Overview of the Centennial Progress in Research on the Arctic–midlatitude Connection

1.   Introduction

The year 2024 marks the centennial anniversary of the Chinese Meteorological Society. This article reviews major advancements in research on Arctic–midlatitude connection over the past century, with a particular focus on the significant achievements and contributions of Chinese scholars in this field.

In winter, the Arctic acts as a crucial cold source in the climate system due to intense radiative cooling and the insulation of the ocean by ice and snow. This modulates global energy balance, profoundly influencing weather events and climate variability within and beyond the Arctic region.

Since the beginning of the 21st century, the Arctic’s ecosystem has undergone profound transformations in response to sustained global warming, including rapid Arctic sea ice decline, the Atlantification of the Arctic Ocean, and amplified Arctic warming. The resulting sea ice loss has expanded the extent of open water areas in the Arctic Ocean and its marginal seas, thereby strengthening air–ice–ocean coupling. By the 2010s, the phenomenon of Arctic amplification and its impacts on mid-latitudes has emerged as a key focus of international research, with numerous studies greatly enhancing our comprehension of Arctic influences.

This review outlines critical advancements in Arctic–midlatitude linkage research across five key areas: (1) three milestones in the study of large-scale Arctic–midlatitude dynamic connections; (2) pathways of winter Arctic cold air outbreaks affecting East Asia; (3) the influence of Arctic sea ice variability on winter weather and climate; (4) the impact of Arctic warming on mid–low latitude; and (5) the connections between the Arctic and mid–low latitude weather and climate during summer. While we strive to provide a comprehensive overview of major research developments over the past century, limitations in scope and expertise may have resulted in the omission of certain relevant findings. We welcome constructive feedback from our readers.

2.   The three major milestones in the study of large-scale Arctic–midlatitude dynamic connections

In the same year the Chinese Meteorological Society was established, Walker (1924) first proposed the concept of the North Atlantic Oscillation (NAO) to describe the oscillation of atmospheric pressure anomalies between the Azores and Iceland. When the NAO is in a positive (negative) phase, the North Atlantic jet stream shifts poleward (equatorward), affecting weather and climate across the mid–high latitude of North Atlantic, the Arctic, and Europe, and also exerting a substantial impact on East Asia (Wu et al., 1999). During the positive phase of the NAO, the westerlies in the mid–high latitudes of Eurasia strengthen, thereby limiting the southward intrusion of cold Arctic air into the mid-latitudes. Conversely, in the negative phase, the southward propagation of Arctic cold air is facilitated, affecting the mid–low latitudes and leading to lower temperatures and heavy snowfall events (Wang and Shi, 2001; Wang et al., 2010; Luo et al., 2014). This atmospheric teleconnection pattern represents the most prominent mode of atmospheric variability in the Northern Hemisphere during winter, effectively demonstrating the dynamical and thermodynamic connections between the Arctic and mid-latitudes. The introduction of the NAO concept has profoundly influenced the development of atmospheric science and represents a landmark event in the study of Arctic-midlatitude dynamic linkages.

Building on the concept of the NAO, Thompson et al. (1998) further proposed the concept of the Arctic Oscillation (AO). They posited that the zonal asymmetry of surface temperature and mid-tropospheric circulation anomalies associated with the AO may directly link to land–sea thermal contrasts. Consequently, they argued that the AO fundamentally differs from the NAO. Generally, the NAO is considered to be a regional manifestation of the AO. Both the AO and the Arctic polar vortex are centered over the Arctic, and AO variabilities are largely reflected through variations of the Arctic tropospheric polar vortex (Zhang et al., 2008). For a long time, the NAO and AO were regarded as primary dynamical factors influencing Arctic atmospheric circulation and the Arctic Ocean (including its marginal seas). This view was only challenged with the emergence of new modes of Arctic atmospheric circulation in the early 21st century. Numerous studies have studied the influence of the AO on East Asian weather and climate variability. A negative phase of the AO during winter is typically associated with a weakening of the westerlies in the mid–high latitudes of Eurasia, thereby facilitating the southward intrusion of Arctic cold air into mid–low latitudes (Gong et al., 2001; Gong and Wang, 2003; Yang and Li, 2008; Wang and Zhou, 2018). However, Wu and Wang (2002) emphasized that the relationship between the winter AO and the Siberian High is not close, and their impacts on the East Asian winter monsoon are also independent.

In the early 2000s, Chinese scholars revealed dominant modes of Arctic atmospheric variability, such as the Arctic Dipole (AD) anomaly (e.g., Wu et al., 2006) and the Barents–Beaufort seas Oscillation (Wu and Johnson, 2007), providing new perspectives and explanations for understanding the Arctic sea–ice–air interactions and the connections between the Arctic and mid–low latitudes. During the Arctic winter (October to March), the AD anomaly corresponds to the EOF2 of sea-level pressure variability in the region north of 70°N. One of the anomaly centers is stably located between the Kara Sea and the Laptev Sea, while the other is situated over the Canadian Arctic Archipelago, extending southeastward through Greenland into the northern European seas. This AD anomaly differs from the “Barents Oscillation”. Due to its strong meridional characteristics, the AD has become a key mechanism driving anomalous Arctic sea ice and freshwater export from the Arctic Ocean, as well as the cold air outbreaks from the Arctic Ocean into the Barents Sea, Northern European Sea and northern Europe. The summer AD reflects the alternating shift of the polar vortex center between the Eastern and Western Hemispheres, thereby influencing East Asian summer precipitation (Wu et al., 2008). The AD sometimes replaces the AO/NAO and significantly contributes to Arctic sea ice melting (Wang et al., 2009). The AD also affects climate variability in Europe, East Asia and North America, and is associated with Arctic warming during 1920–1940. The latest study suggests that the AD is one of the key drivers of the Arctic Ocean Atlantification (Polyakov et al., 2023), affecting not only Arctic sea–ice–air interactions and the ecosystem but also influencing weather and climate outside the Arctic (Wu et al., 2016; Wu, 2017), with potential implications for the evolution of future short-term climate trends.

3.   Pathways of winter Arctic cold air impacting East Asia

3.1   Key pathways of Arctic cold air southward intrusion into East Asia

Due to the inhomogeneous distribution of solar radiation, large-scale cold air masses form in the Arctic. The frequency of Arctic cold air masses has decreased over the past few decades, along with a decline in their intensity and duration (Ye et al., 1995; Hankes et al., 2011). Moreover, Arctic cold air masses during summer have rapidly declined since the 1980s, contributing to increased surface air temperatures and more frequent heat waves in the mid–high latitudes (Liu et al., 2020).

Arctic cold air masses manifest as shallow cold anticyclone at the surface, while at 500 hPa, they appear as a circumpolar cyclonic vortex, i.e., the tropospheric polar vortex. Consequently, the tropospheric polar vortex is often regarded as a proxy for large-scale Arctic cold air activity (Zhang et al., 2008). Chinese meteorologists have studied the tropospheric polar vortex since the 1980s. After 1976, the tropospheric polar vortex began to deepen and strengthen, with its area generally increasing before the mid-1970s and decreasing afterward (Shi and Zhu, 1996; Zhang et al., 2006a). Cold air activities such as cold surges over the Eurasian continent are closely linked to the strength and position of the tropospheric polar vortex, with a larger and stronger vortex favoring the southward intrusion of Arctic cold air (Zhu et al., 2000; Liu and Gao, 2001; Zhang et al., 2006b). Recently, increasing attention has been given to the variability of the stratospheric polar vortex and its impact on weather and climate (Cohen et al., 2021; Tian et al., 2023). However, due to significant uncertainty in the downward propagation of stratospheric vortex anomalies into the troposphere, and a consensus has yet to be reached on this process (Wu, 2024).

Arctic cold air activities play a significant role in the occurrence of cold surges in mid-latitude regions. As early as the1940s, Chinese scholar Lu (1936a, b) conducted preliminary studies on the causes and characteristics of cold waves in China. He proposed that the southward intrusion of Arctic cold air masses into mid-latitudes causes pressure drop and the formation of storms (Lu, 1936a). He further analyzed the sources, pathways of cold air and associated weather changes (Lu, 1936b), and firstly revealed that cold waves originate from the Arctic Ocean and Novaya Zemlya, propagating southeastward through Central Asia. In contrast, cold surges intruding from Northeast China originates from the southward movement of cold air masses over Siberia or the Mongolian Plateau. After the cold air intrusion, the cold air masses move southward along open plains and river valleys. Cold wave events typically begin in October and persist until May of the following year, with the highest frequency occurring in November and December. A decaying low-pressure system from Siberia often precedes the cold wave invasion. During winter, these cold surges result in severe cold winds and snowstorms, while they manifest as sandstorms, gusts and thunderstorms in spring and early summer.

In the 1960s, Chinese scholars began to focus on the occurrence of cold waves, particularly highlighting the sources, pathways and circulation characteristics of cold air masses. Li (1955) carried out preliminary studies on East Asian cold waves, identifying the causes, invasion paths and other related phenomena. Based on these, Tao (1959) systematically classified the pathways of East Asian cold waves. The most common path for cold waves invading China involves cold air originating from the ocean surface west of Novaya Zemlya, crossing the Barents Sea, Siberia and Mongolia before reaching China, often resulting in relatively intense cold waves. A secondary pathway involves cold air migrating southward from Novaya Zemlya and the Kara Sea, then turns southeastward across Siberia and passing through Mongolia into China. A third pathway involves cold air accumulated in the Mongolia–Lake Baikal region directly intruding southward into China. Regarding the atmospheric circulation associated with cold waves, Gu (1956) classified the pathways of high-pressure forming winter cold waves and analyzed the circulation characteristics of three types of cold surges, highlighting the meridional component of atmospheric circulation. Zhu et al. (2000) emphasized that, after intensification, southward-moving cold air associated with cold surges often passes through a key cold surge region (43°–65°N, 70°–90°E), and then mainly intrudes into China via a northwestern route, an eastern route, a western route, or a combined eastern and western route. The weather systems contributing to East Asian cold surges include the Arctic tropospheric vortex, the polar high, surface high pressure and cold fronts. The formation of surface high pressure is directly related to the pathways and accumulation of Arctic cold air. Chinese scholars have categorized the mid-stage processes of East Asian cold surges into three types: the inverted Omega pattern, the eccentric polar vortex pattern, and the eastward-moving large-scale ridges and troughs (Zhu et al., 2000).

Studies have shown that intense cold air originating from the North Pacific, moving westward, can also cause severe cold waves. Takaya et al. (2005) suggested that the westward propagation of the North Pacific anticyclone triggers the development of blocking high-pressure ridges, further enhancing the southward movement of cold air. From mid-January to early February 2012, Eurasia experienced a severe cold surge that resulted in over 700 fatalities across the continent. The evolution of this cold wave displayed a classic “downstream effect” of atmospheric circulation. Prior to the onset, the Aleutian Low rapidly weakened, approaching the intensity of the winter Siberian High. As the Aleutian Low quickly recovered, the Siberian High concurrently intensified (Wu et al., 2017). Following its intensification, the Siberian High continued to expand westward, eventually affecting Europe. Studies have revealed that precursory signals of this extreme cold surge event can be traced back to the mid-latitude eastern Pacific region, and that the outbreak of the cold surge is directly linked to the formation, westward movement, and southward retrogression of polar blocking highs (Fig. 1). Figure 2 outlines the primary pathways of Arctic cold air moving southward, with the dark green arrows indicating that cold air originates from the Pacific sector of the Arctic Ocean, invading East Asia via Russian Far East, then shifting westward.

Download

Larger image

PowerPoint slide

Fig  1.  Evolution of SLP anomalies (relative to 1979–2012) during January 2012. (a) SLP anomalies averaged over 5–7 January, (b)–(i) same as in (a), but for 8–10, 11–13, 14–16, 17–19, 20–22, 23–25, 26–28, and 29–31 January. Date ranges in January are indicated on each plot. 17 January 2012 was the day when a strong cold wave broke out. From Wu et al. (2017).

Download

Larger image

PowerPoint slide

Fig  2.  Schematic diagram of southward paths of winter cold air mass originating from the Arctic, where thin and thick brown arrows respectively represent the southward path of Arctic cold air mass accompanied by the Ural blocking high and the direction of airflow in the middle troposphere. The dark green arrow represents the southward cold air mass coming from the Pacific sector of the Arctic Ocean, moving westward through the middle and high latitudes of East Asia and affecting the Eurasian continent.

In the late 20th century, scholars conducted a series of studies on cold high-pressure systems associated with cold surges. The process of cold air outbreaks is linked to the formation, intensification, southward movement, and modification of the Siberian High. Therefore, it is crucial to study the formation mechanisms of the Siberian High, as well as its dynamic and thermodynamic processes. Its formation and development result from the combined effects of mid- and upper-level air convergence and radiative cooling. As it moves southward to lower latitudes, enhanced sensible and condensation heating in the lower troposphere amplifies the transformation of the cold high. The Siberian High and its associated cold air outbreaks typically exhibit a pronounced low-frequency oscillation and southward propagation with a duration of 10–20 days (Ding et al., 1987; Ding, 1990). In terms of its dynamic structure, the pre-establishment phase of the high-pressure is characterized by positive vorticity in the mid-troposphere, weak convergence in the lower and upper levels, and divergence in the mid-levels. As the anticyclone develops, the upper levels exhibit positive vorticity and convergent flow, while the lower levels experience negative vorticity and divergence, resulting in subsidence throughout the troposphere (Ding et al., 1991). The cold high-pressure systems affecting China primarily originate from weak high-pressure systems in Europe, northern and central Asia. These then strengthen upon reaching western Siberia and Mongolia before moving southward in outbreaks that intrude into China, gradually modifying, warming, and weakening (Liu and Qiu, 1992). It is noteworthy that some cold wave high-pressure systems do not intensify through the eastward propagation of pre-existing cold highs from the Eurasian boundary; rather, they develop independently and are influenced by ascending adiabatic cooling (Zhang and Chen, 1999).

Severe cold wave events occur in both North America and Eurasia, with distinct sources and pathways of cold air masses in each region. In a study of extreme events in North America and Europe (1948–1999), Walsh et al. (2001) determined that cold air in North America originates from the north or northwest, while in Europe, it is sourced from the east via westward movement. Ding and Ma (2007) utilized isentropic potential vorticity to study the strong cold wave in East Asia from December 2004 to January 2005, revealing that cold air originated from high latitudes of Eurasia and the lower stratosphere and upper troposphere of the Arctic. During the winter of 2015/16, under the influence of a strong El Niño, China experienced an intense cold surge event from 20–25 January 2016. This southward movement of intense cold air led to the first snowfall in 70 years in Okinawa, Japan, and multiple fatalities due to freezing in Thailand. During this event, the central value of the Siberian High reached 1089 hPa on 23 January, marking the strongest single-day Siberian High since 1951. Research suggests that the anticyclonic wind field over the Arctic Ocean surface during the summer of 2015, coupled with above-average mean temperatures in the mid- and lower- troposphere, enhanced the negative feedback of reduced Arctic sea ice on winter atmospheric variability. This, in turn, further promoted the intensification of the Siberian High, thereby favoring the occurrence of severe cold surges in East Asia during the winter (Wu et al., 2016; Wu and Yang, 2016). Bueh et al. (2022) analyzed two extreme cold wave events during the 2020/21 winter, revealing that cold air for both events moved southward via a transpolar route and interacted with a northwestward route, resulting in the accumulation of cold air in key regions and its wide latitudinal expansion. Arctic cold air activities also extend to plateau areas, affecting temperature variations in the Qinghai–Xizang Plateau (Jiao et al., 2017).

The influence of Arctic cold air on cold wave events in different regions of China exhibits interannual and decadal variability. Cold waves and strong cold air events are most frequent in January and March, with the lowest occurrence in February. Cold air activities show a cycle of 2–5 years, and the frequency of cold waves has declined since the 1970s, coinciding with global warming (Qiu et al., 1992). From the 1970s to the 1990s, the sources, frequency and intensity of cold air activities in China changed (Li et al., 2006). For the Northeast China, both the frequency and intensity of spring cold surges showed distinct interdecadal shifts in the late 1980s and early 21st century: a decrease in cold surge frequency with an increase in intensity in the 1990s, followed by a rebound in cold surge frequency with reduced intensity in the early 21st century (Tang and Zeng, 2017). In North China, significant differences were observed in the frequency, intensity, and pathways of cold air propagation during the 1990s and 2000s (Bai and Zeng, 2022).

3.2   The crucial role of winter blocking highs in the Arctic–Midlatitude connection

As early as the mid-20th century, meteorologists had already recognized the significance of blocking highs (Elliott and Smith, 1949; Yeh, 1949; Rex, 1950a, b; Gui, 1956; Tao, 1957). Rex (1950a) provided a detailed description of blocking highs at the 500 hPa, which has become a classic depiction of blocking highs. The occurrence of blocking highs typically triggers significant changes in local and downstream weather patterns. Rex (1950b) found that when Atlantic blocking occurs during the winter, temperatures are lower over continental Europe, the British Isles, and the Mediterranean region, while the surface temperatures in areas dominated by the blocking ridge are anomalously warm. In the mid–high latitudes of Eurasia, the formation, persistence, and breakdown of blocking highs are directly linked to the southward intrusion of polar cold air into East Asia. In particular, blocking highs located over the Ural Mountains and western Siberia have historically been identified as critical dynamic systems influencing the southward movement of Arctic cold air. Gui (1956) classified the large-scale circulation patterns in East Asia during Ural blocking into three main types: one dominated by westerly winds, and two involving northerly components. Tao (1957), in his analysis of Eurasian circulation in February 1956, found that the collapse of blocking highs originally situated over the Atlantic and the Urals during cold air outbreaks led to the intrusion of cold air into East Asia. Yeh et al. (1962) proposed that the Ural Mountains, Lake Baikal, and the Sea of Okhotsk regions are three critical areas in Eurasia where blocking highs impact weather and climate in China. Following the breakdown of blocking highs in each of these regions, a notable temperature drop often occurs in northern China within a week. Yi (1982) investigated the frequency and duration of Ural blocking highs during the cold season (October–April) and concluded that the more frequent and longer-lasting blocking highs are associated with relatively lower temperatures in China during the same period. Chen et al. (1991) further explored the impact of Northern Hemisphere blocking on China’s winter temperatures from 1951 to 1999. Their analysis showed that except for January, Ural blocking significantly influenced temperatures across most of China, with stronger blocking highs corresponding to lower temperatures.

Over time and with increasing observational data, further evidence has confirmed that the Ural blocking high is a major circulation system responsible for severe cold surges (Ji et al., 2008; Ma et al., 2008). Narentuya et al. (2001) analyzed the strong cold wave in mid-March 1998 and found that the cold air outbreak was closely linked to changes in blocking conditions in the Ural region, particularly the collapse and re-establishment of the blocking high. Li and Li (2007) also observed that during a severe cold wave in Jiaxing in March 2005, the eastward shift and southward collapse of the Ural blocking triggered the southward movement of cold air. The cold waves during the winter of 2004/05 disrupted the warm winter trend in China that had persisted since 1986. These cold waves were mainly driven by the blocking ridges over the Ural Mountains and the Caspian Sea (Ma et al., 2008). In January 2008, a severe cold wave, accompanied by rain, snow, and ice, affected much of southern China. This event was driven by the prolonged development of the Ural blocking high, which led to persistent cold-air incursions in central, eastern, and southern China (Gao et al., 2008; Gu et al., 2008; Wang et al., 2008).

In winter, the blocking high near the Ural Mountains acts as a dynamic conveyor, transporting warm air from the mid-latitudes to the Arctic while simultaneously advecting cold air from the Arctic to the mid-latitudes (Feng and Wu, 2015; Chen et al., 2018, 2021a; Wu et al., 2022). Arctic warming weakens the meridional potential vorticity gradient, which in turn promotes the development and persistence of blocking highs (Chen and Luo, 2019; Luo et al., 2019a, b; Zhang et al., 2019). With ongoing Arctic warming and sea ice melt, reduced meridional temperature and potential vorticity gradients result in significantly extended lifespans of Eurasian blocking, which contributes to winter cold surges in both Eurasia and North America (Chen and Luo, 2017; Yao et al., 2023).

4.   The impact of Arctic sea Ice variations on winter weather and climate

4.1   Early research findings

Sea ice is a critical component of the climate system, playing a significant role in regulating the ocean’s absorption of solar radiation by modifying its albedo. Its presence also acts as a barrier, inhibiting the exchange of heat, momentum, and moisture between the ocean and the atmosphere. Moreover, sea ice changes are closely related to the ocean’s freshwater cycle, surface buoyancy, and stratification, which may, in turn, affect deep ocean circulation and long-term climate trends.

Before the 1980s, the intricate dynamics of polar ocean currents and atmospheric circulation were not fully understood, although many researchers hypothesized that a significant connection existed between Arctic and global climate. In terms of observations, efforts were focused on identifying ice–atmosphere relationships, with researchers concentrating on local sea ice changes before the advent of satellite observations. Schell (1955) found that summer sea ice in Greenland and the Barents Sea could serve as an indicator for predicting ice conditions and temperatures in southern regions. Later, Schell (1970) confirmed that sea ice was a reliable predictor for precipitation in Northern Europe and the North Atlantic. Miles (1974) discovered a significant positive correlation between the intensity of atmospheric circulation in the North Atlantic and the severity of sea ice along the Newfoundland coast. Rogers and Van Loon (1979) showed that sea ice anomalies in the North Atlantic were part of the Greenland–North European Oscillation, linking cold winters in Greenland and warm winters in Northern Europe to reduced sea ice in the Baltic Sea and increased ice in the Davis Strait the following summer, along with more drift ice near Newfoundland in the spring. Walsh and Johnson (1979) used observational data to investigate the relationship between sea ice and atmospheric circulation changes on monthly and seasonal time scales. They found that atmospheric fields associated with sea ice change indicated that the atmosphere forced sea ice change, such that prior atmospheric circulation patterns from January to April influenced sea ice distribution from February to July. However, during the sea ice growth period (August to January), the influence of sea ice on atmospheric circulation slightly exceeded the atmospheric influence on sea ice. This indicated that Arctic sea ice exerted a more prominent influence on atmospheric circulation during autumn and winter, laying the groundwork for future sea ice impact studies. In a study by Johnson (1980) using January Arctic sea ice data from 1953 to 1977, the 5 years with the most and least sea ice were identified, revealing that when Arctic sea ice extent was greater, the Icelandic Low strengthened and shifted northeast, the subtropical high over the North Atlantic shifted poleward, and both the Aleutian Low and the Pacific subtropical high weakened. Surprisingly, the surface temperature in the sea ice region was not particularly low (relative to normal values), suggesting that in a context of above normal Arctic sea ice, the interannual variability of atmospheric circulation may not be strongly linked to sea ice anomalies.

Early numerical simulations of Arctic sea ice forcing focused on the effects of removing Arctic sea ice on atmospheric circulation changes (Fletcher, 1968; Newson, 1973; Warshaw and Rapp, 1973; Royer et al., 1990). Newson (1973) indicated that the absence of Arctic sea ice would weaken the temperature gradient in middle and high latitudes, facilitating the amplification of planetary waves and promoting the formation of blocking circulations. Fletcher (1968) found that an ice-free Arctic would result in a weak temperature gradient and weaker zonal circulation in the Northern Hemisphere. Royer et al. (1990) demonstrated that removing Arctic sea ice caused a decrease in surface air pressure over the Arctic Ocean and increased precipitation, with moderate cooling effects observed in mid-latitudes of Europe and Asia.

4.2   The influence of Kara–Barents Seas ice on the East Asian Winter Monsoon

Although China is geographically distant from the Arctic, Arctic sea ice anomalies can influence its weather and climate by affecting both atmospheric and oceanic circulations. In the late 1980s and early 1990s, Chinese researchers highlighted that Arctic sea ice plays a significant role in atmospheric circulation changes, even surpassing the influence of sea surface temperatures in the equatorial eastern Pacific (Fang, 1987; Huang et al., 1992; Yang et al., 1992, 1994; Wu et al., 1996). In the late 1990s, Chinese researchers revealed a close relationship between winter sea ice anomalies in the Kara–Barents Seas and the East Asian winter monsoon. Specifically, they found that an anomalously high (low) sea ice extent in this region during winter is associated with a weaker (stronger) Siberian High, a weaker (stronger) East Asian trough, and a weaker (stronger) East Asian winter monsoon, leading to fewer (more) cold air outbreaks into China (Wu and Huang, 1999). This study, through observational analyses and numerical modeling experiments, first revealed the significant role of Barents–Kara Sea ice variability in East Asian winter climate, preceding similar international research by over a decade. This conclusion was further supported by two studies a decade later (Petoukhov and Semenov, 2010; Inoue et al., 2012). Numerical simulations showed that reduced winter sea ice in the Barents–Kara Seas can lead to cold winters in Eurasia, and that atmospheric circulation response to sea ice forcing in this region exhibits non-linear characteristics (Petoukhov and Semenov, 2010). Furthermore, a reduced sea ice extent in the Barents Sea during winter promotes anticyclonic activity in the Barents Sea and along the northern edge of Eurasia, leading to increased air pressure over the northern Eurasian region, which promotes the intensification of the Siberian High (Inoue et al., 2012).

4.3   Two pathways through which rapid Arctic sea ice melting affects weather and climate in mid-latitudes

Since the late 1990s, global warming has intensified, driven by a combination of anthropogenic activities and natural variability, leading to a rapid decline in Arctic sea ice, particularly in summer. Consequently, the causes of Arctic sea ice melting and its impacts on mid-latitudes have become a major focus international research. Some representative studies have emphasized the effects of autumn and winter Arctic sea ice melting on subsequent Eurasian temperatures and the East Asian winter monsoon (Francis et al., 2009; Honda et al., 2009; Wu and Zhang, 2010; Wu et al., 2011; Liu et al., 2012; Mori et al., 2014). Studies have shown that early and late winter cold anomalies over Eurasia are related to reduced Arctic sea ice in the preceding September, which can intensify the Siberian High (Honda et al., 2009). Furthermore, a reduction in Arctic sea ice during summer and autumn is statistically related to subsequent winter atmospheric circulation anomalies resembling a negative phase of the Arctic Oscillation (Wu and Zhang, 2010). Wu et al. (2011) found that persistent and anomalously low sea ice concentration in key Arctic regions (the Barents, Kara, and Laptev Seas, and their northern adjacent areas), along with anomalously high sea surface temperatures in the subarctic and North Atlantic, are associated with a stronger Siberian High and lower winter temperatures over East Asia. A large ensemble of numerical simulations also shows that sea ice loss in the Barents–Kara Seas leads to more than a doubling in the probability of severe winters across central Eurasia (Mori et al., 2014).

Ding et al. (2023) found that, with rapid Arctic warming and persistent sea ice melting and retreat, autumn sea ice characteristics in the Arctic have changed, particularly with a significant increase in the interannual amplitude of sea ice in the East Siberian–Chukchi–Beaufort (EsCB) regions. Reduced sea ice in this region promotes persistent Arctic anticyclonic anomalies, leading to colder temperatures in western and central Eurasia in early winter and frequent extreme low-temperature events in central–western China during mid-winter. The upward propagation of planetary wave 2 anomalies drives the Arctic anticyclonic anomalies, with tropospheric processes directly enhancing these anomalies in early winter, and stratospheric processes indirectly facilitating the intraseasonal development of European anticyclonic anomalies during mid-winter, which cause cold air outbreaks along a “transpolar pathway” toward China, resulting in rapid temperature drops (Ding et al., 2023). Furthermore, a weakened stratospheric polar vortex that persists into late winter extends the influence of Arctic sea ice into early spring, serving as a potential predictor for temperature variability in northern Eurasia (Ding et al., 2021). The connection between autumn EsCB sea ice and mid–low latitude Eurasia has significantly strengthened since the late 1990s, particularly regarding extreme low-temperature events in China. In contrast, the impact of autumn sea ice anomalies in the Barents–Kara Seas has a larger spatial scale and higher intensity, but it primarily affects mid-high latitudes and occurred mainly before 2000.

4.3.1   Impact on mid-latitudes through tropospheric processes

Arctic sea ice melt primarily affects the meridional temperature gradient and zonal westerlies in the Arctic and mid- to high-latitude troposphere, which in turn influences atmospheric storm tracks and jet streams, including their position and intensity (Newson, 1973; Royer et al., 1990; Murray and Simmonds, 1995; Alexander et al., 2004; Deser et al., 2004; Cohen et al., 2014; Smith et al., 2022; Screen et al., 2022). Sea ice anomalies in specific regions, due to their unique geographical locations, can trigger Rossby wave teleconnections, subsequently influencing mid-latitude winter weather. Specifically, reduced sea ice in the Barents–Kara Seas leads to a strengthened Siberian High, accompanied by a “− + −” Eurasian wave train, where an intensified Ural blocking and a deepening East Asian trough promote the southward intrusion of cold air, thus causing winter cold anomalies in Eurasia (Wu et al., 1999; Honda et al., 2009; Petoukhov and Semenov, 2010; Wu et al., 2011; Wu et al., 2013; Mori et al., 2014; Wu et al., 2015, 2017). In contrast, reduced sea ice in the Chukchi–Bering Seas leads to a stronger high-pressure ridge over Alaska, which steers cold air into North America, thus resulting in cold anomalies over North America during winter (Kug et al., 2015). Through these tropospheric processes, Arctic sea ice melt can influence weather and climate in mid-latitudes.

4.3.2   Impact on mid-latitudes through stratosphere–troposphere interactions

Stratosphere–troposphere coupling also represents a potential pathway through which anomalously low Arctic sea ice can influence atmospheric circulation. When there is anomalous low sea ice in the Barents–Kara Sea during early winter, it can trigger atmospheric planetary waves to propagate from the troposphere to the stratosphere. Upon reaching the stratosphere, wave breaking occurs, which then affects the strength of the stratospheric polar vortex, leading to its weakening. Later in the winter, the weakened stratospheric polar vortex propagates downward into the troposphere, causing the tropospheric circulation patterns that resemble a negative phase of the Arctic Oscillation, which in turn affects weather and climate in mid-latitudes (Jaiser et al., 2013; Cohen et al., 2014; Kim et al., 2014; Nakamura et al., 2015; Cohen et al., 2021). Similar upward planetary wave propagation mechanisms have been identified in response to sea ice reduction in the Laptev, East Siberian, Chukchi, Bering, and Okhotsk Sea regions (Sun et al., 2015; Nakamura et al., 2016; Chen and Wu, 2018; Cohen et al., 2018; Kretschmer et al., 2018; Ding et al., 2021; Ding and Wu, 2021), where enhanced upward planetary wave propagation results in EP flux convergence in the upper troposphere and stratosphere, sustaining the weakening of zonal westerlies. Researchers have used numerical methods, such as nudging experiments, to suppress stratosphere–troposphere coupling in models (Wu and Smith, 2016; Zhang P. F. et al., 2018; Chen et al., 2021b), revealing that when the stratospheric pathway is suppressed, the models show only a local warming response to Barents–Kara Sea ice reduction, without a cold anomaly response over Eurasia. When only the stratospheric influence is included, the models show a cold anomaly response over Eurasia but no local warming anomalies in the Barents–Kara Sea. However, when the stratosphere–troposphere coupling is not suppressed in the models, they can simulate the Warm Arctic–Cold Continent pattern (Zhang P. F. et al., 2018). Of course, the complexity of stratosphere–troposphere interactions and the uncertainties surrounding downward stratospheric warming propagation make it more difficult for Arctic sea ice melt to influence mid-latitude regions through this pathway (Wu, 2024).

4.4   Arctic sea ice melt has periodically weakened the Arctic–Asia temperature connections

As previously mentioned, numerous studies have shown that Arctic sea ice melt, particularly in the Barents–Kara Seas during autumn and winter, can induce a Warm Arctic–Cold Eurasia temperature anomaly pattern (Honda et al., 2009; Wu et al., 2011, 2013; Liu et al., 2012; Mori et al., 2014). The mean tropospheric temperature in the lower levels over East Asia (35º–55ºN， 90º–120ºE) during winter has exhibited distinct phasic changes (Fig. 3). The time series of cumulative temperature anomalies at 1000 hPa (Fig. 3a) reveals that over the past 45 winters, the temperature has experienced three distinct phase shifts, occurring during the winters of 1986/87, 2004/05, and 2013/14. From the winter of 1986/87 to 2003/04, the cumulative temperature anomaly showed an increasing trend, during which China experienced 17 warm winters (Fig. 4). From the winter of 2004/05 to 2012/13, the cumulative temperature anomaly over the East Asian region showed a decreasing trend, during which Arctic sea ice entered a phase of rapid melting, coinciding with a winter Warm Arctic–Cold Eurasia pattern. After the winter of 2013/14, the phase shifted to a Warm Arctic–Warm Eurasia pattern. A 9-year sliding t-test further confirms the validity of the phasic temperature division (Fig. 3b).

Download

Larger image

PowerPoint slide

Fig  3.  (a) Winter Asian regional (35°–55°N， 90°–120°E) averaged 1000-hPa temperature (cyan) and its cumulative deviation evolution (orange yellow). (b) 9-year running t-test of winter Asian regional (35°–55°N， 90°–120°E) averaged 1000-hPa temperature (cyan), dashed orange yellow lines denote 0.05 significance level. All data are derived from NCEP/NCAR Reanalysis data I.

Download

Larger image

PowerPoint slide

Fig  4.  (a) Winter 1000-hPa air temperature anomalies derived from the mean averaged over winters of 1986/87–2003/04 minus the average of 1979/80–2023/24 (°C), black contours denote air temperature anomalies at 95% confidence level. (b) and (c) as in (a), but for winters of 2004/05 – 2012/13 and winters of 2013/14–2023/24, respectively. All data are derived from NCEP/NCAR Reanalysis I dataset.

Wu et al. (2022) found that the Arctic–Asia temperature linkage strengthened during the winters of 2004/2005–2012/2013, whereas the Warm Arctic–Warm Eurasia temperature pattern that emerged during the winters of 2013/14–2018/19 resulted in a phasic weakening of this connection. This suggests that, even with anomalously low Arctic sea ice and persistent Arctic warming, the connection between the Arctic and mid-latitudes exhibits alternating periods of strengthening and weakening.

Numerical simulations forced by observed Arctic sea ice concentration suggest that the rapid Arctic sea ice melt and its persistent anomalously low state have led to the phasic evolution of the Arctic–Asian winter temperature connection (Wu and Li, 2022; Wu et al., 2025). The winter temperature in Asia response to the continuous melting of Arctic sea ice exhibits significant low-frequency oscillation. Arctic sea ice melt directly contributes to anomalous Arctic warming, and through tropospheric processes, stratosphere–troposphere coupling, and other unknown mechanisms, can induce a Warm Arctic–Cold Eurasia pattern. With persistent Arctic sea ice melting, the magnitude and extent of Arctic warming further intensify, such that positive temperature anomalies expand southward from the Arctic into high-latitude Eurasia, thus favoring a Warm Arctic–Warm Eurasia pattern. However, the mechanisms through which persistent Arctic sea ice melting generates phasic atmospheric circulation anomalies resembling a positive phase of the NAO [see Fig. 4 in Wu et al. (2022)], as well as the mechanism behind the periodic weakening of the Arctic–midlatitude connection, remain unclear. At this point, the tropospheric processes and stratosphere–troposphere interaction mechanisms triggered by the Arctic sea ice anomalies are no longer effective. Since the winter of 2013/14, significant changes have occurred in the Arctic and high-latitudes climate systems, and the relationship between autumn Arctic sea ice extent and winter NAO and stratospheric polar vortex intensity has also weakened since 2012 (Smith et al., 2022).

4.5   The influence of other factors on modulating the impact of Arctic sea ice

The influence of Arctic sea ice on mid-latitude regions is not only determined by the direct effects of Arctic sea ice but is also closely linked to nonlinear processes associated with Arctic sea ice influence (Petoukhov and Semenov, 2010), the background state of Arctic sea ice (Semenov and Latif, 2015; Luo et al., 2019a), the background conditions of models (Smith et al., 2019), the thermodynamic and dynamic state of the summer Arctic atmospheric circulation (Wu et al., 2016, 2017; Yu and Wu, 2023), the seasonality and magnitude of Arctic sea ice melt (Zhang and Screen, 2021), different climatic backgrounds (He et al., 2023), the synergistic effects of multi-factors across different latitudes (Zhang et al., 2019, 2022a, b; Gao et al., 2024; Zou and Zhang, 2024), and other factors (Sato et al., 2014; Screen et al., 2016; Warner et al., 2020; Rudeva and Simmonds, 2021). These factors also contribute to the uncertainty in the influence of Arctic sea ice on mid-latitudes.

Zhang and Screen (2021), using multi-model large-ensemble sensitivity experiments, investigated the critical sea ice season and magnitude affecting mid-latitudes. Their results indicated that only moderate autumn sea ice loss can induce the Warm Arctic–Cold Eurasia pattern, with stratospheric dynamic processes being a key physical mechanism. They also found that, compared to autumn, winter sea ice loss leads to weaker mid-latitude cold anomalies, and a complete absence of sea ice in winter can even result in warm winters, highlighting the nonlinearity and seasonality of Arctic sea ice impacts. He et al. (2023) pointed out that, over the past century, Arctic and Siberian temperature patterns have shown phasic changes, with a robust Warm Arctic–Cold Eurasia pattern only appearing after 1980 and no significant association between 1945 and 1985. In contrast, a Warm Arctic–Warm Eurasia pattern was observed during 1901–1945. They suggest these phasic changes may be attributed to decadal variability in the stratospheric polar vortex and the background state of the tropospheric Ural blocking high.

4.6   Possible impacts of Arctic sea ice melting on winter extreme events

In recent years, extreme cold and heavy snowfall events in winter have occurred frequently and received extensive attention. In January 2008, a rare freezing rain and snow disaster occurred in southern China, causing significant economic losses. This event is likely related to the reduction of Arctic sea ice. In September 2007, the extent of Arctic sea ice reached its lowest value since satellite observations. Previous studies have found that anomalously low Arctic sea ice extent during the autumn and winter of 2007 caused atmospheric circulation anomalies that resulted in the convergence of strong moisture fluxes from the Bay of Bengal and the Pacific with persistent cold air from eastern Europe and Siberia over central and southern China, leading to severe extreme weather events (Liu et al., 2012; Chen et al., 2013).

Research shows that the state of the summer Arctic atmospheric circulation not only affects concurrent Arctic sea ice but also modulates the lagged influence of anomalously low summer and autumn Arctic sea ice on subsequent winter atmospheric circulation (Wu et al., 2016; Wu F. M. et al., 2019; Yu and Wu, 2023). Therefore, anomalous summer Arctic atmospheric circulation patterns serve as important precursor signals for predicting subsequent winter East Asian temperatures and the winter monsoon. The extreme cold weather experienced over Eurasia from 17 January to 1 February 2012, can likely be attributed to the combined effects of anomalously low Arctic sea ice and the thermodynamic and dynamic characteristics of the Arctic atmospheric circulation in the preceding summer (Wu et al., 2017).

The latest research has confirmed that Arctic sea ice melting is a primary driver of extreme cold and heavy snowfall events in Europe. From February to March 2018, a severe easterly wind swept across Scandinavia, leading to extreme cold weather throughout much of Europe, with strong Arctic cold air outbreaks sweeping the continent and causing snowfall even in southern Europe. Bailey et al. (2021) found that, during this heavy snowfall event, the Barents Sea provided a moisture flux of 9.3 mm day⁻¹, approximately 140 billion tons of water vapor evaporated from the Barents Sea, which accounted for over 88% of the snowfall over northern Europe (corresponding to anomalously low sea ice in this region). This study further revealed that the net evaporation from the Barents Sea in March increased by 70 kg m⁻² per unit sea ice melting between 1979 and 2020 (Bailey et al., 2021). This research directly links the amount of Arctic sea ice melt to sea water evaporation and extreme snowfall.

Regarding the extreme cold and heavy snow events that occurred in the central and southern United States in mid-February 2021, Cohen et al. (2021) proposed that the reduced sea ice in the Barents–Kara Sea in autumn and increased Eurasian snow cover influenced the stretching and deformation of the stratospheric polar vortex, which subsequently affected the central and southern United States, leading to these extreme cold and blizzard conditions. Yao et al. (2023) analyzed the extreme cold and snowstorm events that frequently occurred in North America and Eurasia from November to December 2022. They believed that the atmospheric circulation background causing these extreme cold and snowstorm events was related to the melting of Arctic sea ice.

5.   Influence of Arctic warming on mid–low latitudes

5.1   Characteristics of Arctic warming

Since the Industrial Revolution, the global climate has experienced significant warming, with the most prominent warming in the Arctic region. Based on data from Arctic lakes, tree rings, and ice cores, Kaufman et al. (2009) found that since 1900, the Arctic has reversed the millennial scale cooling trend and has started to exhibit a significant warming trend. This warming process can be divided into two periods: an early period from the 1920s to the 1940s, and a recent period from the 1970s to the present (Bengtsson et al., 2004; Yamanouchi, 2011; Suo et al., 2013; Tokinaga et al., 2017; Bokuchava et al., 2021). Bokuchava et al. (2021) calculated the warming rates of these two periods respectively. The results show that the warming rate during 1916–1945 was approximately 0.47°C (30 yr)⁻¹, which is comparable to 0.56°C (30 yr)⁻¹ during 1976–2005. However, the warming was mainly concentrated in high-latitude regions (Yamanouchi, 2011). Regarding the Arctic warming since 1979, You et al. (2021) pointed out that the annual average warming trend in the Arctic during 1979–2020 was approximately 0.72°C (10 yr)⁻¹. This result is consistent with the 0.75°C (10 yr)⁻¹ calculated by Rantanen et al. (2022). The maximum warming rate in the core regions (near the Arctic Ocean on the Eurasian side, Svalbard, and Novaya Zemlya) reached as high as 1.25°C (10 yr)⁻¹.

In addition, Arctic warming also exhibits seasonal variability. Bekryaev et al. (2010) found that during 1875 –2008, the warming trends in winter, spring, summer, and autumn were 1.73, 1.59, 0.87, and 1.09°C (100 yr)⁻¹, respectively. This indicates that Arctic surface warming is strongest in spring and winter and weakest in summer (Wu et al., 2014). While Arctic warming is evident at the surface, significant warming is also observed within the Arctic troposphere (Graversen et al., 2008; Screen et al., 2010; Cohen et al., 2014). Recently, some studies have further classified the Arctic tropospheric warming into two types: deep warming and shallow warming, and explored their impacts on the weather and climate in Eurasia (He et al., 2020; Li et al., 2023).

Notably, the Arctic amplification is not equivalent to Arctic warming. The Arctic amplification generally refers to the phenomenon where the surface air temperature in the Arctic region increases at more than twice the rate of the global average warming (Zhao et al., 2015; Cohen et al., 2020; Han X. X. et al., 2023). The Arctic amplification was first proposed by Arrhenius (1896). He found that changes in carbon dioxide concentration could amplify surface air temperature anomalies in polar regions. Manabe and Wetherald (1975) utilized a simplified three-dimensional atmospheric circulation model to show that a doubling of carbon dioxide would induce an increase in tropospheric temperature and a decrease in stratospheric temperature, with high-latitude tropospheric warming being 2 to 3 times greater than the overall global warming. By analyzing the station data of Svalbard from 1975 to 2014, Wei et al. (2016) found that the annual average warming rate in this region was four times higher than the global average warming rate. The latest research also shows that since 1979, the surface warming rate in the Arctic region has reached four times the global average warming rate (Rantanen et al., 2022). In addition, the annual average Arctic amplification effect increased sharply in 1986 and 1999 (Chylek et al., 2022). The rapid Arctic warming not only directly affects the local ecological environment, but also has a profound impact on the global climate system through a series of complex physical processes (Chen L. Q. et al., 2003; Cohen et al., 2014, 2020; Zhao et al., 2015; Wu F. M. et al., 2019).

5.2   Impacts on mid–low latitudes

Arctic warming increases the Arctic atmospheric thickness, weakens the meridional temperature gradient and geopotential height gradient between high- latitudes and mid–low latitudes. According to the thermal wind relationship, the zonal wind near 60°N systematically weakens from the troposphere to the stratosphere. Researchers evaluated more than 3000 ensemble experiments of 16 atmospheric circulation models from the Polar Amplification Model Intercomparison Project (PAMIP). They found that, under the forcing of low sea ice in recent years, the 16 atmospheric circulation models consistently showed the weakened westerlies in the mid-latitudes and enhanced westerlies in the subtropics (Screen et al, 2022; Smith et al, 2022). These numerical modeling results provide robust evidence for the connection between Arctic warming and mid-latitude westerly weakening.

The weakened and more meandering jet stream slows down the propagation speed of Rossby waves, causing weather systems to move eastward more slowly (Francis and Vavrus, 2012, 2015). Consequently, weather systems become more persistent. This persistent weather pattern is closely related to the northward extension of atmospheric high-pressure ridges and the southward extension of atmospheric low-pressure troughs. The increased meridional amplitude of the circulation increases the likelihood of extreme weather and climate events (Petoukhov et al., 2013; Screen and Simmonds, 2014; Francis et al., 2018).

The weakened mid-latitude westerlies are often associated with the reduced atmospheric baroclinicity, thereby weakening the storm track. The storm track can act as a bridge connecting the Arctic and mid-latitudes by redistributing momentum, moisture, and heat fluxes (Yu et al., 2024). Yang et al. (2024) found a negative correlation between the Siberian storm track and the Warm Arctic–Cold Eurasia pattern. The weakening of the Siberian storm track can enhance the generation of the Warm Arctic–Cold Eurasia pattern through synoptic eddy feedback. Conversely, the Warm Arctic–Cold Eurasia pattern weakens the Siberian storm track by suppressing the lower-tropospheric baroclinicity. Inoue et al. (2012) found that reduced Barents Sea ice leads to a prevalence of wintertime synoptic-scale anticyclonic anomalies along the Siberian coast, resulting in anomalous warm advection over the Barents Sea and cold advection over eastern Siberia. In low-ice years, the weakened baroclinicity over the Barents Sea inhibits cyclone eastward propagation, leading to a reduction in cyclones and an increase in anticyclones along the Siberian coast.

Arctic warming can also exert impact on the mid-latitude weather and climate by increasing snow cover in mid-latitude continents during winter. Specifically, Arctic warming enhances the atmospheric capacity to hold moisture, and Arctic sea ice melt exposes more open ocean, increasing evaporation. These two factors jointly create favorable water vapor conditions for increased snowfall in mid-latitude continents during winter (Cohen et al., 2012; Liu et al., 2012; Zhang et al., 2019). Due to the high persistence and albedo of snow cover, increased snow cover, by altering surface energy balance and hydrological processes, enhances the Siberian High and excites upward-propagating planetary waves that weaken the polar vortex (Han and Sun, 2018; Xu et al., 2018). This increased snow cover leads to cooling over mid-latitude continents, which to some extent reduces the temperature gradient between the Arctic and mid-latitudes.

Luo (2005) proposed a Nonlinear Multi-scale Interaction model based on the meridional potential vorticity gradient, and then explored how Arctic warming affects the Ural blocking by influencing the atmospheric background conditions, especially the westerlies and vertical wind shear. Arctic warming usually leads to a decrease in the meridional potential vorticity gradient, which is conducive to the maintenance of blocking and the occurrence of extremely cold events in Eurasia (Luo et al., 2017; Yao et al., 2017; Luo et al., 2023). While the potential vorticity gradient theory effectively describes internal atmospheric dynamic processes, it does not capture the role of external forcing and lacks a direct connection with sea ice and snow cover.

Some studies emphasize that only deep Arctic warming has a significant impact on mid–low latitudes (Xu et al., 2019). Observations show that Arctic warming trends are indeed deep, with two maxima, one in the lower troposphere (below 700 hPa) and one in the upper troposphere (around 300 hPa). However, not all model experiments forced by low sea ice anomalies reproduce such deep Arctic warming. Xu et al. (2023) proposed that deep Arctic warming is a critical condition for inducing the Warm Arctic–Cold Continent pattern, and this deep warming is a dynamic response to sea ice loss. Specifically, the upward propagation of planetary waves and their convergence in the stratosphere causes a vortex feedback that triggers clockwise residual circulation anomalies. The descending branch of this circulation then warms the atmosphere through adiabatic heating in the Arctic. Nudging experiments have also confirmed that stratosphere–troposphere coupling plays an important role in this process. Through the analysis of numerous numerical experiments, He et al. (2020) found that Eurasian winter cold anomalies, weakened polar jet stream, and Ural blocking are more likely to occur in winters with deep Arctic warming. Labe et al. (2020) demonstrated a strong linear relationship between the Arctic 1000–500 hPa thickness and the Siberian High Index in model simulations, even with different models or experimental designs. A greater Arctic 1000–500 hPa thickness was associated with a stronger Siberian High, but they also found that models only forced by sea ice changes do not always capture the deep vertical warming structure.

Arctic winter warming exhibits various spatial patterns, with some warming centers located over the Canadian Arctic and Greenland (Zhang et al., 2022a), while others are found over the Arctic Ocean near Eurasia (Wu, 2017). Wu (2017) investigated the first mode of the 1000–500 hPa atmospheric thickness variability north of 60°N during winter. This Arctic warming mode exhibits significant decadal variations. Since the winter of 2004/05, the warming has been notably intensified, which is directly related to the Arctic sea ice decline and the Arctic Ocean Atlantification (Wu et al., 2025). This Arctic warming mode has no connection with the Arctic Oscillation. Instead, it has a closer relationship with the dominant mode of the atmospheric circulation in the mid–low latitudes of the Northern Hemisphere (EOF2 of the winter SLP). This indicates that the anomalous atmospheric circulation in the mid–low latitudes is one of the causes of the Arctic warming.

6.   Linkages between the Arctic and the mid–low latitudes in summer

Although the effects of anomalous wintertime variability in the Arctic sea ice–air system on mid- and low-latitude weather and climate have been extensively investigated, the impact of Arctic sea ice loss and related atmospheric circulation changes on summer weather and climate is equally significant.

6.1   Impacts of Arctic sea ice anomalies on East Asian summer climate

Fang (1990) utilized a simple statistical model to reveal that a larger Arctic sea ice extent in January corresponds to a more southward subtropical high in June. Subsequent studies, utilizing statistical analyses and numerical experiments, further confirmed that Arctic sea ice has a significant impact on the interannual variability of the East Asian summer monsoon, potentially surpassing the influence of sea surface temperature (SST) in the equatorial central–eastern Pacific (Yang et al., 1994; Huang et al., 1992, 1995; Wu et al., 1996). The North Pacific SST plays an essential bridging role in this process (Guo et al., 2014). Arctic sea ice anomalies during winter and spring are likely to change the properties of local air masses, which can trigger Rossby wave trains and exert lagged effects on temperature and precipitation in downstream East Asian region (Fang et al., 1991; Fang and Yang, 2009; Zhang R. N. et al., 2018). Different regions of Arctic sea ice anomalies lead to varying climatic impacts. A synthesis of existing studies reveals that the regions of sea ice variability with the most significant impact on East Asian summer climate include the waters west of Greenland (from Baffin Bay southward to the Newfoundland coast of Canada), the Greenland Sea, the Barents–Kara Seas, and the Bering Sea–Okhotsk Sea.

Winter sea ice in the western Greenland is closely linked to subsequent summer atmospheric circulation anomalies over mid- and high-latitude Eurasia, which in turn influence summer precipitation patterns in East Asia. Wu et al. (2013) emphasized that springtime atmospheric circulation anomalies over the southern Newfoundland region act as a crucial link, connecting persistent winter–spring sea ice anomalies in western Greenland, North Atlantic SST anomalies, and summer atmospheric circulation anomalies over Eurasia. Therefore, winter sea ice in the western Greenland may act as a precursor for predicting summer atmospheric circulation over Eurasia and summer precipitation anomalies in East Asia. Wang et al. (1993) found a significant negative correlation between spring Greenland sea ice and July drought and flood in the middle and upper reaches of the Yellow River. Similarly, Bai et al. (2000) revealed a close relationship between drought and flood in North China and preceding anomalies in Greenland sea ice. Building on these findings, Chen M. X. et al. (2001, 2003) further investigated the relationships between winter–spring Greenland sea ice variability and early summer temperature and precipitation in China. They proposed that variations in sea ice extent near Greenland during winter and spring influence early summer climate anomalies in China by modulating Arctic atmospheric circulation and Northern Hemisphere blocking highs. Specifically, studies have shown that during spring, extensive sea ice anomalies in the Greenland and Barents Seas correspond to anomalous anticyclones over the Lake Baikal and Mongolian Plateau. This is accompanied by a weakening of the Ural blocking high, which suppresses the southward intrusion of polar cold air, resulting in higher-than-average temperatures across China (Wu L. et al., 2019).

Wang and Guo (2004) pointed out that sea ice anomalies in the Barents Sea from May to July can trigger a $"-+-"$ wave train over northern Eurasia, leading to a distinctive early summer precipitation distribution in East China, characterized by increased rainfall in the south and decreased rainfall in the north. Additionally, other studies have indicated a close relationship between reduced sea ice in the Barents Sea and teleconnection patterns such as the Silk Road teleconnection, the Pacific–Japan teleconnection, and East Asian precipitation patterns (He et al., 2018; Li et al., 2018; Han et al., 2021). Based on this, Li et al. (2021) further explored the influence of early spring Barents Sea ice variability on the spatial temperature pattern in eastern China during summer. They found that sea ice anomalies in the Barents Sea in March can influence the north–south dipole temperature anomalies in eastern China during summer through the Silk Road teleconnection.

The Bering Sea and the Sea of Okhotsk are located at the junction of the Arctic Ocean and the Pacific Ocean. Based on reanalysis data, several studies suggest that sea ice variability in these regions during winter and spring may influence summer precipitation in East Asia. Niu et al. (2003) found that sea ice decline in the Bering Sea and the Sea of Okhotsk corresponds to increased precipitation in southern China and decreased precipitation in the north during the summer. Zhao et al. (2004) reached a similar conclusion, indicating that reduced sea ice extent in the Bering Sea and the Sea of Okhotsk during spring is likely to result in increased summer monsoon precipitation in southeastern China. However, the physical mechanisms behind these effects remain unclear. To address this, Tian et al. (2021) combined reanalysis data and atmospheric circulation models and revealed the possible mechanisms by which sea ice variations in the Bering Sea during the melting period (March to June) affect East Asian summer precipitation.

Current research on the link between Arctic sea ice and East Asian summer climate primarily focuses on interannual timescales. However, some studies suggest that this relationship exhibits interdecadal variations. Specifically, the connection between precipitation in northeastern China and sea ice anomalies has significantly strengthened after the 1990s. Since the 1990s, a significant positive correlation has been observed between summer precipitation in northeastern China and the sea ice extent in East Siberia. Additionally, sea ice loss in the Barents Sea during spring is more likely to trigger extreme droughts in northeastern China (Du et al., 2022; Han T. T. et al., 2023).

6.2   Potential connections between Arctic sea ice anomalies and summer extreme events

In recent years, the role of Arctic sea ice anomalies in the frequent occurrence of extreme heatwaves and precipitation has received increasing attention. Tang et al. (2014) found that reductions in Arctic sea ice and snow cover during early summer contribute to an increase in extreme high-temperature events in summer. Zhang et al. (2020) identified, on an interdecadal timescale, that the combined effects of summer Arctic sea ice and Eurasian snow cover drives an increase in European heatwaves. The reduction of sea ice and snow cover is associated with a weakening and poleward shift of the upper-tropospheric jet stream, while amplifying planetary wave activity. This, in turn, enhances the persistence of weather systems, thereby increasing the likelihood of extreme events. Additionally, the location of heatwaves in North China is influenced by sea ice anomalies (Zhang et al., 2022a). When sea ice in the Barents–Kara Sea is above normal, the heatwave center in North China shifts northward, whereas below-normal sea ice leads to a southward shift.

Extreme heatwaves, when coupled with favorable atmospheric circulation conditions, can trigger wildfires, leading to significant ecological damage, a surge in carbon emissions, and substantial socioeconomic losses. Research indicates that the reduction of Arctic sea ice during summer creates conditions conducive to the onset of wildfires. Zou et al. (2021) pointed out that the rising occurrence of large wildfires in the western United States during autumn can be partly attributed to the decline in Arctic sea ice in summer and autumn. The reduction in sea ice initiates anomalous dry and hot weather in the region, conducive to wildfire outbreaks. This feedback mechanism further enhances the absorption of solar radiation by the surface, leading to regional warming and exacerbated dry conditions. Zhang et al. (2024) analyzed the extreme heatwave events in Russia during the summers of 2010 and 2016, revealing the physical mechanisms by which Arctic sea ice reduction contributes to the occurrence of extreme heatwaves. Building on this, Luo B. H. et al. (2024) highlighted that the decline in Arctic sea ice has intensified warming in high-latitude regions, resulting in heatwaves that are more prolonged, widespread, and intense. These intensified heatwaves, in turn, have triggered larger and more persistent wildfire events.

Extreme precipitation, due to its spatiotemporal characteristics, has been less attributed to sea ice anomalies in the previous studies. However, recent studies suggest that the combined effects of Arctic sea ice anomalies and other atmospheric systems can lead to anomalous atmospheric circulation that provides energy for regional weather activity, promoting the formation and development of local circulation systems, and ultimately leading to extreme precipitation events. For instance, studies have linked the extreme precipitation event along the Mei-yu front in the Yangtze River Basin and Japan in 2020 to preceding Arctic sea ice melt (Chen et al., 2021a, 2022). Chen et al. (2021a) argued that the sea ice loss facilitated the development and intensification of the blocking high over East Siberia, allowing cold air to invade the Mei-yu front region, which hindered the seasonal northward migration of the Mei-yu front, and increased the temperature gradient on either side of the front, ultimately resulting in extreme precipitation in the region. Other studies have highlighted the significant role of SST anomalies in the Indian Ocean in modulating this process (Zhou et al., 2021; Chen et al., 2022). For the extreme precipitation in North China in July 2021, Liu et al. (2023) highlighted that record-breaking warm SST anomalies in the tropical Atlantic and the atmospheric teleconnections triggered and modulated by the sea ice loss in the Laptev Sea and East Siberian Sea were the main contributors to this event.

6.3   Summer Arctic cold anomalies in the mid–low troposphere linked to East Asian heatwaves

Since the beginning of this century, heatwaves have occurred more frequently in summer, leading to significant fatalities and economic losses (Meehl and Tebaldi, 2004; Barriopedro et al., 2011; Bador et al., 2017; Wu L. et al., 2019). The number of deaths attributed to heatwaves in the period from 2001 to 2010 increased by more than 2000% compared to the 1991–2000 period (WMO, 2013). Wu L. et al. (2019) revealed the dominant pattern of summer 1000–500-hPa thickness variability north of 30°N and its connection to East Asian heatwaves. This study showed that the EOF2 of summer 1000–500-hPa thickness exhibits strong interannual variability but no significant trend, which contrasts sharply with the EOF1. The positive phase of the EOF2 corresponds to a significant cold anomaly in the mid- and low troposphere of the Arctic during summer, surrounded by warm anomalies. This Arctic cold anomaly is directly linked to the frequency of heatwave events in the middle and low latitude regions of East Asia.

The Arctic has experienced frequent summer cold anomalies in the mid–low-troposphere since 2005. Along with the occurrence of Arctic cold anomalies, a significant enhancement of the upper-level westerlies in the troposphere has dominated most of the Arctic, while westerlies in the middle and low latitudes of Asia have weakened. The acceleration of westerlies in the Arctic lower troposphere substantially intensifies the baroclinicity of the Arctic atmosphere, leading to an increased frequency of anomalously low surface pressure in the Arctic and a decreased frequency over the high latitudes of Eurasia and North America. In the middle and low latitudes of Asia, the weakened upper-troposphere westerlies favor the maintenance of anticyclonic anomalies in the lower and middle troposphere, suppressing summer precipitation and enhancing solar shortwave radiation heating, which facilitates the occurrence of East Asian heatwaves. The study suggests that the systematic northward shift of Asian zonal winds is the intrinsic mechanism linking Arctic cold anomalies to East Asian heatwaves, generating a seesaw structure of zonal wind anomalies between the Arctic and the Qinghai–Xizang Plateau. The systematic meridional displacement of the westerlies in the troposphere connects the Arctic with the Qinghai–Xizang Plateau (Wu L. et al., 2019). Figure 5 illustrates the dynamical connection between summer Arctic cold anomalies and East Asian heatwave events. The study further indicates that the upper-tropospheric zonal wind anomalies over the Arctic in summer are one of the precursor signals for the East Asian winter monsoon. Anomalously strong zonal westerlies in the upper troposphere of the Arctic in summer are conducive to higher surface temperatures in East Asia during the subsequent winter, as well as weaker East Asian winter monsoon, and vice versa (Fig. 6).

Download

Larger image

PowerPoint slide

Fig  5.  Schematic diagram of the dynamic connection between Arctic summer cold anomalies in the middle and lower troposphere and summer heatwave events in the Yangtze River Basin in China.

Download

Larger image

PowerPoint slide

Fig  6.  (a) Normalized time series of the summer 300-hPa Arctic westerly index averaged over north of 70°N (red line) and the ensuing winter (DJF) Siberian high index (blue line). Their correlation is −0.59, significant at 99% confidence level. (b) Winter 500-hPa geopotential height anomalies (gpm), derived from a linear regression on the normalized summer Arctic westerly index. White and black contours represent anomalies at 95% and 99% confidence levels, respectively. (c), (d) as in (b), but for winter SLP (hPa) and SAT (°C) anomalies, respectively. From Wu and Francis (2019).

These findings suggest that the anomalous melting of Arctic sea ice is not necessarily linked to the occurrence of summer heatwaves in the middle and low latitudes of East Asia. Since 2005, the frequent occurrence of summer Arctic cold anomalies in the mid- and low-troposphere has slowed the seasonal melting of Arctic sea ice, contributing to the absence of new record lows in Arctic sea ice extent in September since 2012 (Francis and Wu, 2020). Furthermore, studies indicate that Arctic sea ice melting promotes the development of cold anomalies in the mid- and low-troposphere during summer (Wu and Li, 2022; Wu et al., 2023), thereby establishing atmospheric circulation conditions that inhibit further sea ice melting.

7.   Discussion

7.1   Limited progress in recent studies compared to earlier findings

As early as the late 1970s, Walsh and Johnson (1979) proposed that during the sea ice growth period (August to January), the influence of Arctic sea ice on atmospheric circulation is more pronounced than the reciprocal influence of the atmosphere on sea ice. This finding guided subsequent research on the impact of Arctic sea ice. Before the 1980s, several studies, using numerical simulation experiments, explored the effects of removing Arctic sea ice on atmospheric circulation. These studies found that a reduction in the mid–high latitude temperature gradient following the removal of Arctic sea ice facilitated planetary wave amplification, thereby promoting the formation of blocking circulations, which led to a moderate cooling effect over mid-latitude Eurasia. These conclusions are largely consistent with current findings on the effects of anomalously low Arctic sea ice extent. This suggests that some research conducted since 2000, focusing on the link between anomalously low Arctic sea ice and the increased likelihood of blocking circulation anomalies and associated cold anomalies over high-latitude Eurasia, has not yielded substantial advances compared with studies from the 1960s and 1970s. The main distinction between early and present-day research lies in the utilization of more advanced models, the latest multi-source reanalysis datasets, sea ice concentration data, and outputs from multi-model intercomparison projects.

7.2   Atmospheric circulation models significantly underestimate the impact of Arctic sea ice melting on mid-latitudes

It is important to acknowledge that the understanding of the impact of Arctic sea ice on mid-latitudes has undergone a complex evolutionary process over the past century. Initial assertions of an impact were followed by a period of skepticism, with some studies doubting or even rejecting the influence of Arctic sea ice on mid-latitudes due to the perceived weakness of sea ice forcing and uncertainties in research findings. After a period of further investigation, the ability of Arctic sea ice to affect mid-latitude weather and climate was reaffirmed. A major basis for the skepticism was the observation that atmospheric circulation models consistently underestimate the responses to anomalous Arctic sea ice forcing compared with observed anomalies. This is likely due to simplified parameterizations of crucial Arctic sea ice-related physical processes within these models. For example, the common practice of specifying a fixed Arctic sea ice thickness of 2 m neglects the influence of sea ice thinning, particularly in the marginal seas of the Arctic Ocean. Such a simplified approach artificially reduces ocean turbulent heat fluxes associated with thin ice. The latest study demonstrates that current atmospheric circulation models markedly underestimate the influence of Arctic sea ice on mid-latitude Eurasia (Yu et al., 2024).

7.3   Future research directions

Future research on large-scale Arctic–midlatitude dynamic linkages should focus on the formation mechanisms of Arctic atmospheric circulation modes (e.g., the Arctic Dipole, anomalously cold modes in the mid- to low-troposphere over the Arctic during summer), particularly the underlying atmospheric dynamic processes and their connections with Arctic sea ice and sea surface temperature anomalies. Studies should emphasize the role of Arctic sea ice melt in driving low-frequency variability in atmospheric circulation, with greater attention to the influence of spatial variations in sea ice anomalies and their varying magnitudes. There is a need for quantitative investigations into the role of Arctic sea ice melt in extreme weather and climate events. Furthermore, the predictive and forecasting potential of Arctic air–ice–ocean coupling on East Asian weather and climate should be explored. Finally, future research must address the projected changes in Arctic–midlatitude linkages under various climate change scenarios and explore adaptation and mitigation strategies.

8.   Summary

The expansion of Arctic observational datasets, particularly following the advent of satellite observations in the 1970s, has provided a robust foundation for investigating the role of the Arctic within the climate system. Concurrently, advancements in the understanding of sea–ice–air interactions and the continuous development of sophisticated climate models have become powerful tools for the investigation of Arctic–midlatitude linkages. A century of sustained research efforts has yielded substantial progress in this domain. Representative research advancements are summarized as follows.

(1) Since 1924, three landmark discoveries have been made in the field of large-scale Arctic–midlatitude linkages: the identification of the North Atlantic Oscillation (NAO), the Arctic Oscillation (AO), and the Arctic Dipole (AD) pattern. The Arctic Dipole pattern is not only closely related to Arctic warming and Arctic sea ice decline but can also, at times, supersede the AO/NAO, significantly impacting mid-high latitude atmospheric circulation.

(2) Chinese scholars have made outstanding contributions to the understanding of cold air outbreaks in East Asia, specifically regarding the pathways and mechanisms of Arctic air intrusions, the key regions of cold surge, the formation of the Siberian High, and the development of East Asian cold wave systems. These original research findings have been incorporated as core content in key monographs and textbooks and continue to play a vital role in both research and weather forecasting.

(3) The Arctic has undergone two distinct periods of pronounced warming since the 20th century: the first occurring between the 1920s and 1940s, and the second commencing in the mid-to-late 1990s. Since the early 21st century, Arctic warming and associated reductions in sea ice extent have emerged as salient indicators of climate change. These changes not only profoundly influence Arctic ecosystems but also exert considerable impacts on weather and climate variability in the mid-low latitudes of the Northern Hemisphere.

(4) Anomalous Arctic warming and sea ice reductions influence mid-latitude weather and climate through a complex suite of physical processes. These physical processes include changes in sea surface moisture evaporation, turbulent heat fluxes at the air–sea interface, meridional temperature gradients and tropospheric zonal wind patterns in the mid–high latitude, the location and intensity of synoptic-scale storm tracks, horizontal teleconnection patterns, and the vertical propagation of planetary waves between the troposphere and stratosphere.

(5) Interdecadal variability in Arctic sea ice and its anomalously low extent can induce alternating temperature patterns of Warm-Arctic/Cold-Eurasia and Warm-Arctic/Warm-Eurasia during winter. The former enhances the connections between the Arctic and mid-latitudes of Eurasia, while the latter significantly weakens this linkage.

(6) A directly dynamical connection exists between the frequency of summer heatwaves in China’s mid–low latitudes and the recurrent occurrence of cold anomalies in the mid- and low-troposphere over the Arctic during the summer. The underlying mechanism connecting these Arctic cold anomalies to East Asian heatwaves is the systematic northward displacement of the zonal winds across the Asian continent. Specifically, when cold anomalies occur in the mid- and low-troposphere over the Arctic during summer, they intensify the high-latitude tropospheric zonal westerlies, subsequently contributing to the formation of blocking highs which promote heatwave and wildfire events.