pattern corresponds to an anomalous anticyclone over northern Eurasia during winter, as well as two anomalous cyclones occurring over southern Europe and in the mid- to high latitudes of East Asia. These anomalous cyclones in turn lead to enhanced winter precipitation in these two regions, as well as negative surface temperature anomalies over the mid- to high latitudes of Asia. The intensity of the tripole wind pattern and the frequency of its extreme negative phase are significantly correlated with autumn Arctic sea ice anomalies. Simulation experiments further demonstrate that the winter atmospheric response to Arctic sea ice decrease is dynamically consistent with the observed trend in the tripole wind pattern over the past 24 winters, which is one of the causes of the observed declining winter surface air temperature trend over Central and East Asia. The results of this study also imply that East Asia may experience more frequent and/or intense winter extreme weather events in association with the loss of Arctic sea ice.

1. Introduction

Wind variations, especially surface wind variations, play important roles in regulating air–sea and air–land interactions by means of affecting heat flux and water evaporation, thus further influencing soil moisture and the hydrological cycle (e.g., Wever 2012; McVicar et al. 2012). As the carrier of energy, water vapor, and aerosol, wind anomalies are directly associated with extreme weather events and air environmental variations. During the boreal winter season, wind anomalies are usually connected to cold-air outbreaks and heavy snowfall, such as during the winter of 2011/12 in Eurasia. In late January 2012, extreme cold polar air carried by a suddenly strengthened anticyclone (Siberian high) spreads over nearly all of Eurasia, leading to a very rare extreme cold event, as well as heavy snowfall in Europe. On the other hand, cold-air outbreaks play a critical role in improving the environmental air quality over East Asia, particularly over central and eastern China.

Recently, countries in East Asia have experienced frequent cold winters and extreme weather events. China has been subjected to three successive cold winters during 2009–12. The average surface air temperature from December 2012 to the beginning of January 2013 reached its lowest record in China over the past 28 winters. Additionally, anomalous low temperatures, heavy snowfalls, and frozen precipitation also occurred in East Asia, including heavy snowfall over Japan in December 2005, successive frozen precipitation events over southern China in January–February 2008, extreme heavy snowfall in northwestern China during the winter of 2009/10, and anomalous low temperatures and frequent heavy snowfall events in north and northeast China from November to December 2012. Indeed, recent cold winters in some East Asian countries have been directly connected to the recovery of the Siberian high intensity and its frequent positive anomalies since 2004 (Jeong et al. 2011; Wu et al. 2011).

Although some studies have investigated the predominant patterns of winter wind variability over East Asia and the Arctic (e.g., Wu et al. 2012), some important issues still remain unresolved. For example, what are the dominant patterns of daily wind field variability over northern Eurasia during the winter season? Furthermore, do these dominant patterns exhibit significant trends in their intensity and frequency? Although the Arctic Oscillation (AO) and its regional counterpart, the North Atlantic Oscillation (NAO), directly affect the synoptic regime and climate variability of northern Eurasia, the differences between the AO (or NAO) and the dominant patterns of wind variability are quite significant, due to the fact that they reflect different physical mechanisms (Wu et al. 2012).

The aims of the present study are to explore the leading pattern of winter daily 850-hPa wind variability and its trends over northern Eurasia, and to examine the associations between the intensity and frequency of the leading pattern and previous autumn (September–November) sea ice concentration (SIC) anomalies. Investigating these issues will aid in our understanding the dominant features of winter wind variability and the possible causes, which will be beneficial in improving our prediction ability for extreme weather events.

2. Data and method

The data used in this study include the following: 1) a 33-winter (1979/80–2011/12) dataset of daily sea level pressure (SLP), surface air temperature (SAT), 850-hPa winds, and 500-hPa geopotential heights from 1 December to 28 February of the next year (for a total of 2970 days) obtained from the National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) 40-Year Reanalysis Project and the Japanese 25-yr Reanalysis (JRA-25; Onogi et al. 2007; http://ds.data.jma.go.jp/gmd/jra/download/category-e.html/anl_p25); 2) monthly mean global land precipitation data from 1979 to 2011 (Chen et al. 2002; http://ftp.cpc.ncep.noaa.gov/precip/50yr/gauge/2.5deg/format_bin/); 3) the monthly Arctic SIC dataset (on a 1° latitude × 1° longitude grid) for the period of 1979–2012 obtained from the British Atmospheric Data Centre (BADC; http://badc.nerc.ac.uk/data/hadisst/); and 4) the monthly mean AO index during the period 1979–2012 (http://www.cpc.ncep.noaa.gov/products/precip/CWlink/daily_ao_index/monthly.ao.index.b50.current.ascii).

To reveal the leading pattern of daily 850-hPa wind variability during the boreal winter season, the complex vector empirical orthogonal function (CVEOF) analysis method was applied to normalized wind anomalies that were derived from the daily 850-hPa winds subtracting the corresponding daily climatology over the 33 winters, and the domain of the CVEOF analysis was 40°–70°N, 40°–120°E. It should be noted that if a larger domain was used, for example, over 40°–70°N, 20°–120°E, the results would be similar to that over 40°–70°N, 40°–120°E. The additional motivation for selecting this particular domain is to focus on the influence of the dominant wind patterns over Asia. Instead of surface wind fields, we used the 850-hPa wind field in this study. This is because surface wind fields possess strong inhomogeneities, which may not be related to systematic climate variability (Wever 2012). For more detailed information concerning the statistical and physical meanings of the CVEOF method and its advantages over the traditional EOF method, the reader is referred to Wu et al. (2012). Additionally, the maximum covariance analysis (MCA; see von Storch and Zwiers 1999) is also used to further detect coupled patterns between autumn SIC over the domain 50°–90°N, 0°–360° and the ensuing winter meridional winds at 850 hPa over the domain 40°–70°N, 40°–120°E.

The Monte Carlo method is applied to determine statistical field significance, as described in Livezey and Chen (1983). For an anomalous field derived from linear regression, the percentage of grid points that are statistically significant at the 0.05 level is first identified over a given domain. This process is then repeated 1000 times using different series of numbers randomly selected from a normal distribution. The anomalous field is deemed significant if the percentage of the significant grid points exceeds that derived from the 1000 experimental replications.

Additionally, the ECHAM5 (Roeckner et al. 2003) model (T63 spectral resolution and 19 pressure levels) was applied to explore the impacts of SIC on the model atmosphere. A 30-yr simulation with the climatological monthly sea surface temperature (SST) and observed Northern Hemisphere monthly SIC from 1978 to 2007 as the external forcing was performed (the climatological monthly SST and SIC observational data were obtained online: http://www-pcmdi.llnl.gov/projects/amip/AMIP2EXPDSN/BCS/bcsintro.php), and this experiment was repeated with 12 different atmospheric initial conditions, which were derived from a 30-yr control run.

3. Spatial features of the leading wind pattern and physical significance of the phase evolution

The leading pattern of winter daily 850-hPa wind variability (for a total of 2970 days) accounts for 18% of the total anomalous kinetic energy. To demonstrate the leading pattern's spatial evolution, composite analyses were performed for the following typical four different leading phase ranges: the 0° phase ( < 45° or ≥ 315°), 90° phase (45° ≤ < 135°), 180° phase (135° ≤ < 225°), and 270° phase (225° ≤ < 315°). Table 1 shows the respective frequencies for the typical four different phase ranges.

Table 1.

Frequencies for different phase ranges from 1979 to 2012.

Frequencies for different phase ranges from 1979 to 2012.
Frequencies for different phase ranges from 1979 to 2012.

For the 0° phase, the predominant feature is an anomalous cyclone and anticyclone pair occupying, respectively, the northern Asian continent–Siberian marginal seas of the Arctic Ocean and Europe (Fig. 1a). Meanwhile, there are weak anomalous anticyclonic and cyclonic centers over the East Asian coast and the Mediterranean. Correspondingly, SLP anomalies exhibit a dipole structure with opposite anomalous centers covering, respectively, eastern Europe and the area between Taymyr and Lake Baikal in Siberia (Fig. 2a). Negative SAT anomalies appear in northern Asia and the Arctic Ocean, and there are two separated centers of positive SAT anomalies south of the Barents and Kara Seas and around Lake Baikal (Fig. 3a). It is shown that positive SAT anomalies cover most of East Asia. When the leading wind pattern is in its 90° phase, wind anomalies show a tripole structure and a dominant anomalous cyclone occupies a region from northern Eurasia and the Barents Sea eastward to the Laptev Sea, with its center close to south of the Kara Sea (Fig. 1b). The two weak anomalous anticyclones are located over the southern section of Europe and over northeastern Asia. The spatial distribution of the wind anomalies is dynamically consistent with a dominant monopole structure of the SLP anomalies (Fig. 2b). Furthermore, corresponding SAT anomalies also show a dipole pattern with opposite anomalous centers located over the northern Asian continent and over the Kara Sea (Fig. 3b). Cold- and warm-air advection induced by wind anomalies is responsible for the spatial distribution of SAT anomalies. The anomalous wind, SLP, and SAT patterns corresponding to the 180° (270°) phase show the opposite scenario to that in the 0° (90°) phase (Figs. 1c,d, 2c,d, and 3c,d). It is seen that SAT anomalies in most of East Asia are out of phase with those in the Arctic Ocean and its marginal seas.

Fig. 1.

(a) Composite of winter daily 850-hPa wind anomalies for the 0° ( < 45° or ≥ 315°) phase of the leading wind pattern, (b)–(d) as in (a), but for the 90° (45° ≤ < 135°), 180° (135° ≤ < 225°), and 270° (225° ≤ < 315°) phases, respectively. Units are m s−1.

Fig. 1.

(a) Composite of winter daily 850-hPa wind anomalies for the 0° ( < 45° or ≥ 315°) phase of the leading wind pattern, (b)–(d) as in (a), but for the 90° (45° ≤ < 135°), 180° (135° ≤ < 225°), and 270° (225° ≤ < 315°) phases, respectively. Units are m s−1.

Fig. 2.

As in Fig. 1, but for composite of winter daily SLP anomalies for different phase ranges; intervals are 2 hPa.

Fig. 2.

As in Fig. 1, but for composite of winter daily SLP anomalies for different phase ranges; intervals are 2 hPa.

Fig. 3.

As in Fig. 1, but for composite of winter daily SAT anomalies for different phase ranges; intervals are 1°C.

Fig. 3.

As in Fig. 1, but for composite of winter daily SAT anomalies for different phase ranges; intervals are 1°C.

The above analysis indicates that the leading wind pattern consists of two subpatterns, and the corresponding SLP and SAT anomalies also display different spatial structures. Consequently, the evolution of the leading phase reflects the spatial evolution of the two subpatterns and their frequencies (Table 1). In this study, the two subpatterns are referred to as the dipole and the tripole wind patterns, respectively. As indicated in our previous study (Wu et al. 2012), the real and imaginary parts of the leading complex principal component can be regarded as two intensity indices used to characterize the two wind patterns. Thus, the dipole wind pattern incorporates the 0° and 180° phases of the leading wind pattern, and a positive (negative) phase of the dipole wind pattern corresponds to the 0° (180°) phase. Similarly, the tripole wind pattern incorporates the 90° and 270° phases of the leading wind pattern, and a positive (negative) phase of the tripole wind pattern corresponds to the 90° (270°) phase.

In the midtroposphere, the evolution of height anomalies, from the 0° to 270° phases, displays a coherent westward migration process (Fig. 4). In Asia, east of 60°E, this westward migration process is concurrent with the strengthened amplitudes of height anomalies. In Europe, west of 60°E, along with a westward shift, the amplitude of the height anomalies is weakened. A very similar pattern of evolution also appears in the SLP anomalies (Fig. 2). This westward shift essentially reflects a coherent westward migration of troughs and ridges of height fields in the midtroposphere, which is relevant to Rossby waves (Francis and Vavrus 2012). Indeed, a single dipole or tripole wind pattern is not able to characterize the propagation of Rossby waves.

Fig. 4.

(a) Composite of winter daily 500-hPa geopotential heights and corresponding geopotential height anomalies for the 0° ( < 45° or ≥ 315°) phase of the leading wind pattern, (b)–(d) as in (a), but for the 90° (45° ≤ < 135°), 180° (135° ≤ < 225°), and 270° (225° ≤ < 315°) phases, respectively; contour intervals are 50 and 20 gpm, respectively.

Fig. 4.

(a) Composite of winter daily 500-hPa geopotential heights and corresponding geopotential height anomalies for the 0° ( < 45° or ≥ 315°) phase of the leading wind pattern, (b)–(d) as in (a), but for the 90° (45° ≤ < 135°), 180° (135° ≤ < 225°), and 270° (225° ≤ < 315°) phases, respectively; contour intervals are 50 and 20 gpm, respectively.

4. Time evolution of the two wind patterns

The intensity of the two wind patterns shows strong interannual variability but no significant trend over the entire data record from 1979 to 2012 (Figs. 5a,b). However, some previous studies have shown that winter atmospheric circulation experienced significant changes in the late 1980s (Walsh et al. 1996; Tanaka et al. 1996; Tachibana et al. 1996; Watanabe and Nitta 1999). It is found that both the SLP over the central Arctic and the polar vortex (the area-mean vorticity over 80°–90°N) throughout the troposphere have decreased noticeably since 1988 (Walsh et al. 1996; Tanaka et al. 1996). Thus, this study chooses the period of 1988–2012 to discuss trends since 1988 only. The tripole wind pattern has shown a negative trend intensity (at the 99% significance level) since the winter of 1988/89 (Fig. 5b), implying a coherent strengthening trend in the anomalous anticyclone over northern Eurasia and two anomalous cyclones over southern Europe and northeastern Asia. Thus, this negative trend is dynamical consistent with the recent recovery of the winter Siberian high intensity over the past two decades (Jeong et al. 2011; Wu et al. 2011). If slightly different start winters were considered, for example from 1986/87 to 1990/91, the decline trend still reached the 99% significance level. In the past four winters, the intensity of the tripole wind pattern was −1.0 or less, indicating that strong anomalous anticyclones prevailed over northern Eurasia during those winters. In the winter of 2011/12, the intensity of the tripole wind pattern was the strongest in the entire study period, consistent with the extreme cold winter over Eurasia.

Fig. 5.

Normalized winter mean intensity time series of the (a) dipole and (b) tripole wind patterns. Frequencies of the (c) positive and (e) negative phases of the dipole wind pattern. (d),(f) As in (c),(e), respectively, but for the tripole wind pattern; dashed red lines represent linear trends since the winter of 1988/89.

Fig. 5.

Normalized winter mean intensity time series of the (a) dipole and (b) tripole wind patterns. Frequencies of the (c) positive and (e) negative phases of the dipole wind pattern. (d),(f) As in (c),(e), respectively, but for the tripole wind pattern; dashed red lines represent linear trends since the winter of 1988/89.

The following two aspects are noteworthy: 1) the frequency of the tripole wind pattern was far greater than the dipole wind pattern (Table 1), and 2) neither the intensity nor the frequency of the dipole wind pattern exhibits any significant trend (Figs. 5c,e) since the winter of 1988/89. Based on these findings, we decide to focus primarily on the tripole wind pattern in this study. Although frequencies of the positive phase of the tripole wind pattern did not exhibit any significant trend during the study period, its decline trend was at the 99% significance level after the winter of 1987/88 (Fig. 5d). Frequency reached its record lowest value during the winter of 2011/12 (<10 times). In contrast, frequencies of the negative phase of the tripole wind pattern have shown a positive trend at the 99% significance level after the winter of 1987/88, and the maximum occurred during the winter of 2011/12 (Fig. 5f). This positive trend is dynamically consistent with the recent recovery of the Siberian high intensity and observed negative trend in winter land SAT anomalies over Eurasia (Fig. 6a). The SAT over Central and East Asia exhibits significant negative trends, consistent with previous studies (e.g., Fig. 1 of Cohen et al. 2009; Fig. 5 of Cohen et al. 2012; Figs. 6 and 7 of Wu et al. 2011). Additionally, winter SAT trends are also negative over much of the mid- to high latitudes of Asia during the period 1979–2012 (Fig. 6b). It is noteworthy that significant surface warming trends were observed over the Tibetan Plateau. The causes of winter surface warming trends around the Tibetan Plateau and its impacts on East Asia deserve investigations in the future. It should be pointed out that a very similar tripole wind pattern was also detected from the JRA-25 data (not shown).

Fig. 6.

Spatial distributions of winter mean SAT trends over (a) the past 24 winters (1988/89–2011/12) and (b) the past 33 winters (1979/80–2011/12). The magenta (light blue) and red (blue) areas denote positive (negative) trends at the 95% and 99% significance levels, respectively; intervals are 1°C (10 yr)−1.

Fig. 6.

Spatial distributions of winter mean SAT trends over (a) the past 24 winters (1988/89–2011/12) and (b) the past 33 winters (1979/80–2011/12). The magenta (light blue) and red (blue) areas denote positive (negative) trends at the 95% and 99% significance levels, respectively; intervals are 1°C (10 yr)−1.

According to the daily intensity time series of the tripole wind pattern (for a total of 2970 days; not shown), the extreme tripole wind pattern is defined as its standard deviation being less than −1.28 (the extreme negative phase range) or greater than 1.28 (the extreme positive phase range), and each phase range corresponds to the probability of the tripole wind pattern being less than 10%. It is found that cumulative frequencies for the extreme negative (positive) phase range are 332 (307) times. Although the mean intensities of the two extreme phase ranges do not display any apparent trend (not shown), their frequency time series show distinctive evolution trends from the winter of 1988/89 (both are at the 99% significance level; Figs. 7a,b). The extreme negative phase of the tripole wind pattern prevailed during the winter of 2011/12, and cumulative frequencies were 36 times, occurring mainly during mid-January–mid-February of 2012 (not shown), which led to extreme snowfall events in Europe and Japan. The correlations of the winter AO with two time series in Fig. 7 are 0.35 and −0.35, respectively (at the 95% marginal significance level), suggesting that the negative phase of the winter AO tends to be favorable for the occurrence of the extreme tripole wind pattern in its negative phase.

Fig. 7.

Frequencies of the extreme (a) positive and (b) negative phase ranges of the tripole wind pattern during the winter season. Dashed red lines represent linear trends since the winter of 1988/89.

Fig. 7.

Frequencies of the extreme (a) positive and (b) negative phase ranges of the tripole wind pattern during the winter season. Dashed red lines represent linear trends since the winter of 1988/89.

The tripole wind pattern influences Eurasian precipitation, and its intensity and frequency of negative phases produce similar anomalous precipitation patterns (Figs. 8a,b). Increased precipitation mainly occurs in southern Europe and the mid- to high latitudes of East Asia, particularly in the latter where the increase of precipitation exceeds 20% (Fig. 8b). Comparing Fig. 1d, increased precipitation is dynamically consistent with an anomalous cyclone, with the sole exception being in the midlatitude Asian continent between 80° and 110°E, where increased precipitation is associated with anomalous convergence.

Fig. 8.

(a) Winter mean anomalous precipitation percentages, derived from a linear regression on the inverted normalized winter mean intensity time series of the tripole wind pattern. (b) As in (a), but for regression on the normalized frequency time series of the negative phase of the tripole wind pattern. The magenta (light blue) and red (blue) areas denote negative (positive) precipitation anomalies at the 95% and 99% significance levels, respectively. Units are %.

Fig. 8.

(a) Winter mean anomalous precipitation percentages, derived from a linear regression on the inverted normalized winter mean intensity time series of the tripole wind pattern. (b) As in (a), but for regression on the normalized frequency time series of the negative phase of the tripole wind pattern. The magenta (light blue) and red (blue) areas denote negative (positive) precipitation anomalies at the 95% and 99% significance levels, respectively. Units are %.

5. Possible associations with autumn SIC

Arctic sea ice loss has been evident particularly since the rapid decline in summer Arctic sea ice starting in the late 1990s (e.g., Comiso et al. 2008), which may affect not only Arctic atmospheric temperature and thickness, but also climate variability in remote regions (Francis et al. 2009; Screen and Simmonds 2010; Overland and Wang 2010; Deser et al. 2004, 2007; Dethloff et al. 2006; Wu et al. 2011; Francis and Vavrus 2012; among others). Recently, some regions of Eurasia have experienced cold winters, such as the winters of 2007/08, 2009/10, 2010/11, and 2011/12 (Fig. 9). It appears that cold winters have become more frequent over East Asia. Some studies have shown that less Arctic sea ice in the previous autumn/winter plays a role in the cold Eurasian weather and enhanced Siberian high, through large-scale dynamic and thermal processes (Francis et al. 2009; Honda et al. 2009; Petoukhov and Semenov 2010; Jaiser et al. 2012; Wu et al. 1999, 2011; Hopsch et al. 2012; Tang et al. 2013). On the synoptic time scale, Francis and Vavrus (2012) have provided an original view on the increased frequency and magnitude of weather extremes over North America and the North Atlantic sector. They indicated that the Arctic amplification results in decreases of winter upper-level zonal winds, which cause more persistent weather conditions that could increase the likelihood of certain types of extreme weather.

Fig. 9.

Spatial distributions of winter mean SAT anomalies: (a) 2007/08, (b) 2009/10, (c) 2010/11, and (d) 2011/12. Intervals are 2°C.

Fig. 9.

Spatial distributions of winter mean SAT anomalies: (a) 2007/08, (b) 2009/10, (c) 2010/11, and (d) 2011/12. Intervals are 2°C.

In fact, the intensity of the tripole wind pattern is significantly correlated with the previous autumn's (September–November) SIC (Fig. 10a). Significant positive SIC anomalies are observed in the Pacific sector of the Arctic Ocean and in the area close to northern parts of the Siberian marginal seas. The above (below) normal autumn SIC in these areas tends to correspond to the positive (negative) phase of the tripole wind pattern. The frequency of the extreme negative phase of the tripole wind pattern is also closely correlated with the previous autumn SIC, and significantly negative SIC anomalies appear also in the Pacific sector of the Arctic Ocean and in the north of the Barents Sea across north of the Kara Sea to north of the Laptev Sea (Fig. 10b). Compared with the previous autumn, winter SIC anomalies associated with the tripole wind pattern are obviously weaker (Figs. 10c,d). The result of the Monte Carlo simulations implies that anomalous SIC fields in Figs. 10a,b are at the 95% statistical significance level. The correlation between the tripole wind pattern and the previous autumn SIC is further supported by the leading coupled pattern between autumn SIC and 850-hPa meridional winds in the ensuing winter (Fig. 11). It is seen that decreased autumn Arctic SIC in the Siberian marginal seas of the Arctic Ocean is associated with northerly anomalies over the mid- to high latitudes of the Asian continent and southerly anomalies over most of Europe. Thus, decreasing autumn SIC may favor the more frequent occurrence of the tripole wind pattern with the negative phase.

Fig. 10.

(a) Autumn mean SIC anomalies, derived from a linear regression on the normalized winter mean intensity of the tripole wind pattern. The magenta (light blue) and red (blue) areas denote negative (positive) SIC anomalies at the 95% and 99% significance levels, respectively. (b) As in (a), but regressed on the normalized frequency time series of the extreme negative phase ranges of the tripole wind pattern. (c),(d) As in (a),(b), respectively, but for winter mean SIC anomalies. Intervals are 3%.

Fig. 10.

(a) Autumn mean SIC anomalies, derived from a linear regression on the normalized winter mean intensity of the tripole wind pattern. The magenta (light blue) and red (blue) areas denote negative (positive) SIC anomalies at the 95% and 99% significance levels, respectively. (b) As in (a), but regressed on the normalized frequency time series of the extreme negative phase ranges of the tripole wind pattern. (c),(d) As in (a),(b), respectively, but for winter mean SIC anomalies. Intervals are 3%.

Fig. 11.

(a) Normalized time series of the leading coupled pattern between autumn mean SIC north of 50°N (blue line) and the ensuing winter mean meridional winds at 850 hPa (NCEP–NCAR reanalysis data) over the domain 40°–70°N and 40°–120°E (red solid line; red dashed line represents its trend; their correlation is 0.8). The leading coupled pattern (MCA1) accounts for 59% of the covariance. (b) Autumn mean SIC anomalies, derived from a linear regression on the normalized inverted time series of 850-hPa meridional winds in the leading coupled pattern (%). (c) As in (b), but for the ensuing winter mean 850-hPa meridional wind anomalies (m s−1), derived from a linear regression on the normalized inverted time series of autumn SIC in the leading coupled pattern.

Fig. 11.

(a) Normalized time series of the leading coupled pattern between autumn mean SIC north of 50°N (blue line) and the ensuing winter mean meridional winds at 850 hPa (NCEP–NCAR reanalysis data) over the domain 40°–70°N and 40°–120°E (red solid line; red dashed line represents its trend; their correlation is 0.8). The leading coupled pattern (MCA1) accounts for 59% of the covariance. (b) Autumn mean SIC anomalies, derived from a linear regression on the normalized inverted time series of 850-hPa meridional winds in the leading coupled pattern (%). (c) As in (b), but for the ensuing winter mean 850-hPa meridional wind anomalies (m s−1), derived from a linear regression on the normalized inverted time series of autumn SIC in the leading coupled pattern.

The evidence shows that autumn Arctic SIC decreases and the Arctic amplification (Screen and Simmonds 2010) are associated with significant increases in the winter mean SLP averaged over 1999–2012 in northern Eurasia relative to that in 1988–99 (Fig. 12a). At the midtroposphere, positive height anomalies emerge over the Arctic and around the Ural Mountain and negative height anomalies appear over Europe and the mid- to high latitudes of East Asia (Fig. 12b). It is seen that over northern Eurasia, the spatial distribution of SLP (500-hPa height) anomalies resembles that in Fig. 2d (Fig. 4d).

Fig. 12.

Winter mean (a) SLP and (b) 500-hPa height anomalies during 1999–2012 (13 winters) relative to 1988–99 (11 winters). The magenta (light blue) and red (blue) areas denote positive (negative) anomalies at the 95% and 99% significance levels, respectively. Intervals are 1 hPa in (a) and 10 gpm in (b). (c),(d) As in (a),(b), respectively, but for the simulated winter mean SLP and 500-hPa height anomalies during 1999–2007 relative to 1988–99, derived from simulation experiments of the ECHAM5 model forced by observed Northern Hemisphere SIC during 1978–2007. (e) As in (d), but for the simulated winter mean SAT anomalies. (f) Significance test of winter mean SAT anomalies in (e); the meaning for color areas is same as in (a). Intervals are 0.3 hPa in (c), 5 gpm in (d), and 0.5°C in (e).

Fig. 12.

Winter mean (a) SLP and (b) 500-hPa height anomalies during 1999–2012 (13 winters) relative to 1988–99 (11 winters). The magenta (light blue) and red (blue) areas denote positive (negative) anomalies at the 95% and 99% significance levels, respectively. Intervals are 1 hPa in (a) and 10 gpm in (b). (c),(d) As in (a),(b), respectively, but for the simulated winter mean SLP and 500-hPa height anomalies during 1999–2007 relative to 1988–99, derived from simulation experiments of the ECHAM5 model forced by observed Northern Hemisphere SIC during 1978–2007. (e) As in (d), but for the simulated winter mean SAT anomalies. (f) Significance test of winter mean SAT anomalies in (e); the meaning for color areas is same as in (a). Intervals are 0.3 hPa in (c), 5 gpm in (d), and 0.5°C in (e).

To further study these relationships, we applied the same analysis process to the simulation results of the ECHAM5 model forced by observed Northern Hemisphere SIC during 1978–2007. Here, only the simulated winter mean SLP, 500-hPa height, and SAT anomalies during the period of 1999–2007 relative to those of 1988–99 are analyzed. Thus, 12 simulation experiments contain 96 (132) winter cases during the period 1999–2007 (1988–99). The simulated winter mean SLP and 500-hPa height anomalies, however, derived from all simulation experiments, do not hold the same relationships as in the observations (not shown). Through inspecting differences in winter mean SLP and 500-hPa height between two periods (1999–2007 minus 1988–99) in each 30-yr simulation, it is found that only five of the experiments, to a great extent, can reproduce the major features of the observed atmospheric circulation anomalies over northern Eurasia. Thus, the five simulation experiments contain 55 (40) winter cases during 1988–99 (1999–2007), and composite anomalies are shown in Figs. 12c–f. Over northern Eurasia, the spatial distributions of simulated SLP and 500-hPa height anomalies closely resemble the observations though the amplitudes of these anomalies are relatively weaker. At the surface, accompanying Arctic sea ice loss, significant warming is observed over the Arctic Ocean and its marginal seas, and concurrent negative SAT anomalies emerge over much of the Asian continent (Figs. 12e,f). Thus, the ECHAM5 simulations suggest that a similar mechanism, as indicated from the observations, may be working in the model. Due to the differences between the observed and simulated patterns as shown in Fig. 12, however, and due to the fact that only five simulations reproduce the major features of the observed atmospheric circulation anomalies over northern Eurasia, the influence of internal variability and model uncertainties is suggested to be rather large. Further, the initial atmospheric conditions may also influence the atmospheric response to the SIC forcing, and further investigations are needed.

Many previous studies have suggested that decreased autumn Arctic sea ice would lead to an intensified heat loss from the ocean and a stronger heating effect on the overlying atmosphere, which strengthens the atmospheric baroclinicity and instability (Alexander et al. 2004; Jaiser et al. 2012; Porter et al. 2012). As the seasons progress, through a negative feedback process, baroclinic atmospheric processes diminish and barotropic interactions connected with the development of long planetary waves become more important, resulting in frequent anticyclonic anomalies over northern Eurasia and the Siberian marginal seas of the Arctic Ocean (Alexander et al. 2004; Deser et al. 2004, 2007, 2010; Jaiser et al. 2012; Wu et al. 2011). Thus, we summarize in Fig. 13 the schematic of how reduced Arctic sea ice influences the winter SAT and precipitation across Eurasia. Through the negative feedback process, persistent negative autumn/winter Arctic SIC anomalies cause the frequent occurrence of the winter anomalous atmospheric blocking pattern over northern Eurasia and some marginal seas of the Arctic Ocean. This atmospheric blocking pattern induces cold-air outbreaks southward from the polar region, leading to negative SAT anomalies in the mid- to high latitudes of Asia. Meanwhile, through the anomalous atmospheric blocking pattern affecting the prevailing pattern of winter atmospheric variability, the reduced autumn/winter Arctic SIC influences weather events over the midlatitudes of Eurasia. This feedback process indeed involves the roles of sea surface temperature (SST) anomalies in the northern North Atlantic and subarctic areas (Deser et al. 2007; Wu et al. 2011). In addition, other than the Arctic SIC, other factors, such as SST in the North Atlantic and the Atlantic multidecadal oscillation (AMO), may also contribute to increases in winter anomalous blocking patterns over northern Eurasia (Li 2004; Peng et al. 2003).

Fig. 13.

Schematic of how reduced Arctic sea ice affects winter SAT and precipitation tendencies across Eurasia. Arrows denote the spatial distribution of the anomalous anticyclone and cyclone associated with the negative phase of the tripole wind pattern in the lower troposphere. The brown line represents the 500-hPa height isoline. Yellow and green areas indicate less and more precipitation, respectively. Red and purple areas, respectively, depict positive and negative SAT anomalies.

Fig. 13.

Schematic of how reduced Arctic sea ice affects winter SAT and precipitation tendencies across Eurasia. Arrows denote the spatial distribution of the anomalous anticyclone and cyclone associated with the negative phase of the tripole wind pattern in the lower troposphere. The brown line represents the 500-hPa height isoline. Yellow and green areas indicate less and more precipitation, respectively. Red and purple areas, respectively, depict positive and negative SAT anomalies.

On the other hand, a short-term trend in the tripole wind pattern could reflect the nature of the interdecadal (or multidecadal) variability of the regional climate system, which contributes a great deal to the short-term trend. In fact, a coherent pattern of interdecadal (or multidecadal) variability in the atmosphere–ocean–ice system in the Arctic and northern Eurasia requires that Arctic sea ice anomalies have an important impact on the atmosphere (Mysak et al. 1990; Mysak and Venegas 1998). It is possible that Central and East Asia may experience more frequent cold winters under the background of an increasing trend in the tripole wind pattern and a declining trend in autumn sea ice. We are left to ask the following: How long will the increasing trend in the tripole wind pattern persist? What is the underlining forcing(s) necessary to maintain this increasing trend—nature decadal–interdecadal variability, or does the enhanced anthropogenic forcing also played a role, or some combination of the two? These questions are relevant for the proposed dynamic feedback mechanism and for the seasonal prediction of the East Asian winter monsoon, and they need to be studied further in the future.

6. Summary

This paper reveals the leading pattern of daily 850-hPa wind variability during the winter over northern Eurasia from a dynamical perspective. It is found that the leading pattern accounts for 18% of the total anomalous kinetic energy and consists of two subpatterns: the dipole and tripole wind patterns. The dipole wind pattern does not exhibit any significant trend in the study period. The tripole wind pattern, however, has displayed significant trends in its intensity and frequency since the late 1980s. The negative phase of the tripole wind pattern corresponds to anomalous cyclones over southern Europe and over the mid- to high latitudes of East Asia, thereby enhancing winter precipitation in those regions. The intensity of the tripole wind pattern and the frequency of its extreme negative phase are significantly correlated with autumn Arctic SIC. The simulated results from the ECHAM5 model forced by observed Northern Hemisphere SIC during 1978–2007 suggest that the decreased Arctic SIC causes strengthening of the winter anomalous blocking pattern over northern Eurasia, supporting the observational association. Thus, autumn Arctic SIC may provide a potential precursor for the ensuing winter daily tripole wind pattern and precipitation in southern Europe and the mid- to high latitudes of East Asia. The results of this study imply that East Asia may experience more frequent and/or intense winter extreme weather events in association with the loss of Arctic sea ice.

Acknowledgments

The authors are grateful to Dr. Jennifer A. Frances, Dr. Judah L. Cohen, and an anonymous reviewer for their support and constructive suggestions, which helped to significantly improve this paper. The authors thank Dr. Jingzhi Su for kindly providing simulation experiments with the ECHAM5 model forced by Northern Hemisphere sea ice concentration. This study was supported by the National Key Basic Research Project of China (2013CBA01804), the National Natural Science Foundation of China (41221064), the Chinese Project (GYHY200906017), and the State Oceanic Administration Project (201205007). AH was supported by the Office of Science (BER), U.S. Department of Energy, Cooperative Agreement DE-FC02-97ER62402. The National Center for Atmospheric Research is sponsored by the National Science Foundation.

REFERENCES

REFERENCES
Alexander
,
A.
,
S.
Bhatt
, and
J.
Walsh
,
2004
:
The atmospheric response to realistic sea ice anomalies in an AGCM during winter
.
J. Climate
,
17
,
890
905
.
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
.
Cohen
,
J.
,
M.
Barlow
, and
K.
Saito
,
2009
:
Decadal fluctuations in planetary wave forcing modulate global warming in late boreal winter
.
J. Climate
,
22
,
4418
4426
.
Cohen
,
J.
,
J.
Furtado
,
M.
Barlow
,
V.
Alexeev
, and
J.
Cherry
,
2012
:
Arctic warming, increasing snow cover and widespread boreal winter cooling
.
Environ. Res. Lett.
,
7
,
014007
,
doi:10.1088/1748-9326/7/1/014007
.
Comiso
,
J.
,
C.
Parkinson
,
R.
Gersten
, and
L.
Stock
,
2008
:
Accelerated decline in the Arctic sea ice cover
.
Geophys. Res. Lett.
,
35
,
L01703
,
doi:10.1029/2007GL031972
.
Deser
,
C.
,
G.
Magnusdottir
,
R.
Saravanan
, and
A.
Phillips
,
2004
:
The effects of North Atlantic SST and sea ice anomalies on the winter circulation in CCM3. Part II: Direct and indirect components of the response
.
J. Climate
,
17
,
877
889
.
Deser
,
C.
,
A.
Tomas
, and
S.
Peng
,
2007
:
The transient atmospheric circulation response to North Atlantic SST and sea ice anomalies
.
J. Climate
,
20
,
4751
4767
.
Deser
,
C.
,
R.
Tomas
,
M.
Alexander
, and
D.
Lawrence
,
2010
:
The seasonal atmospheric response to projected arctic sea ice loss in the late twenty-first century
.
J. Climate
,
23
,
333
351
.
Dethloff
,
K.
, and
Coauthors
,
2006
:
A dynamical link between the Arctic and the global climate system
.
Geophys. Res. Lett.
,
33
,
L03703
,
doi:10.1029/2005GL025245
.
Francis
,
J. A.
, and
S. J.
Vavrus
,
2012
:
Evidence linking Arctic amplification to extreme weather in mid-latitudes
.
Geophys. Res. Lett.
,
39
,
L06801
,
doi:10.1029/2012GL051000
.
Francis
,
J. A.
,
W.
Chan
,
D. J.
Leathers
, and
J. R.
Miller
,
2009
:
Winter Northern Hemisphere weather patterns remember summer arctic sea-ice extent
.
Geophys. Res. Lett.
,
36
,
L07503
,
doi:10.1029/2009GL037274
.
Honda
,
M.
,
J.
Inous
, and
S.
Yamane
,
2009
:
Influence of low Arctic sea-ice minima on anomalously cold Eurasian winters
.
Geophys. Res. Lett.
,
36
,
L08707
,
doi:10.1029/2008GL037079
.
Hopsch
,
S.
,
J.
Cohen
, and
K.
Dethloff
,
2012
:
Analysis of a link between fall Arctic sea ice concentration and atmospheric patterns in the following winter
.
Tellus
,
64A
,
18624
,
doi:10.3402/tellusa.v64i0.18624
.
Jaiser
,
R.
,
K.
Dethloff
,
D.
Handorf
,
A.
Rinke
, and
J.
Cohen
,
2012
:
Impact of sea ice cover changes on the Northern Hemisphere atmospheric winter circulation
.
Tellus
,
64A
,
11595
,
doi:10.3402/tellusa.v64i0.11595
.
Jeong
,
J.-H.
,
T.
Ou
,
H. W.
Linderholm
, and
B.-M.
Kim
,
2011
:
Recent recovery of the Siberian high intensity
.
J. Geophys. Res.
,
116
,
D23102
,
doi:10.1029/2011JD015904
.
Li
,
S.
,
2004
:
Impact of northwest Atlantic SST anomalies on the circulation over the Ural Mountains during early winter
.
J. Meteor. Soc. Japan
,
82
,
971
988
.
Livezey
,
R.
, and
W. Y.
Chen
,
1983
:
Statistical field significance and its determination by Monte Carlo techniques
.
Mon. Wea. Rev.
,
111
,
46
59
.
McVicar
,
T.
, and
Coauthors
,
2012
:
Global review and synthesis of trends in observed terrestrial near-surface wind speeds: Implications for evaporation
.
J. Hydrol.
,
416
,
182
205
.
Mysak
,
L.
, and
S.
Venegas
,
1998
:
Decadal climate oscillation in the Arctic: A new feedback loop for atmosphere–ice–ocean interactions
.
Geophys. Res. Lett.
,
25
,
3607
3610
.
Mysak
,
L.
,
D.
Manak
, and
R.
Marsden
,
1990
:
Sea-ice anomalies observed in the Greenland and Labrador Seas during 1901–1984 and their relation to an interdecadal Arctic climate cycle
.
Climate Dyn.
,
5
,
111
133
.
Onogi
,
K.
, and
Coauthors
,
2007
:
The JRA-25 Reanalysis
.
J. Meteor. Soc. Japan
,
85
,
369
432
.
Overland
,
J. E.
, and
M.
Wang
,
2010
:
Large-scale atmospheric circulation changes are associated with the recent loss of Arctic sea ice
.
Tellus
,
62A
,
1
9
, doi:10.1111/j.1600-0870.2009.00421.x.
Peng
,
S.
,
W. A.
Robinson
, and
S.
Li
,
2003
:
Mechanisms for the NAO response to the North Atlantic SST tripole
.
J. Climate
,
16
,
1987
2004
.
Petoukhov
,
V.
, and
V.
Semenov
,
2010
:
A link between reduced Barents–Kara sea ice and cold winter extremes over northern continents
.
J. Geophys. Res.
,
115
,
D21111
,
doi:10.1029/2009JD013568
.
Porter
,
D.
,
J.
Cassano
, and
M.
Serreze
,
2012
:
Local and large-scale atmospheric responses to reduced Arctic sea ice and ocean warming in the WRF model
.
J. Geophys. Res.
,
117
,
D11115
,
doi:10.1029/2011JD016969
.
Roeckner
,
E.
, and
Coauthors
,
2003
: The atmospheric general circulation model ECHAM5. Part I: Model description. Max Planck Institute for Meteorology Rep. 349, 127 pp. [Available from MPI for Meteorology, Bundesstr. 53, 20146 Hamburg, Germany.]
Screen
,
J. A.
, and
I.
Simmonds
,
2010
:
The central role of diminishing sea ice in recent Arctic temperature amplification
.
Nature
,
464
,
1334
1337
.
Tachibana
,
Y.
,
M.
Honda
, and
K.
Takeuchi
,
1996
:
The abrupt decrease of the sea ice over the southern part of the Sea of Okhotsk in 1989 and its relation to the recent weakening of the Aleutian low
.
J. Meteor. Soc. Japan
,
74
,
579
584
.
Tanaka
,
H. L.
,
R.
Kanohgi
, and
T.
Yasunari
,
1996
:
Recent abrupt intensification of the northern polar vortex since 1988
.
J. Meteor. Soc. Japan
,
74
,
947
954
.
Tang
,
Q. H.
,
X. J.
Zhang
,
X. H.
Yang
, and
J. A.
Francis
,
2013
:
Cold winter extremes in northern continents linked to Arctic sea ice loss
.
Environ. Res. Lett.
,
8
,
014036
,
doi:10.1088/1748-9326/8/1/014036
.
von Storch
,
H.
, and
F. W.
Zwiers
,
1999
: Statistical Analysis in Climate Research. Cambridge University Press, 494 pp.
Walsh
,
J.
,
W.
Chapman
, and
T.
Shy
,
1996
:
Recent decrease of sea level pressure in the central Arctic
.
J. Climate
,
9
,
480
486
.
Watanabe
,
M.
, and
T.
Nitta
,
1999
:
Decadal changes in the atmospheric circulation and associated surface climate variations in the Northern Hemisphere winter
.
J. Climate
,
12
,
494
510
.
Wever
,
N.
,
2012
:
Quantifying trends in surface roughness and the effect on surface wind speed observations
.
J. Geophys. Res.
,
117
,
D11104
,
doi:10.1029/2011JD017118
.
Wu
,
B.
,
R.
Huang
, and
D.
Gao
,
1999
:
Effects of variation of winter sea-ice area in Kara and Barents Seas on East Asian winter monsoon
.
Acta Meteor. Sin.
,
13
,
141
153
.
Wu
,
B.
,
Z.
Su
, and
R.
Zhang
,
2011
:
Effects of autumn–winter Arctic sea ice on winter Siberian high
.
Chin. Sci. Bull.
,
30
,
3220
3228
,
doi:10.1007/s11434-011-4696-4
.
Wu
,
B.
,
J.
Overland
, and
R.
D'Arrigo
,
2012
:
Anomalous Arctic surface wind patterns and their impacts on September sea ice minima and trend
.
Tellus
,
64A
,
18590
,
doi:10.3402/tellusa.v64i0.18590
.