1. Introduction
Numerous studies have made unremitting explorations on the physical characteristics of Arctic sea ice and their impacts on weather events and climate variations. Initially, numerical simulation experiments demonstrated that anomalies in Arctic sea ice have impacts on atmospheric circulation, resulting in a moderate cooling in mid- and high latitudes (Fletcher 1968; Newson 1973; Warshaw and Rapp 1973; Royer et al. 1990). Subsequently, some studies suggested that impacts of Arctic sea ice variations are comparatively weaker relative to the atmospheric internal variability (Alexander et al. 2004; Mori et al. 2014; Peings and Magnusdottir 2014), and certain literature has further questioned or even denied impacts of Arctic sea ice (Screen et al. 2013, 2014; Perlwitz et al. 2015; Ogawa et al. 2018; Blackport et al. 2019; Peings 2019; Galytska et al. 2023). Blackport et al. (2019) argue that only when decreases of sea ice are accompanied by abnormal heat transfer from the atmosphere to the ocean (atmosphere drives sea ice), can the model capture the major characteristics of observational connections (actually, this work does not separate the role of heat transportation from Arctic sea ice loss either). Therefore, they concluded that abnormal atmospheric circulation drives both midlatitude cold winters and mild Arctic winters, and the reduction of Arctic sea ice has the least impact on midlatitude severe winters. Recently, some studies once again demonstrated that Arctic sea ice variations do impact on the atmosphere circulation (Overland et al. 2021; Bailey et al. 2021; Ding et al. 2021; Smith et al. 2019, 2022; Outten et al. 2023).
The lack of consistency among different literatures can be attributed, on the one hand, to differences in used data, analysis methods, simulation experiment designs, and the researcher’s perspectives. On the other hand, the response of atmospheric circulation to Arctic sea ice anomalies is not solely determined by the direct impact of Arctic sea ice. It also closely depends on nonlinear processes (Honda et al. 2009; Petoukhov and Semenov 2010), Arctic sea ice conditions (Semenov and Latif 2015), background states of numerical models (Smith et al. 2019), and model atmospheric initial conditions (Wu et al. 2016, 2017; Yu and Wu 2023), as well as other influencing factors such as tropical forcing (Sato et al. 2014; Screen and Francis 2016; Warner et al. 2020; Rudeva and Simmonds 2021; Cohen et al. 2021). Consequently, it is impossible that the response of winter atmospheric circulation to Arctic sea ice melting to exhibits a specific spatial paradigm.
Previous studies have outlined the possible mechanisms by which Arctic sea ice anomalies affect midlatitude weather and climate, including tropospheric processes (Alexander et al. 2004; Deser et al. 2004) and stratospheric–tropospheric interaction processes (Jaiser et al. 2013; Kretschmer et al. 2018; McKenna et al. 2018; Zhang et al. 2018; Siew et al. 2020; Cohen et al. 2021; Overland et al. 2021; Zhang et al. 2022; Tian et al. 2023). However, both of these possible mechanisms still exhibit notable limitations. For instance, neither can fully account for the extreme heavy snowfall event in Europe from February to March 2018 or the alternating occurrences of warm Arctic–cold Eurasia and warm Arctic–warm Eurasia post 2004, while they are closely linked to Arctic sea ice anomalies (Bailey et al. 2021; Wu et al. 2022).
The stratospheric process can even influence the Siberian high, sea surface temperature (SST), and Arctic SICs (Zhang et al. 2018; Zhang et al. 2022), but there is still great uncertainty regarding the impact of Arctic sea ice melting on midlatitudes through the stratospheric pathway. McKenna et al. (2018) found that for both large-magnitude and moderate-magnitude Arctic sea ice melting, their impacts on the winter stratospheric polar vortex are contrasting. In the case of moderate-magnitude sea ice melting, opposite tropospheric Arctic Oscillation responses are observed. Conversely, for large-magnitude sea ice melting, the tropospheric response resembles a strong negative phase of the Arctic Oscillation. This suggests that as sea ice melting intensifies, the tropospheric mechanism becomes more important than the stratospheric process.
It is commonly understood that following a sudden warming of the stratosphere, cold and snowstorm weather often ensues in various parts of Europe and the United States. Regarding the severe cold and blizzard weather process that occurred in North America in February 2021, Cohen et al. (2021) contend that increases in snow cover over Eurasia and decreases in sea ice in the Barents–Kara Seas during autumn can impact the stretching deformation of the stratospheric polar vortex, consequently influencing midlatitudes and leading to extreme cold and snowstorm weather in North America. However, through simulation experiments, Davis et al. (2022) found that a sudden warming of the stratosphere has limited impact on surface temperature over the subsequent 4 weeks. They argue that the stratospheric warming that occurred in January 2021 is not related to the severe cold and blizzard weather in the United States in February of the same year. Instead, they suggest that the state of the tropospheric atmosphere is the primary factor behind the record low temperatures in North America in February 2021. Siew et al. (2020) also demonstrate that the stratospheric path connecting autumn Arctic sea ice and winter North Atlantic Oscillation is highly intermittent, with only 16% of observed samples showing a complete stratospheric path. The intermittency of this path aligns with the weak influence of Arctic sea ice melting on numerical simulation experiments, indicating that the role of Arctic sea ice melting in the Arctic–midlatitude connection is easily influenced by other factors. The latest research results show that the probability of sudden warming of the stratosphere primarily relies on the favorable combination of geomagnetic activity, solar activity, and quasi-biennial oscillation phases (Vokhmyanin et al. 2023), rather than the Ural blocking high anomaly and Arctic sea ice melting, further confirming the weak connection between Arctic sea ice anomalies and stratospheric processes.
While significant advancements have been achieved regarding the impact of Arctic sea ice anomalies on the midlatitudes, several critical questions still lack clarity. For instance, what are the shared characteristics in sea ice melting in the Barents–Kara Seas during autumn and winter seasons, and how has the evolution of sea ice in this sea region influenced short-term climate pulsation processes and interdecadal shifts in winter atmospheric circulation? Furthermore, how can we comprehend the ongoing melting of Arctic sea ice weakening winter temperature linkage between the Arctic and Eurasia? The motivation of this study is to explore these questions. Given significant uncertainty of the impact of the stratospheric process on the surface and the Siberian high, as mentioned earlier, this study specifically concentrates on the tropospheric process associated with the melting of Arctic sea ice. The results show that the interdecadal reduction of sea ice in the Barents–Kara Seas, which began in the early 2000s, has not only altered the thermal and moisture structure of the Arctic lower troposphere but has also amplified short-term pulsation processes of the Siberian high. This amplification has led to increased intensity of the Siberian high compared to the average before the winter of 2004/05. Furthermore, the evolution of sea ice in the Barents–Kara Seas can initiate different interdecadal shifts in the dominant modes of winter temperature variability. One such shift occurred in the context of the anomalous low Arctic sea ice post-2010, leading to a transition from warm Arctic–cold Eurasia to warm Arctic–warm Eurasia, which weakens the temperature linkage between the Arctic and Eurasia. These findings have important implications for understanding the roles of Arctic sea ice melting in the changing climate system.
2. Data and methods
The data used in this study include monthly Arctic sea ice concentration (SIC) data spanning from 1979 to 2022 on a 1° × 1° grid, acquired from Met Office Hadley Centre datasets (Rayner et al. 2003). Additionally, the global precipitation climatology project monthly precipitation dataset from 1979 to 2022 was obtained from National Centers for Environmental Information (Adler et al. 2018). Evaporation, sea level pressure (SLP), specific humidity, air temperature, and zonal winds were obtained from fifth major global reanalysis produced by (ERA5) dataset (Hersbach et al. 2020) from 1979 to 2022. The winter (DJF) Siberian high intensity is defined as the regionally averaged SLP over the domain 80°–120°E and 40°–60°N (Wu and Wang 2002).
To extract the predominant patterns of winter air temperature variability north of 20°N, both in ERA5 reanalysis data and in the ensemble means of simulation experiments, along with identifying the leading coupled pattern of regionally averaged SIC in the Barents–Kara Seas between autumn (SON) and the subsequent winter (DJF), the empirical orthogonal function (EOF; Lorenz 1956) analysis was conducted. Furthermore, the wavelet power spectrum analysis method (Torrence and Compo 1998) was also applied to identify dominant periods in the time series of regionally averaged SICs. To determine the transition points that occurred in the early 2000s, three different methods are used in this study, including the cumulative deviation time series, the running t test, and Mann–Kendall test (Wang et al. 2020). Additionally, the significance of comparisons between two periods is assessed through a Student’s t test.
To investigate possible impacts of the melting of Arctic sea ice on winter atmospheric circulation variations, simulation experiments were conducted using the Community Atmosphere Model, version 5.4 (CAM5.4). The horizontal resolution of CAM5.4 is 1.9° × 2.5°, with 30 levels in the vertical direction (Vertenstein et al. 2013). Table 1 summarizes the specific information of the control runs and simulation experiments.
Configurations of the control runs and simulation experiments.
3. Observational analyses
a. The melting of sea ice in the Barents–Kara Seas has enhanced the Arctic hydrological cycle
In autumn and winter, the regionally averaged SICs in the Barents–Kara Seas show decline trends (both trends are at 95% confidence level) (Figs. 1a,b). To extract predominant characteristics of the coupling pattern between autumn and the subsequent winter regionally averaged SICs in the Barents–Kara Seas (Fig. 1b), EOF analysis is carried out. It is found that the leading coupling pattern accounts for the 87.5% of the covariance, and corresponding spatial loadings are 0.71 and 0.71. A striking common feature of the sea ice evolutions in both seasons is the same interdecadal shift that occurred in the early of this century, transitioning from more sea ice stage before the winter of 2003/04 to less sea ice phase afterward (Fig. 1c). This result well agrees with the study of Close et al. (2015), and they have shown that the timing of onset of rapid decline in SICs in the Barents Sea is 2003/04. This interdecadal shift is completely consistent with the onset time of the Atlantification of the Arctic Ocean (Mohamed et al. 2022).
(a) Autumn (SON) mean SIC (%) differences between the mean averaged over 2004–21 and the mean averaged over 1979–2003; the red box indicates the Barents–Kara Seas (20.5°–90.5°E and 70.5°–84.5°N), which is used to calculate the regionally averaged SICs in autumn and winter. The purple contours denote differences at 95% confidence level. (b) Interannual variations of autumn (SON, blue) and winter (DJF, purple) regionally averaged SICs in the Barents–Kara Seas [the red box in (a)], and corresponding dashed lines denote their trends. (c) PC1 time series of EOF1 of autumn and winter regionally averaged SICs shown in (b); the leading EOF accounts for 87.5% of the covariance. Blue dashed lines denote the means averaged over 1979–2003 and 2004–21, respectively.
Citation: Journal of Climate 38, 4; 10.1175/JCLI-D-24-0163.1
The hydrological cycle in the Barents–Kara Seas also shows a notable interdecadal strengthening phenomenon (Figs. 2a–c). Winter open water evaporation in the Barents–Kara Seas increased significantly (at 99% confidence level) after the winter of 2005/06 relative to the before, and their averages (standard deviations) in two phases are 781.41 (99.29) and 1048.18 (210.92) km3, respectively (Fig. 2a). Correspondingly, the regional averaged specific humidity at 925 hPa (Fig. 2b), precipitation (Fig. 2c), and snowfall (not shown) also showed an interdecadal increase after the winter of 2004/05 relative to the before. The mean winter precipitation averaged over the Barents–Kara Seas was increased by 35.72 mm in the latter phase (at 99% confidence level). This interdecadal shift is not limited to the Barents–Kara Seas, and an averaged hydrological cycle in a large domain of the Arctic, such as 70°–80°N, 30°E–180°–150°W, also exhibits very similar interdecadal evolution (not shown).
(a) Winter cumulative evaporation amount over the Barents–Kara Seas (the red box in Fig. 1a) (km3). Red dashed lines represent winter means averaged over 1979/80–2004/05 and 2005/06–2021/22, respectively. (b) Winter 925-hPa specific humidity averaged over the Barents–Kara Seas (g kg−1). (c) Observed winter precipitation averaged over the Barents–Kara Seas (mm). (d) Cumulative deviation time series of the winter Siberian high intensity (hPa). (e) Normalized time series of the winter Siberian high intensity. In (b), (c), and (e), two dashed lines denote winter means averaged over 1979/80–2003/04 and 2004/05–2021/22, respectively. Gray shading denotes winters from 2004/05 to 2021/22.
Citation: Journal of Climate 38, 4; 10.1175/JCLI-D-24-0163.1
At interannual time scales, the winter hydrological cycle in the Barents–Kara Seas exhibits a more pronounced correlation with the simultaneous winter SIC in the same region compared to the preceding autumn SIC (Table 2). In contrast, the winter Siberian high intensity shows a stronger association with the preceding autumn SIC in the Barents–Kara Seas, rather than with the simultaneous winter SIC in the same region. This result is corroborated by the simulation experiments that incorporate observed SIC data, as detailed in the subsequent section.
Correlation coefficients of SIC in the Barents–Kara Seas with winter atmospheric variables over the same region and Siberian high intensity. Asterisk (*) denotes the correlation coefficient at 95% confidence level.
b. Possible impacts of sea ice melting in the Barents–Kara Seas on midlatitudes
This interdecadal shift phenomenon in the Arctic inevitably extends southward into the midlatitudes, consequently impacting the winter Siberian high. Analysis reveals different phase evolution characteristics of the winter Siberian high (Figs. 2d,e). The cumulative deviations’ evolution indicates that the maximum and minimum respectively occurred in winters of 1985/86 and 2003/04, implying that the Siberian high was weakened during the winters of 1986/87–2003/04, followed by a strengthened phase thereafter (Fig. 2d). From 1986 to 2003, the winter Siberian high experienced a normal to weaker stage in 16 winters, with their standard deviations being less than 0.5, and only winters of 1995/96 and 1999/2000 exceeded 0.5. The mean spanning from 1986/87 to 2003/04 was −0.53, weaker than the mean averaged over winters from 2004/05 to 2021/22 (0.39) (Fig. 2e). Since 2004, with the rapid melting of sea ice in the Barents–Kara Seas, the strengthened and weakened Siberian high has alternately occurred, which is not entirely consistent with the interdecadal variation of the Arctic.
The winter atmospheric baroclinicity, characterized by differences in mean zonal winds between 300 and 850 hPa, is weak (strong) over high latitudes (mid- and low latitudes) (Fig. 3). Regionally averaged baroclinicity has also experienced an interdecadal shift since 2004, which are dynamically consistent with strengthening of the Siberian high (Fig. 2e), indicating more effective potential energy is transformed into kinetic energy in mid- and low latitudes.
(a) The vertical shear of winter zonal winds (m s−1) between 300 and 850 hPa, averaged from 1979/80 to 2021/22; the black box (90°–120°E and 35°–45°N) is chosen to assess regional variations of the vertical shear. (b) Interannual variations of the vertical shear averaged within the black box in Fig. 3a (m s−1); blue dashed lines denote winter means averaged over 1979/80–2003/04 and 2004/05–2021/22, respectively. The red dashed line is the cumulative deviation time series of the regionally averaged vertical shear.
Citation: Journal of Climate 38, 4; 10.1175/JCLI-D-24-0163.1
4. Simulation experiments forced by observed Arctic SICs
a. Enhancement of the Arctic hydrological cycle
In this section, all figures depicting the time evolution of ensemble means of simulated variables do not incorporate the spreads among different simulation experiments because the results derived from a single simulation experiment primarily reflect atmospheric internal variability and stochastic processes that dilute the impact of Arctic sea ice forcing. The evolution of SICs in the Barents–Kara Seas directly alters the thermal structure in the lower troposphere (Fig. 4). It is seen that more sea ice stage in this sea region directly reinforced the inversion layer in the lower troposphere because sea ice obstructs heat exchanges from the ocean to the atmosphere and strengthens the radiative cooling effect. Weakened inversion layers below both 850 and 925 hPa since 2004 completely coincides with the interdecadal reduction of sea ice in the Barents–Kara Seas. The average strength of the inversion layer below 850 hPa is reduced by 3.85°C, decreasing from a mean of 2.44° to −1.41°C (Fig. 4a, blue dashed lines). This implies that the continuous melting of sea ice in the Barents–Kara Seas is more conducive to enhancing the instability of atmospheric stratification and development of convective activity, which is further supported by the evolutions of the vertical profiles of temperature and specific humidity (Figs. 4b,c). In more sea ice stage, the strongest inversion layer appears between 925 and 1000 hPa, inhibiting the vertical moisture transport, resulting in the maximum specific humidity appearing at 925 hPa. In the rapid melting phase of sea ice, weakening of the temperature inversion layer is evident, and the difference in mean specific humidity between 925 and 1000 hPa becomes negligible.
(a) The blue (red) curve shows differences in ensemble means of simulated winter air temperature averaged over the Barents–Kara Seas (the red box in Fig. 1a) between 850 hPa (925 hPa) and the surface (former minus latter) (°C); blue (red) dashed lines denote winter means averaged over 1979/80–2003/04 and 2004/05–2016/17, respectively. Gray shading denotes winters from 2004/05 to 2016/17. (b) The blue (red) line shows the vertical profile of ensemble means of simulated the winter air temperature averaged over the Barents–Kara Seas (the red box in Fig. 1a) for winters of 1979/80–2003/04 (winters of 2004/05–2016/17) (°C). (c) As in (b), but for the specific humidity (g kg−1).
Citation: Journal of Climate 38, 4; 10.1175/JCLI-D-24-0163.1
This interdecadal shift in Arctic sea ice results in significant changes in both temperature and specific humidity throughout the troposphere and stratosphere over the Arctic (Fig. 5). Notable positive temperature anomalies are observed in most of the lower troposphere, except for the Atlantic sector of the Arctic, where significant positive temperature anomalies extend upward to 400 hPa. In the upper troposphere and the stratosphere, positive and negative temperature anomalies coexist (Fig. 5a). The largest positive specific humidity anomalies are observed around the Barents–Kara Seas. Accompanied by significant positive specific humidity anomalies covering most of the troposphere, negative specific humidity anomalies occupy the entire stratosphere and the upper troposphere (Fig. 5b). It seems that the interdecadal change in Arctic sea ice has a greater impact on humidity than temperature.
(a) The pressure–longitude cross section of the meridional average over 70°–80°N of differences in ensemble means of simulated winter mean air temperature between 1979/80–2003/04 and 2004/05–2016/17 (red contours, °C); the shading area denotes the differences at 95% confidence level. (b) Same as in (a), but for the specific humidity (g kg−1).
Citation: Journal of Climate 38, 4; 10.1175/JCLI-D-24-0163.1
In the Barents–Kara Seas, winter sea ice and evaporation are completely out of phase (Fig. 6a; their correlation coefficient is −0.98), and enhanced evaporation has been observed since the winter of 2004/05, consistent with the observed evaporation evolution (Fig. 2a). Simulated winter precipitation (Fig. 6b), snowfall (not shown), and surface temperature (not shown) in the Barents–Kara Seas clearly exhibit the identical interdecadal shift since the winter of 2004/05, consistent with the observed interdecadal shift. Compared with the impact of sea ice melting on evaporation and precipitation, its impact on 500-hPa heights is more complicated (Fig. 6c). It is seen that there are only three winters where simulated 500-hPa heights exceeded 5110 gpm, and all of them occurred after the winter of 2003/04, indicating that the intensity and frequency of positive anomalous fluctuations are increasing in the context of Arctic sea ice melting. Meanwhile, we also note that in the context of Arctic sea ice melting, the simulated response in 500-hPa heights is weak; for example, in the winter of 2011/12, the response is notably lower than the average before the winter of 2003/04. In autumn 2016, SIC in the Barents–Kara Seas was only higher than that in autumn 2012 and 2020, and in the winter of 2016/17, the corresponding SIC in this region reached its minimum (Fig. 1a). However, the simulated height response was obviously lower, even lower than that in many winter cases prior to 2004. Therefore, under the context of persistently low Arctic sea ice, the individual response of the atmospheric height field exhibits great uncertainties and is not solely determined by the anomaly of sea ice.
(a) The blue and green lines represent ensemble means of simulated winter evaporation (mm) averaged over the Barents–Kara Seas (the red box in Fig. 1a) and observed winter SIC (%) averaged in the Barents–Kara Seas, respectively. (b) Ensemble means of simulated winter precipitation averaged over the Barents–Kara Seas (mm); dashed lines represent the means over winters of 1979/80–2003/04 and winters of 2004/05–2016/17, respectively. (c) As in (b), but for simulated winter 500-hPa geopotential heights (gpm). (d) Ensemble means (purple) of the simulated winter Siberian high intensity and the observed winter Siberian high intensity (red); purple dashed lines denote the means of ensemble means averaged over 1987/88–2003/04 and 2004/05–2016/17, respectively. (e) Cumulative deviations of ensemble means of the simulated winter Siberian high intensity (black) and the observed winter Siberian high intensity (brown). Gray shading represents winters from 2004/05 to 2016/17.
Citation: Journal of Climate 38, 4; 10.1175/JCLI-D-24-0163.1
b. Strengthening of climatic pulsations in midlatitudes
Figure 6d shows the evolution of ensemble means of the simulated winter Siberian high intensity, along with its comparison with reanalysis data. Although the response of the Siberian high to the continuous melting of Arctic sea ice is noticeably weaker than that in the reanalysis data, major features of the phase evolution are clear. Thickness of Arctic sea ice is prescribed as 2 m in CAM5, and thus the model underestimates the heat flux through thin sea ice in the marginal seas of the Arctic Ocean, which is one of the reasons for the underestimated Siberian high. The average values of the Siberian high intensity during the winters of 1987/88–2003/04 and 2004/05–2016/17 are 1031.93 and 1032.37 hPa, respectively, with a difference of 0.44 hPa between their averages (at 99% confidence level). Additionally, their standard deviations in the two stages are 0.23 and 0.29 hPa, respectively, indicating an enhanced variability in the latter phase. It is striking that the cumulative deviations for the simulated and observed Siberian high intensity exhibit very similar evolution features, with their maximums in the mid-1980s and their minimums occurring simultaneously in the winter of 2003/04 (Fig. 6e). This means that the melting of Arctic sea ice plays a crucial regulating role in the interdecadal climate shift in midlatitudes of Eurasia.
Figure 6d clearly shows that since the winter of 2004/05, the response of the Siberian high has exhibited frequent and irregular pulsating processes. During the trough phase, the Siberian high intensity even falls below that of individual cases before 2003. In September 2012, the Arctic sea ice extent reached the lowest value on the record, and in the ensuing autumn and winter seasons of the same year, the average SICs in the Barents–Kara Seas also reached their lowest values before 2012 (Fig. 1a). However, the simulated Siberian high in the winter of 2012/13 was even weaker than some winters prior to 2003. Therefore, although Arctic sea ice melting has enhanced climate pulsating processes in midlatitudes of Eurasia, it is not possible to predict when these pulsations will reach their peak or trough in midlatitudes.
Further analysis shows that the simulated winter Siberian high intensity has a stronger relationship with the preceding autumn SIC in the Barents–Kara Seas compared to the simultaneous winter SIC in the same region. The correlation coefficients are −0.29 (at 90% confidence level) for autumn SIC and −0.20 for winter SIC, respectively. Consequently, observational analyses (Table 2) and simulation experiments consistently demonstrate that autumn sea ice in the Barents–Kara Seas plays a more significant role in influencing the subsequent winter Siberian high than the simultaneous winter sea ice. Thus, these results do not support the conclusion of Blackport and Screen (2019), who suggested that the winter atmospheric response is mainly due to winter sea ice forcing rather than autumn sea ice. As to the mechanism for the association between autumn SICs and the winter Siberian high, some studies have explored this question (e.g., Deser et al. 2004; Honda et al. 2009; Wu et al. 2011; Jaiser et al. 2013; Cohen et al. 2021), and this study will not delve further into the matter.
Figure 7 shows a comparison of the differences in phasic means between the reanalysis data and ensemble means. In the rapid melting stage of Arctic sea ice, SLP increases in the northern North Atlantic, the central and northern Eurasia, and the East Asian coast relative to that in winters from 1987/88 to 2003/04, and the difference in SLP between mid- and high latitudes of Eurasia exceeds 99% confidence level (Fig. 7a). Major differences between the simulated response and the reanalysis data are found in mid- and high latitudes of East Asia, the North Pacific, and the northern region of the North American continent where negative SLP anomalies in the response are replaced by positive anomalies in the reanalysis data (Fig. 7b). The cause of these differences is very likely to be the impact of the prescribed climatological SST forcing in simulation experiments, which artificially ignored the impacts of SSTs in the North Pacific and the tropical Pacific on winter SLP over the northern North Pacific (Kosaka and Xie 2013; Ding et al. 2014; Screen and Francis 2016). The 1000-hPa temperature differences between the two phases in both simulations and the reanalysis data consistently show a warm Arctic–cold Eurasia structure (Figs. 7c,d). It is seen that the simulated response to Arctic sea ice melting hardly affects Tibetan Plateau.
Differences in ensemble means of simulated winter mean (a) SLP and (c) 1000-hPa temperature between 1987/88–2003/04 and 2004/05–2016/17 (latter minus former); purple contours represent differences at 95% confidence level. (b), (d) As in (a) and (c), respectively, but for differences derived from ERA5 reanalysis data.
Citation: Journal of Climate 38, 4; 10.1175/JCLI-D-24-0163.1
The atmospheric baroclinicity not only reflects the potential conversion ability between effective potential energy and kinetic energy but also describes the conduciveness of atmospheric circulation to synoptic-scale development. Arctic sea ice melting has a direct impact on the baroclinicity evolution of the atmospheric circulation over Eurasia (Fig. 8). It is seen that the atmospheric baroclinity is positive over Eurasia and the northwestern Pacific sector, with a belt region in the mid- and low latitudes showing strong baroclinity (Fig. 8a). Accompanied by the evolution of Arctic sea ice, the atmospheric baroclinity also has experienced different interdecadal shifts in high latitudes and mid- and low latitudes (Figs. 8b,c). In high latitudes, there was an interdecadal weakening from a high-baroclinicity phase (1987/88–2003/04) to a low-baroclinicity stage (2004/05–2016/17), with a reverse evolution in mid- and low latitudes. The weakened baroclinicity over high latitudes corresponds to a reduction in the conversion of atmospheric effective potential energy to kinetic energy. In mid- and low latitudes of East Asia and the northwest Pacific Ocean, the strengthened baroclinicity is dynamically coincident with an interdecadal enhancement of the Siberian high. The identical interdecadal shift observed in the simulated response and reanalysis data (Figs. 3b and 8b,c) demonstrates that Arctic sea ice melting plays an important role in producing interdecadal fluctuations.
The vertical shear of winter zonal winds (m s−1) of ensemble means between 300 and 850 hPa; the black boxes (30°–120°E and 55°–65°N, 120°–160°E and 25°–40°N) are chosen to calculate regional variations of the vertical shear. (b) Interannual variations of the vertical shear averaged within 30°–120°E and 55°–65°N, derived from ensemble means. (c) As in (b), but for the region (120°–160°E and 25°–40°N). In (b) and (c), dashed lines denote the phasic means averaged over 1987/88–2003/04 and 2004/05–2016/17, respectively.
Citation: Journal of Climate 38, 4; 10.1175/JCLI-D-24-0163.1
c. Triggering distinct interdecadal shifts
An EOF analysis of ensemble means of simulated winter 1000-hPa temperature is conducted to explore impacts of Arctic sea ice melting on winter temperature variations. The leading EOF reflects a warm Arctic–cold Eurasia temperature configuration, with a clear interdecadal shift toward more frequent positive anomalies since 2004/05 (Figs. 9a,b). It is seen that the negative phase prevailed during winters from 1987/88 to 2003/04, when the Arctic was abnormally cold, while mid- and high latitudes of Eurasia (especially East Asia) were abnormally warm. After 2004, the positive phase frequently occurred, which was completely consistent with the phase transition of Arctic SICs, once again highlighting that Arctic sea ice melting is one of the primary causes for this interdecadal shift in the leading mode of winter temperature variations. Compared to the corresponding EOF1 derived from the reanalysis data shown in Figs. 9c and 9d, simulation experiments effectively capture the key features of the observed connection, particularly the identical interdecadal shift that reinforces warm Arctic–cold Asia pattern. The extent and amplitude of negative anomalies shown in Fig. 9c are evidently smaller and weaker relative to those in Fig. 9a, implying the presence of other factors that mitigate the impact of Arctic sea ice melting.
(a) The spatial distribution of EOF1 of normalized ensemble means of simulated winter 1000-hPa temperature north of 20°N and (b) its corresponding PC1 time series. In (b), blue dashed lines denote the phasic means averaged over 1987/88–2003/04 and 2004/05–2016/17, respectively. The leading EOF accounts for 26.4% of the variance. (c), (d) As in (a) and (b), respectively, but derived from ERA5 reanalysis data, and it accounts for 20.0% of the variance. In (d), the purple dashed lines represent the phasic means averaged over 1979/80–1998/99, 1999/2000–2003/04, and 2004/05–2021/22, respectively.
Citation: Journal of Climate 38, 4; 10.1175/JCLI-D-24-0163.1
Compared to EOF1 (Fig. 9), EOF2 of ensemble means of simulated winter 1000-hPa temperature north of 20°N displays distinct spatiotemporal characteristics (Fig. 10). Spatially, temperature anomalies in mid- and high latitudes of Eurasia are in phase with those in the Arctic Ocean and the Siberian marginal seas. In terms of temporal evolution, the positive phase was prevalent during winters from 1994/95 to 2010/11, followed by dominantly negative phases. Consequently, warm Arctic and warm Eurasia coexisted during winters of 2011/12–2016/17. Simulated winter 1000-hPa temperature anomalies averaged over winters of 2011/12–2016/17 clearly demonstrate the occurrence of the warm Arctic–warm Eurasia pattern (see Fig. 4d of Wu et al. 2022), resulting in a weakened temperature linkage between the Arctic and Asia. Comparing to Fig. 9, it is evident that the interdecadal transition in EOF2 significantly contributes to the weakening of the winter temperature linkage between the Arctic and Asia. In reality, EOF2 of winter 1000-hPa temperature variations, extracted from the reanalysis data, shows positive anomalies over mid- and high latitudes of Eurasia and most of the Siberian marginal seas (Fig. 10c), with frequently positive phases prevailed during winters of 2013/14–2021/22 (Fig. 10d). Temperature anomalies averaged over winters from 2013/14 to 2021/22 exhibit warm Arctic–warm Eurasia configuration (Fig. 11), essentially confirming the result of simulation experiments. Consequently, the melting of Arctic sea ice has triggered distinct interdecadal shifts since 2000. The alternating occurrence of the warm Arctic–cold Eurasia and warm Arctic–warm Eurasia patterns would be expected in the context of Arctic sea ice melting.
(a), (b) As in Figs. 9a and 9b, respectively, but for EOF2, accounting for 14.0% of the variance. In (b), blue dashed lines represent the phasic means averaged over three phases: 1981/82–1993/94, 1994/95–2010/11, and 2011/12–2016/17. (c), (d) As in Figs. 9c and 9d, respectively, but for EOF2. In (d), the black dashed line represents 0.0, and the red dashed line denotes a 3-yr running mean of PC2 time series. EOF2 accounts for 13.6% of the variance.
Citation: Journal of Climate 38, 4; 10.1175/JCLI-D-24-0163.1
Winter 1000-hPa temperature anomalies averaged over 2013/14–2021/22, relative to the mean over 1979/80–2021/22. Data are derived from ERA5 reanalysis data.
Citation: Journal of Climate 38, 4; 10.1175/JCLI-D-24-0163.1
5. Discussion of possible mechanisms
a. Enhanced climatic pulsation processes in midlatitudes
In addition to the direct dynamic forcing of ocean currents and surface wind fields on sea ice, the evolution of sea ice primarily exhibits slow-varying processes, while the atmospheric evolution consists of more stochastic fast-varying processes, and their interactions are inevitably complex and variable. For such highly complex processes, we discuss the possible causes of climate fluctuations through highly idealized processes. We hypothesized that in midlatitudes, the atmospheric response to Arctic sea ice melting is considered an interaction process between the atmospheric internal intrinsic variation and the forced variation induced by Arctic sea ice melting.
The wavelet power spectrum analysis of autumn SICs averaged in the Barents–Kara Seas, and the black dots denote spectrum values at 95% confidence level.
Citation: Journal of Climate 38, 4; 10.1175/JCLI-D-24-0163.1
Thus, the atmospheric response can be written as Cos(ωi × t + θ) × Cos(ωa × t + ϕ).
Figures 13a–c show different evolution processes of the forced variation, the atmospheric internal intrinsic variation, and their interaction process. This may imply that short-term climatic pulsations are caused by an interaction process between long-period oscillations of the forced variation and the atmospheric internal intrinsic variation. This study only provides an idealized scenario, but the conclusion is enlightening.
Schematic diagram showing a highly idealized possible interaction to interpret the generation of climatic pulsations induced by Arctic sea ice melting. (a) An assumed atmospheric response with a 7-yr cycle induced by Arctic sea ice melting, x-axis coordinate is time (year). (b) An atmospheric internal variation with a 3-yr cycle. (c) An interaction between the atmospheric response in (a) and the atmospheric internal variation in (b) produces climatic pulsations. (d) The interaction process between the atmospheric response with a 4-yr cycle and the atmospheric internal variation with a 3-yr cycle produces an oscillation with a 12-yr cycle.
Citation: Journal of Climate 38, 4; 10.1175/JCLI-D-24-0163.1
b. A recent interdecadal shift
The wavelet power spectrum analysis of autumn SICs in the Barents–Kara Sea reveals a trend wherein the dominant periods have been gradually shortened since 1979. Specifically, the analysis shows a transition from a 7-yr period during the early 1980s to approximately a 4-yr period by 2008 (Fig. 12). When the period of the forced variation is reduced to 4 yr, approaching the period of the atmospheric internal intrinsic variation, their interaction process produces a clear decadal oscillation (Fig. 13d). Although the processes and mechanisms by which anomalously low Arctic sea ice causes interdecadal shifts remain unclear, Figs. 12 and 13 imply that the interdecadal shift in winter temperature variations in the background of abnormally low Arctic sea ice, as shown in Figs. 10 and 11, may be attributed to the shortened cycle period of sea ice melting. Therefore, it is very likely that the phasic response of winter atmospheric circulation to the melting of Arctic sea ice is to be determined by their different prevalent cycle periods.
6. Discussion and conclusions
Using Arctic SIC data, atmospheric reanalysis data, and numerical simulation experiments forced by observed Arctic SIC data, this paper investigates some major features of the evolution of SICs averaged in the Barents–Kara Seas during autumn and the subsequent winter, as well as their impacts on winter atmospheric circulation variations. It is found that both autumn and the subsequent winter sea ice in the Barents–Kara Seas underwent a coincident interdecadal shift in the early 2000s, entering a phase of rapid melting since 2004. As a direct result of the interdecadal reduction of sea ice, the atmospheric stratification in the lower troposphere over the Barents–Kara Seas has undergone a significant change. The winter inversion in the lower troposphere over the sea region has significantly weakened or even disappeared. The enhancement of evaporation and the diffusion of water vapor in the Arctic troposphere have led to increases in atmospheric humidity and precipitation, strengthening Arctic winter hydrological cycle. The impact of sea ice melting on Arctic winter atmospheric humidity in the troposphere is more important than air temperature.
The accelerated melting of Arctic sea ice since 2004 has strengthened short-term pulsation processes of the Siberian high, making the mean intensity of the Siberian high stronger relative to the average over winters of 1979/80–2003/04. The melting of Arctic sea ice has dynamically weakened the atmospheric baroclinicity in high latitudes of Eurasia and strengthened the atmospheric baroclinicity in mid- and low latitudes of East Asia and the northwest Pacific.
More importantly, the continuously low level of Arctic sea ice has triggered a recent interdecadal transition in the second pattern of winter 1000-hPa temperature variations since the early 2010s, which has produced a phasic warm Arctic–warm Eurasia temperature anomalous pattern that weakens the temperature linkage between the Arctic and mid- and high latitudes of Eurasia. Accompanied by shortening of dominant periods in autumn sea ice melting in the Barents–Kara Seas, it may be anticipated that interdecadal transitions will become more frequent in the future.
The observed evolution of Arctic sea ice is the result of atmosphere–ocean–ice interactions under the background of global warming. However, this study directly uses observed Arctic sea ice as an external forcing to drive the atmospheric circulation model (CAM5.4), which has its limitations in methodology. On the one hand, this design may exaggerate the impact of Arctic sea ice melting. On the other hand, without atmosphere–ocean–ice interactions, the results here may underestimate the impact of Arctic sea ice melting (Deser et al. 2016; Jenkins and Dai 2021). Additionally, observed Arctic sea ice–forced simulation experiments may make it harder to interpret the causality underlying observational associations between Arctic sea ice and atmospheric variability because some sea ice signals may be regarded as a response to atmospheric forcing (Ding et al. 2022). Consequently, it is an urgent need to design a more comprehensive approach to assess the impact of Arctic sea ice evolution on the winter atmospheric circulation variability. Moreover, a more comprehensive exploration of the mechanisms for distinct interdecadal transitions, including the role of SSTs, is beyond the scope of this study but will be pursued in future work.
Additionally, comparison simulation experiments were conducted with observed monthly SICs from 1979 to 2021 only within the Barents–Kara Seas (see the red box in Fig. 1a), and SICs in the remaining sea regions and global SSTs were prescribed as their climatological values. This simulation experiment was repeated 30 times, each with different model initial conditions. The results indicate that observed SIC forcing confined to the Barents–Kara Seas was insufficient to reproduce the major features of phasic and interdecadal transitions similar to the observational associations (see the supplemental information in the online supplemental material). This implies that when investigating the impacts of Arctic sea ice melting on interdecadal variations of winter atmospheric circulation, we should consider the evolution of the entire Arctic sea ice, rather than just focusing the sea ice in the Barents–Kara Seas.
Acknowledgments.
This study was supported by the National Natural Science Foundation of China (42375023), the National Key Research and Development Project of China (2022YFF0801701), the Key Program of National Natural Science Foundation of China (Grant 41730959), and the program of CAMS (2015CB453202).
Data availability statement.
European Centre for Medium-Range Weather Forecasts Reanalysis 5 (ERA5) dataset is available at https://cds.climate.copernicus.edu/cdsapp#!/search?type=dataset/. Met Office Hadley Centre datasets are available at https://hadleyserver.metoffice.gov.uk/hadisst. The global precipitation climatology project monthly precipitation dataset is available https://www.ncei.noaa.gov/data/.
REFERENCES
Adler, R. F., and Coauthors, 2018: The Global Precipitation Climatology Project (GPCP) monthly analysis (new version 2.3) and a review of 2017 global precipitation. Atmosphere, 9, 138, https://doi.org/10.3390/atmos9040138.
Alexander, M. A., U. S. Bhatt, J. E. Walsh, M. S. Timlin, J. S. Miller, and J. D. Scott, 2004: The atmospheric response to realistic Arctic sea ice anomalies in an AGCM during winter. J. Climate, 17, 890–905, https://doi.org/10.1175/1520-0442(2004)017<0890:TARTRA>2.0.CO;2.
Bailey, H., A. Hubbard, E. S. Klein, K.-R. Mustonen, P. D. Akers, H. Marttila, and J. M. Welker, 2021: Arctic sea-ice loss fuels extreme European snowfall. Nat. Geosci., 14, 283–288, https://doi.org/10.1038/s41561-021-00719-y.
Blackport, R., and J. A. Screen, 2019: Influence of Arctic sea ice loss in autumn compared to that in winter on the atmospheric circulation. Geophys. Res. Lett., 46, 2213–2221, https://doi.org/10.1029/2018GL081469.
Blackport, R., J. A. Screen, K. van der Wiel, and R. Bintanja, 2019: Minimal influence of reduced Arctic sea ice on coincident cold winters in mid-latitudes. Nat. Climate Change, 9, 697–704, https://doi.org/10.1038/s41558-019-0551-4.
Cohen, J., L. Agel, M. Barlow, C. I. Garfinkel, and I. White, 2021: Linking Arctic variability and change with extreme winter weather in the United States. Science, 373, 1116–1121, https://doi.org/10.1126/science.abi9167.
Close, S., M.-N. Houssais, and C. Herbaut, 2015: Regional dependence in the timing of onset of rapid decline in Arctic sea ice concentration. J. Geophys. Res. Oceans, 120, 8077–8098, https://doi.org/10.1002/2015JC011187.
Davis, N. A., J. H. Richter, A. A. Glanville, J. Edwards, and E. LaJoie, 2022: Limited surface impacts of the January 2021 sudden stratospheric warming. Nat. Commun., 13, 1136, https://doi.org/10.1038/s41467-022-28836-1.
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, https://doi.org/10.1175/1520-0442(2004)017<0877:TEONAS>2.0.CO;2.
Deser, C., L. Sun, R. A. Tomas, and J. Screen, 2016: Does ocean coupling matter for the northern extratropical response to projected Arctic sea ice loss? Geophys. Res. Lett., 43, 2149–2157, https://doi.org/10.1002/2016GL067792.
Ding, Q., J. M. Wallace, D. S. Battisti, E. J. Steig, A. J. E. Gallant, H.-J. Kim, and L. Geng, 2014: Tropical forcing of the recent rapid Arctic warming in northeastern Canada and Greenland. Nature, 509, 209–212, https://doi.org/10.1038/nature13260.
Ding, Q., A. Schweiger, and I. Baxter, 2022: Nudging observed winds in the Arctic to quantify associated sea ice loss from 1979 to 2020. J. Climate, 35, 6797–6813, https://doi.org/10.1175/JCLI-D-21-0893.1.
Ding, S., B. Wu, and W. Chen, 2021: Dominant characteristics of early Autumn Arctic sea ice variability and its impact on winter Eurasian climate. J. Climate, 34, 1825–1846, https://doi.org/10.1175/JCLI-D-19-0834.1.
Fletcher, J. O., 1968: The influence of the Arctic pack ice on climate. Causes of Climatic Change, Meteor. Monogr., No. 30, Amer. Meteor. Soc., 93–99, https://doi.org/10.1007/978-1-935704-38-6_7.
Galytska, E., K. Keigel, D. Handorf, R. Jaiser, R. Köhler, J. Runge, and V. Eyring, 2023: Evaluating causal Arctic-midlatitude teleconnections in CMIP6. J. Geophys. Res. Atmos., 128, e2022JD037978, https://doi.org/10.1029/2022JD037978.
Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 1999–2049, https://doi.org/10.1002/qj.3803.
Honda, M., J. Inoue, and S. Yamane, 2009: Influence of low Arctic sea-ice minima on anomalously cold Eurasian winters. Geophys. Res. Lett., 36, L08707, https://doi.org/10.1029/2008GL037079.
Jaiser, R., K. Dethloff, and D. Handorf, 2013: Stratospheric response to Arctic sea ice retreat and associated planetary wave propagation changes. Tellus, 65A, 19375, https://doi.org/10.3402/tellusa.v65i0.19375.
Jenkins, M., and A. Dai, 2021: The impact of sea-ice loss on Arctic climate feedbacks and their role for Arctic amplification. Geophys. Res. Lett., 48, e2021GL094599, https://doi.org/10.1029/2021GL094599.
Kosaka, Y., and S.-P. Xie, 2013: Recent global-warming hiatus tied to equatorial Pacific surface cooling. Nature, 501, 403–407, https://doi.org/10.1038/nature12534.
Kretschmer, M., D. Coumou, L. Agel, M. Barlow, E. Tziperman, and J. Cohen, 2018: More-persistent weak stratospheric polar vortex states linked to cold extremes. Bull. Amer. Meteor. Soc., 99, 49–60, https://doi.org/10.1175/BAMS-D-16-0259.1.
Lorenz, E. N., 1956: Empirical orthogonal functions and statistical weather prediction. MIT Department Meteorology, Statistical Forecasting Project Science Rep. 1, 48 pp., https://hogback.atmos.colostate.edu/group/dave/pdf/EOFs.pdf.
McKenna, C. M., T. J. Bracegirdle, E. F. Shuckburgh, P. H. Haynes, and M. M. Joshi, 2018: Arctic sea ice loss in different regions leads to contrasting Northern Hemisphere impacts. Geophys. Res. Lett., 45, 945–954, https://doi.org/10.1002/2017GL076433.
Mohamed, B., F. Nilsen, and R. Skogseth, 2022: Marine heatwaves characteristics in the Barents Sea based on high resolution satellite data (1982–2020). Front. Mar. Sci., 9, 821646, https://doi.org/10.3389/fmars.2022.821646.
Mori, M., M. Watanabe, H. Shiogama, J. Inoue, and M. Kimoto, 2014: Robust Arctic sea-ice influence on the frequent Eurasian cold winters in past decades. Nat. Geosci., 7, 869–873, https://doi.org/10.1038/ngeo2277.
Newson, R. L., 1973: Response of a general circulation model of the atmosphere to removal of the Arctic ice-cap. Nature, 241, 39–40, https://doi.org/10.1038/241039b0.
Ogawa, F., and Coauthors, 2018: Evaluating impacts of recent Arctic sea ice loss on the Northern Hemisphere winter climate change. Geophys. Res. Lett., 45, 3255–3263, https://doi.org/10.1002/2017GL076502.
Outten, S., and Coauthors, 2023: Reconciling conflicting evidence for the cause of the observed early 21st century Eurasian cooling. Wea. Climate Dyn., 4, 95–114, https://doi.org/10.5194/wcd-4-95-2023.
Overland, J. E., and Coauthors, 2021: How do intermittency and simultaneous processes obfuscate the Arctic influence on midlatitude winter extreme weather events? Environ. Res. Lett., 16, 043002, https://doi.org/10.1088/1748-9326/abdb5d.
Peings, Y., 2019: Ural blocking as a driver of early-winter stratospheric warmings. Geophys. Res. Lett., 46, 5460–5468, https://doi.org/10.1029/2019GL082097.
Peings, Y., and G. Magnusdottir, 2014: Response of the wintertime Northern Hemisphere atmospheric circulation to current and projected Arctic sea ice decline: A numerical study with CAM5. J. Climate, 27, 244–264, https://doi.org/10.1175/JCLI-D-13-00272.1.
Perlwitz, J., M. Hoerling, and R. Dole, 2015: Arctic tropospheric warming: Causes and linkages to lower latitudes. J. Climate, 28, 2154–2167, https://doi.org/10.1175/JCLI-D-14-00095.1.
Petoukhov, V., and V. A. Semenov, 2010: A link between reduced Barents-Kara sea ice and cold winter extremes over northern continents. J. Geophys. Res., 115, D21111, https://doi.org/10.1029/2009JD013568.
Rayner, N. A., D. E. Parker, E. B. Horton, C. K. Folland, L. V. Alexander, D. P. Rowell, E. C. Kent, and A. Kaplan, 2003: Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res., 108, 4407, https://doi.org/10.1029/2002JD002670.
Royer, J. F., S. Planton, and M. Deque, 1990: A sensitivity experiment for the removal of Arctic sea ice with the French spectral general circulation model. Climate Dyn., 5 (1), 1–17, https://doi.org/10.1007/BF00195850.
Rudeva, I., and I. Simmonds, 2021: Midlatitude winter extreme temperature events and connections with anomalies in the Arctic and tropics. J. Climate, 34, 3733–3749, https://doi.org/10.1175/JCLI-D-20-0371.1.
Sato, K., J. Inoue, and M. Watanabe, 2014: Influence of the Gulf Stream on the Barents Sea ice retreat and Eurasian coldness during early winter. Environ. Res. Lett., 9, 084009, https://doi.org/10.1088/1748-9326/9/8/084009.
Screen, J. A., and J. A. Francis, 2016: Contribution of sea-ice loss to Arctic amplification is regulated by Pacific Ocean decadal variability. Nat. Climate Change, 6, 856–860, https://doi.org/10.1038/nclimate3011.
Screen, J. A., I. Simmonds, C. Deser, and R. Tomas, 2013: The Atmospheric response to three decades of observed Arctic sea ice loss. J. Climate, 26, 1230–1248, https://doi.org/10.1175/JCLI-D-12-00063.1.
Screen, J. A., C. Deser, I. Simmonds, and R. Tomas, 2014: Atmospheric impacts of Arctic sea-ice loss, 1979–2009: Separating forced change from atmospheric internal variability. Climate Dyn., 43, 333–344, https://doi.org/10.1007/s00382-013-1830-9.
Semenov, V. A., and M. Latif, 2015: Nonlinear winter atmospheric circulation response to Arctic sea ice concentration anomalies for different periods during 1966–2012. Environ. Res. Lett., 10, 054020, https://doi.org/10.1088/1748-9326/10/5/054020.
Siew, P. Y. F., C. Li, S. P. Sobolowski, and M. P. King, 2020: Intermittency of Arctic–mid-latitude teleconnections: Stratospheric pathway between autumn sea ice and the winter North Atlantic Oscillation. Wea. Climate Dyn., 1, 261–275, https://doi.org/10.5194/wcd-1-261-2020.
Smith, D. M., and Coauthors, 2019: The Polar Amplification Model Intercomparison Project (PAMIP) contribution to CMIP6: Investigating the causes and consequences of polar amplification. Geosci. Model Dev., 12, 1139–1164, https://doi.org/10.5194/gmd-12-1139-2019.
Smith, D. M., and Coauthors, 2022: Robust but weak winter atmospheric circulation response to future Arctic sea ice loss. Nat. Commun., 13, 727, https://doi.org/10.1038/s41467-022-28283-y.
Tian, W., J. Huang, J. Zhang, F. Xie, W. Wang, and Y. Peng, 2023: Role of stratospheric processes in climate change: Advances and challenges. Adv. Atmos. Sci., 40, 1379–1400, https://doi.org/10.1007/s00376-023-2341-1.
Torrence, C., and G. P. Compo, 1998: A practical guide to wavelet analysis. Bull. Amer. Meteor. Soc., 79, 61–78, https://doi.org/10.1175/1520-0477(1998)079<0061:APGTWA>2.0.CO;2.
Vertenstein, M., and Coauthors, 2013: CESM user’s guide (CESM1.2 release series user’s guide). NCAR Tech. Note, 104 pp., https://www2.cesm.ucar.edu/models/cesm1.2/cesm/doc_cesm1_2_1/usersguide/ug.pdf.
Vokhmyanin, M., T. Asikainen, A. Salminen, and K. Mursula, 2023: Long-term prediction of sudden stratospheric warmings with geomagnetic and solar activity. J. Geophys. Res. Atmos., 128, e2022JD037337, https://doi.org/10.1029/2022JD037337.
Wang, F., W. Shao, H. Yu, G. Kan, X. He, D. Zhang, M. Ren, and G. Wang, 2020: Re-evaluation of the power of the Mann-Kendall test for detecting monotonic trends in hydrometeorological time series. Front. Earth Sci., 8, 14, https://doi.org/10.3389/feart.2020.00014.
Warner, J. L., J. A. Screen, and A. A. Scaife, 2020: Links between Barents-Kara Sea ice and the extratropical atmospheric circulation explained by internal variability and tropical forcing. Geophys. Res. Lett., 47, e2019GL085679, https://doi.org/10.1029/2019GL085679.
Warshaw, M., and R. R. Rapp, 1973: An experiment on the sensitivity of a global circulation model. J. Appl. Meteor., 12, 43–49, https://doi.org/10.1175/1520-0450(1973)012<0043:AEOTSO>2.0.CO;2.
Wu, B., and J. Wang, 2002: Winter Arctic Oscillation, Siberian high and East Asian winter monsoon. Geophys. Res. Lett., 29, 1897, https://doi.org/10.1029/2002GL015373.
Wu, B., J. Su, and R. Zhang, 2011: Effects of autumn-winter Arctic sea ice on winter Siberian High. Chin. Sci. Bull., 56, 3220–3228, https://doi.org/10.1007/s11434-011-4696-4.
Wu, B., K. Yang, and J. A. Francis, 2016: Summer Arctic dipole wind pattern affects the winter Siberian High. Int. J. Climatol., 36, 4187–4201, https://doi.org/10.1002/joc.4623.
Wu, B., K. Yang, and J. A. Francis, 2017: A cold event in Asia during January–February 2012 and its possible association with Arctic sea ice loss. J. Climate, 30, 7971–7990, https://doi.org/10.1175/JCLI-D-16-0115.1.
Wu, B., Z. Li, J. A. Francis, and S. Ding, 2022: A recent weakening of winter temperature association between Arctic and Asia. Environ. Res. Lett., 17, 034030, https://doi.org/10.1088/1748-9326/ac4b51.
Yu, Q., and B. Wu, 2023: Summer Arctic atmospheric circulation and its association with the ensuing East Asian Winter Monsoon variability. J. Geophys. Res. Atmos., 128, e2022JD037104, https://doi.org/10.1029/2022JD037104.
Zhang, J., and Coauthors, 2022: Responses of Arctic sea ice to stratospheric ozone depletion. Sci. Bull., 67, 1182–1190, https://doi.org/10.1016/j.scib.2022.03.015.
Zhang, P., Y. Wu, I. R. Simpson, K. L. Smith, X. Zhang, B. De, and P. Callaghan, 2018: A stratospheric pathway linking a colder Siberia to Barents-Kara Sea sea ice loss. Sci. Adv., 4, eaat6025, https://doi.org/10.1126/sciadv.aat6025.
