China’s Recent Progresses in Polar Climate Change and Its Interactions with the Global Climate System※

Xichen LI ,Xianyao CHEN ,Bingyi WU ,Xiao CHENG ,Minghu DING ,Ruibo LEI ,Di QI,Qizhen SUN,Xiaoyu WANG,Wenli ZHONG,Lei ZHENG,Meijiao XIN,Xiaocen SHEN,Chentao SONG,and Yurong HOU

1Institute of Atmospheric Physics,Chinese Academy of Sciences,Beijing 100029,China

2Frontier Science Center for Deep Ocean Multispheres and Earth System and Physical Oceanography Laboratory,Ocean University of China,Qingdao 266100,China

3Qingdao National Laboratory for Marine Science and Technology,Qingdao 266061,China

4Department of Atmospheric and Oceanic Sciences/Institute of Atmospheric Sciences,Fudan University,Shanghai 200438,China

5CMA-FDU Joint Laboratory of Marine Meteorology,Shanghai 200438,China

6School of Geospatial Engineering and Science,Southern Marine Science and Engineering Guangdong Laboratory(Zhuhai),Sun Yat-sen University,Zhuhai 519082,China

7State Key Laboratory of Severe Weather,Chinese Academy of Meteorological Sciences,Beijing 100081,China

8Key Laboratory for Polar Science of the MNR,Polar Research Institute of China,Shanghai 200136,China

9Polar and Marine Research Institute,Jimei University,Xiamen 361021,China

10Polar Research and Forecasting Division,National Marine Environmental Forecasting Center,Beijing 100081,China

11College of Earth and Planetary Sciences,University of Chinese Academy of Sciences,Beijing 100049,China

ABSTRACT During the recent four decades since 1980,a series of modern climate satellites were launched,allowing for the measurement and record-keeping of multiple climate parameters,especially over the polar regions where traditional observations are difficult to obtain.China has been actively engaging in polar expeditions.Many observations were conducted during this period,accompanied by improved Earth climate models,leading to a series of insightful understandings concerning Arctic and Antarctic climate changes.Here,we review the recent progress China has made concerning Arctic and Antarctic climate change research over the past decade.The Arctic temperature increase is much higher than the global-mean warming rate,associated with a rapid decline in sea ice,a phenomenon called the Arctic Amplification.The Antarctic climate changes showed a zonally asymmetric pattern over the past four decades,with most of the fastest changes occurring over West Antarctica and the Antarctic Peninsula.The Arctic and Antarctic climate changes were driven by anthropogenic greenhouse gas emissions and ozone loss,while tropical—polar teleconnections play important roles in driving the regional climate changes and extreme events over the polar regions.Polar climate changes may also feedback to the entire Earth climate system.The adjustment of the circulation in both the troposphere and the stratosphere contributed to the interactions between the polar climate changes and lower latitudes.Climate change has also driven rapid Arctic and Southern ocean acidification.Chinese researchers have made a series of advances in understanding these processes,as reviewed in this paper.

Key words: polar climate change,recent progress in China,Arctic amplification,tropical-polar interactions,global sea level rise,stratospheric circulation

1.Introduction

In the recent four decades since the satellite era,many dramatic climate changes have been observed over the Arctic and Antarctic regions (Zhang,2015;Yao et al.,2017;Yu et al.,2018a;Wang et al.,2019d;Francis and Wu,2020;Li et al.,2021b).These polar climate changes affect changes in the global climate system (Chen and Tung,2018;Luo et al.,2019;Chen et al.,2020a;Chen and Luo,2021;Wu and Li,2022;Yuan et al.,2022).The climate changes over the Arctic and the Antarctic are characterized by different features.The Arctic climate change is much faster than that of the globe,a phenomenon known as the Arctic amplification(Wu,2017;Dai et al.,2019).The surface air temperature trend over the Arctic region is at least two to four times as fast as the global mean warming rate (Wu,2017;Wei et al.,2016),which results in a rapid decline and altered physical properties of the Arctic sea ice (Lei and Wei,2020).

The climate-related changes around the Antarctic show a zonally asymmetric pattern,with the strongest changes happening around the West Antarctic (Ding et al.,2011;Wang et al.,2019b,d;Li et al.,2021b;Zhang et al.,2021c).The surface warming around the West Antarctic and the Antarctic Peninsula is about two to three times that of the global mean rate,whereas the warming trend over the East Antarctic has not been not significant in recent decades (Yuan et al.,2015;Turner et al.,2020;Li et al.,2021b).The overall Antarctic sea-ice extent had been increasing before 2014,known as the Antarctic sea-ice paradox (Li et al.,2014;Yuan et al.,2017;Yu et al.,2018a;Wang et al.,2019c;Li et al.,2021b),except that the sea-ice extent around the Amundsen-Bellingshausen seas had dramatically decreased (Li et al.,2014;Yu et al.,2018a).However,this positive sea-ice trend reversed after 2015,as was followed by rapid declines around the entirety of Antarctica (Wang et al.,2019d;Li et al.,2021b;Ding et al.,2022a).

The polar climate changes interact with the climate variability over the lower latitudes.First,tropical-polar teleconnections influence polar climate changes (Ding et al.,2011,2014;Li et al.,2014,2015b,2019c,2021b).On interannual time scales,El Ni?o dominates the tropical Antarctic teleconnection (Yuan and Martinson,2001;Ding et al.,2011;Li et al.,2021b;Zhang et al.,2021b),with the Indian Ocean Dipole mode also playing an important role (Rao and Ren,2020;Yu et al.,2022).On decadal and longer time scales,the Pacific Decadal Oscillation (PDO) and the Atlantic Multi-decadal Oscillation (AMO) both contribute to the Antarctic multi-decadal changes (Li et al.,2014,2015a,b,2021b).In particular,the positive phase of the AMO since the early 1980s and the following negative phase of the PDO since the late 1990s both served to deepen the Amundsen Sea Low (Li et al.,2014,2021b).These circulation changes contribute to the asymmetric Antarctic climate changes,including the rapid warming over the West Antarctic,the sea-ice paradox,and the accelerated ice shelf melting around the West Antarctic.The climate variability over the tropics and the mid-latitudes may also contribute to the Arctic changes through both atmospheric bridges and oceanic pathways (Ding et al.,2014;Hu et al.,2016;Wu,2017;Luo et al.,2019;Liu et al.,2021b;Xiao and Ren,2021;Wu and Ding,2022).The tropical Pacific decadal variability may cause rapid warming around northeastern Canada and Greenland (Ding et al.,2014),while the AMO may also contribute to the evolution of the Arctic Amplification (Fang et al.,2022).On the other hand,the North Atlantic warming may drive a warmer Arctic Ocean as well as an Arctic sea-ice decline (Wang et al.,2017a;Zhang et al.,2022e) through oceanic processes,known as the Arctic Atlantification.The thinner and younger Arctic sea ice is more sensitive to atmospheric and oceanic forcing,which leads to the fact that the Arctic sea ice is physically more like the Antarctic sea ice,leading to the phenomenon of Antarctification of Arctic sea ice (Granskog et al.,2020).A strengthened Siberian high well corresponds to a systematic northward shift of atmospheric circulation over northern Eurasia,the North Atlantic,and northern North America,leading to significant increases in the atmospheric total energy transport into the Barents-Kara Seas (Wu and Ding,2022).Therefore,atmospheric circulation anomalies in the middle and lower latitudes play important roles in generating Arctic warm anomalies (Wu,2017).

The Arctic and Antarctic climate changes may feedback to the lower latitudes and contribute to changes in the Earth’s climate system (Gong et al.,2017;Wu and Francis,2019;Chen et al.,2020b;Dou et al.,2020;Wu et al.,2022a;Yuan et al.,2022).The rapid warming over the Arctic is usually associated with cooling over the land areas of the midlatitudes in the Northern Hemisphere,known as the Warm Arctic Cold Continents pattern (Sun et al.,2016b;Wu,2017;Guan et al.,2020),or the Warm Arctic Cold Eurasian pattern (Overland et al.,2011;Wu et al.,2011;Luo et al.,2016;Xu et al.,2018).In particular,the Arctic climate variability may largely contribute to the climate variability around East Asia,impacting the surface temperature (Li et al.,2019b;Zhang et al.,2020;Bi et al.,2022),the precipitation(Zhang et al.,2021f;Wu and Li,2022) and the air pollution(Ji and Fan,2019b;Liu et al.,2020b) around China.On the other hand,it is reported that the variability of the Southern Ocean SST and the Antarctic sea ice may also influence the tropical convection and inter-hemispheric energy balance(Xiang et al.,2018),as well as precipitation over southern China (Dou et al.,2020;Yuan et al.,2022).Finally,both the Greenland ice sheet and the Antarctic ice sheet experienced accelerated melting during the past three decades (Zhang et al.,2021c),largely contributing to the global mean sea level rise (Chen et al.,2014,2017),as well as the freshwater cycle over the high-latitude oceans.

In this paper,we first review the recent research progress of the observed Arctic and Antarctic climate changes in section 2.We then review the tropical—polar teleconnections and their contributions to polar climate changes in section 3.The impacts of the polar climate changes on the lower latitudes are reviewed in section 4.The dynamics related to stratospheric processes are then summarized in section 5,followed by a conclusion and discussion.

2.Recently observed polar climate changes

New techniques have recently been developed,which largely increased our capability to monitor the climate changes over the polar regions,especially during the modern satellite era since 1979.A series of dramatic climate changes have been observed over the polar regions.The Arctic climate changes are characterized by rapid surface warming,accompanied by the accelerated melting of both land and sea ice,a process known as Arctic amplification.The Antarctic climate changes exhibit clear regionally and seasonally dependent features,with the strongest changes usually occurring around West Antarctica.The recent progress in the observed Arctic and Antarctic climate changes are reviewed as follows.

2.1.The Arctic Amplification

During the past four decades,the Arctic has experienced rapid climate change.The surface air temperature over the Arctic has increased rapidly in all seasons (Fig.1).The warming rate over the Arctic region is more than four times faster than that of global warming,a phenomenon termed Arctic amplification.Some regions,like Svalbard Island,can experience warming four times greater than the global average(Wei et al.,2016).This Arctic warming is usually accompanied by anomalous cooling over the mid-and high-latitudes of North America and the Eurasian continent and an anomalous warming over the North Pacific,North Africa,and the lower latitudes of North America and Asia (Wu,2017)(Fig.1).

Fig.1.The seasonal surface air temperature (SAT) trend over the Northern Hemisphere from 1979 to 2020.The color shadings in (a—d) show the SAT trend for boreal winter (DJF,a),spring (MAM,b),summer (JJA,c),and autumn (SON,d),respectively.Dotted regions are significant at p<0.05 based on a Mann-Kendall test.

The rapid warming over the Arctic region is associated with an accelerated sea-ice decline (Ding et al.,2017;Luo et al.,2021) (Fig.2).Recent studies indicated that the summertime atmospheric warming and moistening over the Arctic region,synchronized by an atmospheric circulation adjustment over the high-latitude northern hemisphere,may increase the downwelling longwave radiation over the polar region,contributing to the Arctic sea-ice loss,especially in summer (Ding et al.,2017;Luo et al.,2021;Liang et al.,2022;Shen et al.,2022b).Similarly,the SST anomalies over the mid-and high-latitude North Pacific and North Atlantic may drive anomalous advection of both temperature and water vapor and enhance water vapor and cloud radiative feedback processes,thus contributing to the accelerated Arctic sea-ice decline in boreal autumn (Yu and Zhong,2018).Additionally,increased snowfall or rain-on-ice events,associated with increased warm air mass intrusions from low latitudes,together with thinner sea ice (Liang et al.,2022),would have the net effect of enhancing the precipitation contribution to the sea—ice mass balance.Thinner sea ice and additional ice,in turn,promote an increasing winter oceanto-ice heat flux in the Arctic Ocean (Zhong et al.,2022) and delay the onset of winter basal growth of sea ice (Lei et al.,2022).Moreover,the Arctic sea-ice loss shows a strong regional feature.For example,a positive phase of the Arctic dipole may intensify the ice-albedo feedback,thus contributing to rapid sea-ice loss over the Pacific sector of the Arctic Ocean in boreal summer (Lei et al.,2016).In addition,recent studies (Francis and Wu,2020) indicated that cold anomalies have frequently occurred in the middle and lower levels of the troposphere over the Arctic region since 2005.Such atmospheric changes may have led to a slowdown of the Arctic sea-ice decline over the recent decade (Francis and Wu,2020).

The Arctic amplification was reported to be associated with a series of physical processes and feedback (Xiao et al.,2020a;Fang et al.,2022),including an intensified transport of heat and moisture between the Arctic and lower latitudes(Wang et al.,2020e;Xu et al.,2021c;Wang and Chen,2022),and an increase in the outgoing longwave radiation induced by the recent Arctic sea-ice loss (Dai et al.,2019).The rapid surface warming and sea-ice decline are also partially attributed to the strengthened subsidence flow associated with an intensified polar cell (Qian et al.,2016).In addition,the decline of the Arctic sea ice and its associated surface warming may influence the atmospheric lapse rate and thus adjust the stability of the boundary layer over the Arctic region (Zhang et al.,2021d).By modulating the lower tropospheric stability,the Arctic sea ice may also impact the pathway of liquid water,leading to an adjustment of the lowlevel clouds (Yu et al.,2019b).Such feedback between the sea ice and the atmosphere may further contribute to Arctic amplification,especially during the initial ice-freezing season(Zhang et al.,2021d).

An important consequence of the Arctic amplification is the accelerated melting of the Greenland ice sheet and the associated freshwater discharge,which has significantly raised the sea level and affected ocean stratification (Yan et al.,2013,2014;Ran et al.,2021).The Greenland ice sheet,especially the glaciers on its edge,has experienced severe melting in the past decades (Yang et al.,2019b;Zhang et al.,2022c),primarily due to the warm sub-surface water intrusion from the Atlantic Ocean.

Recently,the surface melting and flow of the glaciers are accelerating,which also contributes to the accelerated mass loss and internal structure change of the Greenland Ice Sheet (Liu et al.,2016;Wang et al.,2019a;Huai et al.,2020;Chen and Luo,2021;Wang and Luo,2022;Luo and Lin,2023).The surface mass balance of the Greenland ice sheet is tightly associated with the surface temperature and precipitation changes.The ice surface temperature in the northwest portion of the Greenland Ice Sheet has increased since the 21st century with a strong seasonal feature (Hall et al.,2013;Huai et al.,2021),accelerating the surface melting in this region (Wei et al.,2022).The total precipitation over Greenland shows a positive long-term trend,with the peak season having shifted during the past decade (Ding et al.,2020;Huai et al.,2022).

The melting of the Greenland ice sheet has released large amounts of freshwater into the North Atlantic Ocean,leading to an adjustment of the ocean circulation (Hu et al.,2011).Simulations show that the Atlantic Meridional Overturning Circulation (AMOC) is weakened due to this freshwater discharge,which may also affect global climate by adjusting meridional heat transport (Yang et al.,2016a;Yu et al.,2016;Liu et al.,2018).Similar ocean circulation adjustments have occurred over the Arctic Ocean,associated with increased sea-ice melt,Arctic river discharge,and changes in atmospheric and oceanic circulations.For example,an anomalous accumulation of freshwater in the western Arctic Ocean leads to a spin-up of the Beaufort Gyre (Zhong et al.,2019b) and the redistribution of Pacific water in the deep basin (Zhong et al.,2019a).The transport of water masses from the East Siberian Sea into the Makarov Basin associated with the cyclonic Arctic circulation may contribute to the development of the halocline in the Makarov Basin (Wang et al.,2021c).

2.2.Asymmetric Antarctic climate changes

During the past four decades,coincident with the modern satellite era,Antarctica has experienced a series of dramatic climate changes (Wang et al.,2017b,2019d;Zhang et al.,2021c),showing strong zonally asymmetric features (Li et al.,2021b).These changes involve rapid surface warming in West Antarctica and the Antarctic Peninsula (Gao et al.,2018;Ding et al.,2020;Lin et al.,2020;Xu et al.,2021a),an opposite trend of snow accumulation between the Antarctic Peninsula and the western part of West Antarctica (Wang et al.,2017b),a pre-2014 expansion of the Antarctic sea-ice extent (Yu et al.,2018a;Wang et al.,2019c),followed by a rapid sea-ice decline (Wang et al.,2019d),and the accelerated land-ice melting (Wang et al.,2019b;Zhang et al.,2021c)especially around the West Antarctic.

Since the second half of the 20th century,the surface air temperature trends over the Antarctic have been characterized by a seesaw-like pattern,with rapid warming over the West Antarctic and the Antarctic Peninsula (Fig.3a) (Ding et al.,2011;Li et al.,2021b) and mild cooling over mainland East Antarctica (Fig.3a).The warming trend over the West Antarctic was about twice the global warming rate (Li et al.,2014,2015a,b,2021b).At the same time,the East Antarctic has experienced negligible warming or even a weak cooling trend (Fig.3a) (Li et al.,2014,2015a,b).Recently,scattered station observations show some reversed temperature trends since the beginning of the 21st century (Figs.3b—e),including the disappearance (vanishing) of the rapid warming over the Antarctic Peninsula and the West Antarctic (Figs.3b,c),accompanied by rapid warming of the South Pole since the early 2000s (Figs.3d,e) (Turner et al.,2020;Li et al.,2021b).

The snow accumulation rate over the Antarctic also shows a strong regional dependence.Observations based on ice cores reveal a positive trend in the accumulation rate over the Antarctic Peninsula since the 1950s,whereas that over the western part of West Antarctica has significantly decreased during 1900—2010 (Wang et al.,2017b,2019b,2020g).Moreover,the snow melting rate over the Ross Sea region has dramatically decreased in the past several decades (Li et al.,2017b;Zheng et al.,2020).During the same period,both the intensity and frequency of the extreme precipitation events over the Antarctic region have increased significantly,largely due to global warming caused by anthropogenic greenhouse gas emissions (Ren,2002;Xiao et al.,2008;Yu et al.,2018b).

The ocean heat content in the Southern Ocean has increased over the past several decades (Gao et al.,2018;Guo et al.,2019;Xu et al.,2021a),with much of the ocean warming concentrated in the high latitudes of the Southern Oceans (Gao et al.,2018).This increase is tightly associated with the thickening,deepening,and warming of the Subantarctic Mode Water (SAMW) (Gao et al.,2018;Xu et al.,2021a),as well as the shallowing and freshening (Yao et al.,2017) of the Antarctic Intermediate Water.In contrast,although there is extensive warming in the sub-surface Southern Ocean,the surface water shows a cooling trend (Song,2020),which is primarily attributed to the changes in the surface heat flux between the ocean and the atmosphere.Moreover,sea spray over the Southern Ocean may intensify the heat and moisture exchange between the atmosphere and ocean,resulting in a reduction of the surface air temperature(Song et al.,2022).

The changes in Antarctic sea ice show strong regional and seasonal features (Li et al.,2014,2015a,b).The improvement of both the observational techniques (Lei et al.,2009;Shi et al.,2021b;Liao et al.,2022) and the numerical model experiments (Shu et al.,2020;Shi et al.,2021a) help us to better understand the complexity of the Antarctic sea-ice changes.Unlike Arctic sea ice,which experienced a rapid decline in the past decades,the total Antarctic sea-ice extent increased until 2015 (Fig.4) (Yu et al.,2018a;Wang et al.,2019c).The sea-ice concentration around the West Antarctic is accompanied by a redistributed total (areal) expansion,with significant declines over the Weddell Sea,the Bellingshausen Sea,and the Amundsen Sea,but dramatic increases over the Ross Sea (Fig.4) (Li et al.,2014,2015a,b,2021b;Wang et al.,2019c).After 2015,the Antarctic sea ice experienced a rapid decline (Fig.5),with the annual mean total extent of the sea ice decreased by at least 1.6 × 106km2(Wang et al.,2019d).The sea-ice changes around Antarctica may further contribute to the ocean heat content and oceanic stratification.For example,in recent decades,many polynyas (Wang et al.,2017c;Hou and Shi,2021;Wang et al.,2021d,2022d) are triggered (formed) over the Ross Sea,the Prydz Bay,and the Weddell Sea regions,mainly driven by the strengthened wind forcing.These processes favor the formation of Antarctic sea ice,releasing more salinity and further intensifying deep oceanic convection (Guo et al.,2016;Ma et al.,2020b) through the brine effect,contributing to a warming of the Antarctic bottom water (Wang et al.,2017c).

Fig.4.The seasonal Antarctic sea-ice concentration (SIC) changes from 1979 to 2014.The color shadings in (a—d)show the SIC changes for austral spring (SON,a),summer (DJF,b),autumn (MAM,c),and winter (JJA,d),respectively.Dotted regions are significant at p<0.1.

Fig.5.The monthly Antarctic sea-ice extent (SIE) anomalies during 1979—2021.The red line shows the SIE trend from 1979 to 2014.

Over the past several decades,the melting rate of the Antarctic ice sheets has accelerated.This change shows a clear zonally asymmetric feature,with the strongest melting around West Antarctica (Zhang et al.,2021c;Li et al.,2022).Recent studies indicated that the West Antarctic Ice Sheet experienced a rapid mass loss (Wang et al.,2019b;Zhang et al.,2021c),with a strong seasonal feature (Yang et al.,2016b),primarily due to the warm water intrusion from the sub-surface Southern Oceans.The mass loss over the East Antarctic is relatively weak (Zhang et al.,2021c).

The complexity of the Antarctic climate changes is primarily attributed to a combined effect of the anthropogenic ozone loss and greenhouse gas increase (Thompson et al.,2011;Swart et al.,2018;Li et al.,2021b),atmospheric circulation changes (Wang et al.,2019d,2022d;Zhang et al.,2021c),cloud-radiation feedback (Wang et al.,2019c),as well as the atmosphere—sea ice—ocean interaction (Li et al.,2014;Ma et al.,2020b),the sea ice—ocean coupling (Ma et al.,2020b;Shi et al.,2021a),and the sea ice—ice shelf coupling(Guo et al.,2019),etc.

Notably,the adjustment of the atmospheric circulation over the high-latitude Southern Hemisphere plays an essential role in modulating these complicated climate changes over the Antarctic region,mainly through atmospheric thermal advection (Li et al.,2021b) and mechanical forcing (Wang et al.,2019d;Zhang et al.,2021c,2022d).Over the past four decades,the atmospheric circulation changes over the Antarctic have been characterized by a positive phase of the Southern Annular Mode (SAM) and a deepening of the Amundsen Sea Low (ASL) (Ding et al.,2020).The former was mainly driven by the depletion of stratospheric ozone and increased CO2concentrations (Thompson et al.,2011;Swart et al.,2018;Li et al.,2021b),whereas the latter was more closely related to tropical—polar teleconnections (Li et al.,2021b),largely contributing to the climate changes over the West Antarctic including both the rapid warming and the sea-ice redistribution (Ding et al.,2011;Li et al.,2017b,2021b;Wang et al.,2019d;Guo et al.,2020;Zhang et al.,2021c).Furthermore,the interannual variability of the surface wind over Antarctica and the Southern Ocean is tightly associated with the Southern Hemisphere large-scale circulation modes (Yu and Zhong,2019b),while the extreme strong wind events are more prevalent over the coastal region around East Antarctica (Yu and Zhong,2019a).

2.3.Progresses in observational techniques

Because of the unique geological location and the harsh environment,climate observations over the polar regions are difficult to obtain.For a long time,limited observations and relatively low-quality measurements restricted our capability to retrieve the climate change signals over the polar regions.Over the past four decades,a series of modern climate satellites were launched.Meanwhile,China engaged itself in polar expeditions,with a series of new techniques developed to better detect the changing polar environment(Cheng et al.,2022),including a series of polar expeditions with the two icebreakers (“Xuelong” and “Xuelong 2”).A satellite (BNU-1) has been launched recently to provide timely observations and monitor the rapid changes in polar regions (Zhang et al.,2021g).Sled car-based and airplanebased radars have been used to observe the environment under the ice sheet (Sun et al.,2009;Cui et al.,2020).Unmanned aerial systems (UASs) have also been used to resolve the detailed surface features of polar sea ice (Li et al.,2019a) and obtain atmosphere profiles over the Southern Ocean,which may benefit numerical weather forecasts (Sun et al.,2020).Along the PANDA (Prydz Bay-Amery Ice Shelf-Dome A) Transect,CHINARE (Chinese National Antarctic Research Expedition) deployed～650 stakes to monitor the surface mass balance from the coastal Zhongshan Station to Dome A (Minghu et al.,2011;Ding et al.,2015),the summit of East Antarctica.In addition,Automatic Weather Stations (AWSs),applied to extremely cold environments,have also been installed along PANDA Transect.Currently,12 AWSs operate continuously and provide near realtime data to the public (Ding et al.,2022b).

A series of new techniques have been developed to better observe the sea-ice properties over the polar regions (Lei et al.,2009,2017,2022;Wu and Liu,2018;Chen et al.,2022c).A high-precision ice thickness measuring device was invented and applied to observe the landfast ice mass balance in East Antarctica (Lei et al.,2009).Based on the measured data,the influence of tides on sea-ice growth was obtained.A new type of sea Ice Mass balance Buoy (IMB),an unmanned ice station,has been developed,which has been applied to the construction of the buoy array during the Multidisciplinary drifting Observatory for the Study of Arctic Climate (MOSAiC) expedition in 2019/20 (Lei et al.,2022).Compared with the traditional IMB,upper-ocean observations have increased,allowing for the observation and assessment of the effects of the subcooled water under winter ice and the summer freshwater layer on the growth and ablation of sea ice (Zhang et al.,2022c).Combining the shipbased and ice-based observational systems,especially the measurement technology of underwater robots,a three-dimensional observation system for measuring the morphological parameters of sea ice is constructed (Lei et al.,2017).A satellite-based sea-ice navigation system (SatSINS) has been developed (Hui et al.,2017a),which integrates remotely sensed and in-situ observations to optimize the marine navigational routes in sea ice-covered waters.Wang et al.(2022f)compared the sea-ice motion products from the Ocean and Sea Ice Satellite Application Facility and indicated that eleven Arctic sea-ice motion products are of reasonably high quality if the uncertainties are appropriately considered.By combining observations from multiple satellites,a recent study revealed that the sea-ice melting time is linked to the evolution of the summer sea ice (Zheng et al.,2021).

The assessment of the total mass balance of the Arctic and Antarctic ice sheets has also been significantly improved (Wang et al.,2021a),largely due to the launch of altimetry and gravity satellites (Cheng et al.,2015;Xiao et al.,2020b),which have great implications for analyzing the global mean sea level rise.Although the GRACE data tends to underestimate the ice mass loss (Gao et al.,2019a),combining these satellite measurements with the in-situ observations may improve the quality of these mass-balance datasets.Additionally,Yuan et al.(2020) developed an algorithm based on the convolutional neural network to detect the supraglacial lakes over the Greenland ice sheet.

One of the most important examples of progress is a better estimation of the ice flow and the surface melting over the Antarctic and Greenland ice sheets using satellite-based measurements.Cheng and Xu (2006) developed a technique based on four-pass differential synthetic aperture radar(SAR) interferometry aboard the JERS-1 and ERS satellite platforms to retrieve the motion of Antarctic glaciers.This method has been used to detect and validate the ice flow of Grove Mountain and several East Antarctic Glaciers (Cheng et al.,2007).Recently Li et al.(2018) further retrieved the velocity of ice flow over the Antarctic ice sheet using MODIS-based Mosaic datasets.Hui et al.(2017b) produced a database (Antarctic Land Cover Database for the Year 2000,Antarctica LC2000) based on combining computeraided and manual interpretation methods to classify the landcover types over Antarctica.Liao et al.(2019) developed an algorithm to derive the snow depth and ice thickness from the SIMBA (Snow and Ice Mass Balance Array) ice mass balance buoy data.In addition,Zheng et al.(2022) developed a new strategy to quantify the daily surface melting over the Greenland ice sheet based on satellite observations.

Based on ICESat-1/GLAS satellite observations,Liu et al.(2014) developed a method to detect the peak stress points where an ice shelf is most likely to break.This method has been used to investigate the mechanism of iceshelf retreat (Liu et al.,2015) and detect the ice crevasses around the Antarctic ice sheet.Qi et al.(2021) built an iceberg calving product over all Antarctic ice shelves based on this method,which identifies every calving event larger than 1 km2in the past 15 years.

2.4.Rapid acidification in the Arctic Ocean and Southern Ocean

With the climate change in the past decades,the Arctic Ocean has shown widespread ocean acidification and aragonite undersaturation (Ωarag<1).Based on the observations of the Arctic Ocean (47 Arctic research cruises) over the past three decades (1994—2021),Qi et al.(2017) illustrated that low Ωaragwater had expanded northwards to 85°N and deepened to a 250-m depth in the Western Arctic Ocean,and concluded that the Pacific Winter Water (PWW) transport,upon converging and deepening as a result of the intensification of the Beaufort Current eddies,is primarily responsible for the expansion of acidified waters in the upper Arctic Ocean.Later,by estimating changes in pH and Ωaragof the Arctic surface seawater,Qi et al.(2022a) reported a rapid acidification of the Arctic Ocean with a rate 3～4 times higher than other ocean basins.Meanwhile,the mechanism“ice melt—driven enhanced anthropogenic CO2acidification”was unveiled (Qi et al.,2022a): sea-ice melt exposes seawater to the atmosphere and promotes the rapid uptake of atmospheric CO2,lowering its alkalinity and buffer capacity,thus leading to sharp declines in pH and Ωarag(Fig.6).

Fig.6.Conceptual model illustrating warming and climate change-driven sea-ice retreat and associated changes in CO2-system.

As the most productive sea,the Chukchi Sea has been found to contribute more than 50% of the carbon sink in the Arctic Ocean,and the carbon absorption at the surface is still increasing year after year (Tu et al.,2021).However,severe ocean acidification was observed in the subsurface water of the Chukchi Sea.A control mechanism was proposed that the biological activities driven by the Pacific Ocean inflowing water lead to the acidification of subsurface water in the Chukchi Sea (Qi et al.,2020a,2022b).These result s are expected to provide important background information for the ecological assessment of the Chukchi Sea.

In addition,Wu et al.(2021) developed and improved a walk-around pH observation system that adapts to low-temperature environments.Simultaneously,a high-quality dataset was produced by compiling and revising data from twelve Arctic scientific expeditions,thirty-five Antarctic scientific expeditions,and international databases.With that,they further demonstrated the mechanism of the Arctic Ocean acidification,revealing that the Arctic melt-ice dilution and sea-air CO2exchange induced the rapid acidification of the water column,wherein aragonite saturation (a biological indicator of acidification) first decreased to a biohazard threshold (Wu et al.,2021).Aided by the comparison of Southern Ocean data from BGC-Argo and ship-collected datasets,a tool CORS (integrated analysis of seawater dissolved gases CO2and O2) was developed as a quality control for ocean float data.They found that carbon sinks in the high-latitude region of the Southern Ocean were underestimated by 50%by BGC-Argo and proposed that data quality control and correction are necessary for some automatic observation platforms such as buoys (Wu et al.,2022b).

3.Impacts of Tropics and mid-latitudes on polar climate

Climate variability over the tropics and mid-latitudes plays a crucial role in driving Arctic and Antarctic climate changes on intra-seasonal to multi-decadal time scales (Wu,2017;Wang et al.,2018;Li et al.,2019c,2021b;Dou and Zhang,2022;Wu and Ding,2022).In the Southern Hemisphere,atmospheric bridges,especially Rossby wave dynamics,play a key role in these teleconnections (Ding et al.,2011;Li et al.,2014,2021b;Zhang et al.,2021b).The largescale anomalous circulation driven by these teleconnections largely contributes to the asymmetric Antarctic climate changes through a series of atmosphere-ice-ocean interactions (Yuan and Martinson,2000;Li et al.,2014;Yu et al.,2018a;Zhang et al.,2021b).In the Northern Hemisphere,both atmospheric and oceanic processes are crucial in linking the Arctic climate system and the lower latitudes (Hall et al.,2013;Li et al.,2015b;Liu et al.,2018;Wang et al.,2021a).

3.1.Tropics and mid-latitudes impact Antarctic climate

Recently,a series of studies indicated that the tropical and mid-latitude climate variability might be contributing to the Antarctic climate changes (Li et al.,2021b;Zhang et al.,2021b;Chen et al.,2022a;Wang et al.,2022c) on interannual,decadal,and longer timescales.The physical pathways linking the Antarctic climate changes and the lower latitudes are complex (Yang et al.,2020;Zhang et al.,2021b,e;Dou and Zhang,2022),although most of these teleconnections are through the atmospheric bridges,in particular the adjustment of the three-cell circulation,the jets system,and of stationary Rossby waves (Li et al.,2021b;Zhang et al.,2021b,e).In this section,we review how climate variability over lower latitudes may impact the atmospheric circulation adjustment over the high latitudes of the Southern Hemisphere (Li et al.,2020a,2021b;Zhang et al.,2021b),and how these large-scale circulation changes further contribute to the asymmetric climate changes over the Antarctic through a series of atmosphere-ice-ocean interactions (Li et al.,2021b;Zhang et al.,2021e;Dou and Zhang,2022).

On interannual time scales,ENSO plays a dominant role in contributing to climate variability over the high-latitude Southern Hemisphere.During the El Ni?o events,the warm sea surface temperature (SST) anomaly over the eastern tropical Pacific generally triggers a Rossby wave train emanating from the tropical Pacific to the South American region,inducing a shallow Amundsen Sea Low (ASL) (Yuan and Martinson,2000,2001;Li et al.,2020a;Li et al.,2021b)(Fig.7a).The changes of ASL further impact the Antarctic surface air temperature (Li et al.,2021b) and sea ice (Yuan and Martinson,2001;Yuan and Li,2008;Li et al.,2014,2020a;Zhang et al.,2021b,e) through wind forcing and thermal advection (Li et al.,2014,2020a).A recent study indicated that the Rossby wave train triggered by the central Pacific El Ni?o shifted to the west by about 20° compared to an eastern Pacific El Ni?o (Zhang et al.,2021b).In addition,based on CMIP6 simulation results,the projected Southern Ocean warming in the 21st century is tightly associated with changes in the amplitude of ENSO events (Wang et al.,2022a).Meanwhile,the amplitude of the Antarctic dipole is expected to be reduced under greenhouse warming,which is predominantly attributed to the opposing roles of increased ENSO variability and decreased SAM variability (Li et al.,2021a).

Fig.7.Teleconnection patterns triggered by interannual and decadal sea surface temperature variability,respectively.(a) Schematic of tropical—Antarctic atmospheric teleconnection patterns induced by El Ni?o events.(b) Schematic of the atmospheric teleconnection patterns induced by a positive phase of the Atlantic Multidecadal Oscillation (AMO;red arrow) and a negative phase of the Interdecadal Pacific Oscillation (IPO;blue arrow).

The mechanisms outlined above operate not only for the ENSO-polar connection but also for tropical-polar connections from other ocean basins (Li et al.,2014;Rao and Ren,2020;Yang et al.,2020;Chen et al.,2022a).For example,Rossby wave trains can be generated by convective heating associated with the tropical Indian Ocean (Rao and Ren,2020;Yu et al.,2022),the Maritime Continent (Chen et al.,2022a),and the tropical and North Atlantic (Li et al.,2014).In addition,these mechanisms operate on various time scales.The negative trend of the South Pacific Oscillation index in austral autumn,triggered by an anomalous wave train from the Indian Ocean to the Southern Ocean,can,in part,induce decreased SIC in the Bellingshausen Sea and increased SIC over the Ross Sea (Yu et al.,2021).On intraseasonal timescales,Rossby wave trains excited by the Madden-Julian Oscillation (Yang et al.,2020) propagate to southern high latitudes and can influence temperature and sea ice in as quickly as a few days to a week (Wang et al.,2022b,c).

On decadal and longer time scales,the Interdecadal Pacific Oscillation (IPO) (Ding et al.,2011) and the Atlantic Multidecadal Oscillation (AMO) (Li et al.,2014,2021b)play a key role in driving the multi-decadal changes over the Antarctic through similar mechanisms,namely the adjustment of the stationary Rossby wave trains (Li et al.,2021b)(Fig.7b).During the positive phase of the IPO,the anomalous warm SST over the eastern tropical Pacific triggers a Rossby wave train that induces a shallow ASL (Ding et al.,2011;Li et al.,2021b).During the positive phase of the AMO,the north and tropical Atlantic warming generates a Rossby wave train,which is guided around the Southern Oceans by the subtropical jet,ultimately influencing the atmospheric circulation around Antarctica,intensifying the SAM,and deepening the ASL (Li et al.,2014,2015a,b,2021b).

Notably,these interannual and interdecadal teleconnections are not independent of one another but rather closely interact (Ding et al.,2016;Dou and Zhang,2022).In addition,interdecadal shifts exist in these interactions,including the interdecadal changes in the lagged relationship between the summer Pacific—South American (PSA) pattern and subsequent ENSO events in the following summer (Ding et al.,2016),as well as the weakening of the ENSO-Antarctic dipole relationship over the past two decades (Dou and Zhang,2022).

Moreover,midlatitude climate variability may also influence the Antarctic climate (Wen et al.,2021;Zhang et al.,2021e).Recent studies have revealed that the warm (cool)SST anomalies in the western South Atlantic can increase the SIC over the eastern Ross Sea in austral spring while,at the same time,decreasing the SIC over the northwestern Weddell Sea (Zhang et al.,2021e).In addition,the model simulation shows that long-term changes in the Antarctic bottom water (AABW) can be affected by remote forcing from topographic changes of the Tibetan Plateau and also through Rossby wave dynamics (Wen et al.,2021).In particular,increased anthropogenic aerosol concentration in the Northern Hemisphere can induce atmospheric circulation changes in Southern Hemisphere,weakening the subtropical and the subpolar jets in austral winter (Wang et al.,2020b).

These teleconnections modify the large-scale circulation over the high-latitude Southern Hemisphere and subsequently impact the Antarctic climate variability through a series of atmosphere-ice-ocean interactions (Yuan and Martinson,2001;Li et al.,2014,2021b;Zhang et al.,2021e).The Antarctic SIC is strongly influenced by the remote forcing from the tropical and midlatitudes (Li et al.,2014,2020a;Zhang et al.,2021b,e;Dou and Zhang,2022).La Ni?a events (Li et al.,2020a;Zhang et al.,2021b),the negative phase of IPO (Li et al.,2020a),and a positive phase of AMO (Li et al.,2014) lead to a deepening of the ASL through a Rossby wave train,and vice versa.The deepened ASL can further increase the SIC over the western Ross Sea and decrease the SIC over the Amundsen,Bellingshausen,and Weddell Seas(Li et al.,2014,2020a;Zhang et al.,2021b),a phenomenon known as the Antarctic sea-ice dipole.Moreover,the Antarctic SIC is also modulated by the Southern Annular Mode(SAM) (Zhang et al.,2018b;Li et al.,2020a,2021a),which is closely related to the Antarctic internal variability (Wang et al.,2020f).

The tropical-polar teleconnections may also contribute to the ocean bottom pressure (OBP) in the South Pacific,with a positive OBP anomaly related to the positive phase of the second Pacific South American mode (PSA2) (Qin et al.,2022).Temperature extremes on the Antarctic Peninsula are closely related to the intraseasonal oscillations and the fast synoptic-scale Rossby wave (Wang et al.,2022c).Additionally,tropical volcanic eruptions (Wu et al.,2018;Liu et al.,2020a) may also impact the Antarctic surface temperature by increasing sulfur aerosols in the Antarctic stratosphere(Wu et al.,2018).

3.2.Tropics and mid-latitudes impact Arctic climate

The tropics and the mid-latitude climate variability may impact Arctic climate changes through atmospheric and oceanic pathways (Ding et al.,2014;Feng and Wu,2015;Hu et al.,2016;Liu et al.,2021b;Fang et al.,2022;Wu and Ding,2022),whose process is even more complicated than that of the tropical—Antarctic teleconnection.

The tropical SST variability may contribute to the Arctic SAT changes on interannual and decadal time scales through atmospheric bridges.The tropical Pacific decadal variability may generate stationary Rossby wave trains and influence the atmospheric circulation around the high-latitude Atlantic (Ding et al.,2014),contributing to a prominent surface and tropospheric warming around Northeastern Canada and Greenland since 1979.The SST cooling in the tropical central and eastern Pacific may also contribute to the recently observed Arctic warming anomalies (Wu,2017).Recent studies indicated that the East Pacific El Ni?o events may have led to a cooling anomaly over the Barents-Kara Seas in February,while the Central Pacific events usually contribute to an SAT warming over northeastern Canada and Greenland (Li et al.,2019c).The equatorial Pacific warming related to Central Pacific El Ni?o events could inhibit summer Arctic warming and sea-ice melting by deepening the tropospheric Arctic polar vortex and enhancing the circumpolar westerly wind (Hu et al.,2016).The AMO may contribute to the evolution of the Arctic Amplification on multidecadal and longer timescales (Fang et al.,2022).On the other hand,a recent study reveals a multidecadal seesaw of the cold wave frequency between central Eurasia and Greenland,which is likely driven by the AMO (Liu et al.,2022).

The atmospheric teleconnections may also impact the Arctic sea-ice extent.The decline of summer-time Arctic sea ice has accelerated over the past four decades,which is,in part,driven by thermal advection and the mechanical forcing associated with the negative phases of the Arctic Oscillation and the PDO,as well as the positive phases of the NAO,the Arctic Dipole,and the AMO (Cai et al.,2021).It has been suggested that the recent persistent positive PNA pattern has led to increased heat and moisture fluxes,accelerating the sea-ice decline over the western Arctic Ocean (Liu et al.,2021b).In contrast,the phase changes of the PDO may affect the Arctic dipole pattern,modulating the variations of summertime SIC over the Pacific sector.In particular,the PDO recently shifted to its positive phase,which may temporarily slow down the observed decline of the summertime SIC within the Pacific sector of the Arctic Ocean (Bi et al.,2021).In addition,the Arctic sea-ice variability during the melt season is tightly associated with the phase change of the AMO (Yu et al.,2019a).On the other hand,the enhancement of North Atlantic westerly winds associated with the SST dipole to the south and north of the Gulf Stream Extension may promote long-lived Ural blocking and high-latitude European blocking,further contributing to the loss of sea ice in the Barents-Kara Seas and also to Eurasian cooling(Luo et al.,2019).

Oceanic processes represent an important pathway by which features in the tropics and mid-latitudes can induce Arctic climate changes (Zhang,2015;Shu et al.,2019,2021;Zhang et al.,2022e).Zhang (2015) revealed that three key factors,including the oceanic heat transport from the Atlantic and the Pacific Oceans and the Arctic Dipole pattern,have played a crucial role in driving the summer-time Arctic seaice decline in recent decades.The Atlantic water is particularly important in the thermal balance of the Arctic Ocean(Shu et al.,2019).North Atlantic warming is usually associated with a warmer Arctic Ocean and sea-ice declines(Zhang et al.,2022e).The Barents-Kara Seas releases most of the ocean heat that originates from the North Atlantic and is a known cooling machine of the Arctic Ocean.It has been reported that this cooling machine has expanded poleward,a phenomenon termed the poleward shift of the Arctic Atlantification (Shu et al.,2021).The transport of heat and salinity from the Pacific also significantly contributes to the climate variability of the Arctic Ocean.For example,the variability of the Bering Strait inflow largely contributes to the recent freshening of the Arctic Ocean (Shu et al.,2018).

On the other hand,the Arctic Ocean may feedback upon the Atlantic and the Pacific by adjusting their ocean circulations.In particular,the freshwater outflow from the Arctic Ocean plays an important role in modulating the Atlantic Meridional Overturning Circulation (AMOC).A recent study (Chen and Tung,2018) indicated that the role of AMOC may be altered in the presence of greenhouse-gas heating.Usually,the AMOC transports surface heat northwards,heating Europe and North America.Under global warming,the AMOC may start to store heat in the deeper Atlantic,buffering the surface warming for the planet as a whole(Chen and Tung,2018).On the other hand,according to CMIP model simulations,the projected decline of anthropogenic aerosols may also potentially weaken the AMOC,leading to enhanced oceanic heat uptake in the subpolar North Atlantic (Ma et al.,2020c).

Simulation experiments forced by winter Eurasian regional ground albedo changes consistently demonstrate that regional cooling directly contributes to Arctic warm anomalies.Sometimes the cooling can enhance the Siberian high and generate positive 500-hPa height anomalies over and around the Ural Mountains through atmospheric subsidence anomalies and wave-energy propagation,ultimately leading to warm anomalies over the Barents-Kara Seas (Wu and Ding,2022).

4.Polar changes contribute to a changing global climate system

Climate change over the Arctic and Antarctic regions may both impact the lower latitudes,primarily through the adjustment of the large-scale atmospheric circulations and associated atmosphere-ice-ocean interactions.Both the Greenland and Antarctic ice sheets experienced accelerated melting during the past three decades,contributing to a rise in global mean sea level.These global impacts of the polar climate changes are reviewed as follows.

4.1.Arctic changes impact lower-latitudes

Rapid warming and sea-ice loss over the Arctic are associated with anomalous cooling over the mid-and high-latitude continental areas of the Northern Hemisphere for boreal winter and spring.This dipole-like,anomalous SAT pattern is pronounced from interannual to multi-decadal timescales,known as the “warm Arctic cold continental” pattern (Luo et al.,2016;Wu,2017;Gu et al.,2018;Ding et al.,2021).In particular,the cooling over the central Eurasian continent is linked to warming in the Arctic,named the “warm Arcticcold Eurasia” pattern (Xu et al.,2019;Dai and Song,2020;Dai and Deng,2022) (Fig.8).The Arctic sea-ice loss,especially those over the Barents-Kara seas (Fig.2a),releases additional heat and moisture from the ocean to the atmosphere,which can increase Siberian snow cover and further enhance the Siberian high (Wu et al.,2011;Zuo et al.,2016).This anticyclonic circulation anomaly transports cold air from the Arctic to the mid-latitude Eurasian continent(Fig.1a and Fig.8).The retreat of the Barents-Kara sea ice can trigger a surface temperature dipole pattern between Eurasia and North America by adjusting the mid-and high-latitude atmospheric circulation in boreal winter (Hou et al.,2022).In addition,the anomalous snowfall over northern Eurasia may mediate the effect of the Arctic sea ice on Eurasian cooling (Xu et al.,2018).The autumn sea-ice loss over the Arctic Ocean may contribute to an anomalous “north-south” dipole mode of the snow cover over the mid-to-high latitude Eurasian continent,potentially contributing to the winterspringtime cold events over mid-latitude Eurasia (Zhang et al.,2019a,2022g).Moreover,the phase changes of the AMO may influence the effect of the Arctic sea-ice loss on the anomalous cooling over Siberia (Luo et al.,2017;Chen et al.,2021).

Fig.8.Schematic diagram of the mechanism for the Warm Arctic—Cold Eurasia(WACE) pattern associated with sea-ice loss over the Barents-Kara Seas in boreal winter.Declines in sea-ice heat the surface and tropospheric atmosphere and excite an anomalous high-pressure center over northern Siberia,driving cold advection from the Arctic to central Eurasia.

In particular,the Arctic sea-ice loss may change winter weather patterns and the frequency of extreme cold events over the Eurasian continent.During wintertime,the intensity of the tripole wind pattern and the frequency of its extremely negative phase over Eurasia is significantly correlated with autumn Arctic sea-ice anomalies (Wu et al.,2013).The reduction of the meridional temperature gradient associated with Arctic sea-ice loss leads to the vertical propagation of planetary wave energy,a weakening of the stratospheric polar vortex,and the genesis of an atmospheric blocking system over the Ural regions (Luo et al.,2018).The winter warming in the Barents-Kara seas and the associated sea-ice declines may contribute to a large increase in the mean duration of the Ural blocking events,thus affecting the recent winter Eurasian cold extreme events (Luo et al.,2018).The Arctic atmospheric circulation anomalies during the summertime and the Arctic sea-ice declines may also impact the intraseasonal evolution of extreme cold events over Eurasia (Wu et al.,2017).Notably,the zonal movement of the Ural blocking exerts important feedback on the sea-ice variability in the Barents-Kara seas (Chen et al.,2018).The Arctic polar vortex shifted persistently towards the Eurasian continent in February over the past three decades,which induced a cooling anomaly over the mid-latitude Eurasian continent (Zhang et al.,2016).In addition,the phase change of the PDO may contribute to the influence of the Arctic sea-ice loss on the extreme winter events over the Eurasian continent (Zhang et al.,2022d).Recent studies have shown that in recent decades,the linkage between the Arctic sea-ice variability,the Arctic Oscillation,and the Eurasian extreme cold events has significantly intensified (Chen et al.,2019;He et al.,2019;Ding and Wu,2021).

Recent studies revealed a seesaw pattern of extreme temperature events on sub-seasonal timescales (Ma and Zhu,2020),with a cold event over East Asia accompanied by a warm event in North America,usually lasting for several weeks (Sung et al.,2021).In addition,increased wildfire events over the western U.S.in boreal autumn are fueled by more fire-favorable weather associated with Arctic sea-ice declines during the preceding months on both interannual and interdecadal time scales,inducing regional circulation changes accompanied by a poleward shift of the polar jet stream (Zou et al.,2021).

It has been reported that an increase of the winter seaice extent over the Greenland-Barents Seas may trigger an atmospheric wave train propagating southeastward from high-latitude Eurasia towards the subtropical North Pacific,with cyclonic wind anomalies over the subtropical North Pacific,to include the induction of an El Ni?o event in the following winter (Chen et al.,2020a).One recent study (Chen et al.,2020b) found that the relationship between the spring Arctic Oscillation and winter-time ENSO events has weakened since the early 1990s.The boreal winter Arctic Oscillation is also reported to have an interannual relationship with the summer SST anomalies over the western tropical Indian Ocean (Gong et al.,2017) and may impact October East African precipitation (Gong et al.,2016).

Additionally,the Arctic-midlatitude linkage also displays strong phasic fluctuations (or decadal fluctuations),and Arctic sea-ice loss may be conducive to the alternative occurrence of warm Arctic-cold Eurasia (2007/2008—2012/2013) and warm Arctic-warm Eurasia (2013/2014—2018/2019) events (Wu et al.,2022a).

The Arctic changes may impact climate variability over East Asia.It is reported that the Arctic Oscillation may work together with the ENSO system,contributing to the occurrence of cold events over East Asia during boreal winter(Song and Wu,2022).Both the negative phase of the Arctic Oscillation in El Ni?o winters and the Ural blocking in La Ni?a winters are followed by a Rossby wave train over Eurasia continents that strengthens the Siberian high,leading to cold anomalies over East Asia (Song and Wu,2022).The winter sea-ice loss over the Barents-Kara Seas also largely contributes to the hiatus of winter warming in China (Li et al.,2019b).The interactions between the Arctic sea-ice anomalies and the SST anomalies in the equatorial central-eastern Pacific are important in influencing the East Asian winter monsoon system (Sun et al.,2016a;Zhang et al.,2020).In addition,the widespread North Atlantic-Arctic warming could modulate and interact with the atmospheric background flow over the North Atlantic-Arctic Ocean,which further enhances the quasi-stationary Rossby waves and results in a strengthening of the East Asian winter monsoon and cooling over East Asia (Zhang et al.,2022e).

In addition,declines in Arctic sea ice play an important role in the summer atmospheric circulation and the extreme climate events over East Asia (Wu and Li,2022).The Arctic sea-ice loss may intensify the East Asia summer monsoon(Wu and Li,2022),deepening the low-pressure system over the Eurasian continent and increasing the frequency of the heat waves and heavy precipitation over the mid-latitudes of East Asia (Wu and Li,2022).The winter-time Arctic seaice variability plays an important role in the influence of the North Atlantic Oscillation on the East Asia summer-time precipitation (Zhang et al.,2021f).The combined effects of Indian Ocean warming and Arctic sea-ice loss also contributed to a record-breaking Meiyu-Baiu rainfall over East Asia in June-July 2020 through enhanced mid-tropospheric westerlies and anomalous meridional wind convergence(Chen et al.,2022b).In addition,the anomalous cooling over the mid-and low-level troposphere over the Arctic region favors a heat wave event over East Asia and Europe in boreal summer (Wu and Francis,2019).

Furthermore,extreme cold events over the Tibetan Plateau are affected by Rossby wave trains triggered by Arctic sea-ice loss (Bi et al.,2022),and they are regulated by seaice variability in the Beaufort and Laptev Seas (Bi et al.,2022).The Arctic sea-ice loss may also trigger stationary Rossby wave trains,propagating to the Tibetan Plateau area and impacting its air pollution by changing the atmospheric advection (Bi et al.,2022).

The winter sea-ice cover over the Barents Sea has been reported to have an interannual linkage with the spring Normalized Difference Vegetation Index over Eurasia (Ji and Fan,2019a).They are also important predictors of dust weather frequency in East Asia,particularly over northern China (Ji and Fan,2019b;Liu et al.,2020b).In addition,the positive phase of the preceding wintertime Arctic polar vortex intensity tends to increase the spring Normalized Difference Vegetation Index in Europe and Lake Baikal but may lead to a significant decrease in this parameter over Siberia (Li et al.,2017a).

4.2.Antarctic changes impact lower latitudes

The Antarctic climate variability may also largely impact the lower latitude climate (Bai et al.,2016;Xiang et al.,2018;Fu et al.,2019;Jiang et al.,2022).Antarctic climate changes can directly affect tropical climate variability through atmospheric bridges (Xiang et al.,2018;Jiang et al.,2022).For example,the Antarctic sea ice significantly impacts the global air temperature by changing the intensity of atmospheric baroclinic disturbances to modify wave energy transmission (Jiang et al.,2022).Antarctic ozone depletion also plays an important role in increasing the Southern Hemisphere extratropical precipitation (Bai et al.,2016).On the other hand,the Antarctic climate variability can transport energy to the upper-layer ocean over the mid-high latitude Southern Hemisphere,affecting SST variability,which further triggers the remote climate effect through atmospheric teleconnections (Hwang et al.,2017;Xiang et al.,2018;Yang et al.,2018;Fu et al.,2019).

The Southern Oceans,a major region for global heat uptake,has a pronounced influence on global-scale atmospheric circulations,tropical SST variability,and tropical precipitation (Hwang et al.,2017).In addition,the Southern Ocean may affect the position of the ITCZ from the perspective of global energy balance (Xiang et al.,2018).Perturbing the Southern Ocean with an external heat forcing may drive cross-equatorial ocean heat transport and lead to a shift of the ITCZ (Liu et al.,2021a).In addition,based on the CMIP6 simulation results,the projected changes of the extratropical surface heat flux associated with the AMOC and Southern Ocean SIC play an important role in the tropical climate response to global warming (Geng et al.,2022).Antarctic Intermediate Water,originating from the Southern Oceans,also has a potential impact on interannual and interdecadal variability of the Indonesian Throughflow (Yang et al.,2018),as well as the oceanic climate variability around the western boundary current (Fu et al.,2019).

Antarctic climate variability also has a pronounced impact on the weather and climate systems around China(Fang et al.,2019;Chang et al.,2020;Dou et al.,2020;Yuan et al.,2022).An interdecadal dipole pattern between the summer geopotential heights in Northeast Asia and the Antarctic (Fig.9) has been revealed in a recent study (Fang et al.,2019),which provides important evidence that the atmospheric circulation around the Antarctic may influence China.The Antarctic Oscillation may impact the precipitation and temperature of Southern China,especially around the Yangtze River Basin (Fig.10) (Dou et al.,2020;Yuan et al.,2022).In spring,the enhanced atmosphere-ocean heat flux associated with the Antarctic Oscillation may intensify a meridional Indian Ocean tripole mode and thus strengthen precipitation over the Maritime Continent area,further influencing the summer precipitation in Southern China (Fig.10)through atmospheric bridges (Dou et al.,2020).In addition,the autumn Antarctic Oscillation may regulate both the drycold northerly advection and the moist-warm southerly advection in both the troposphere and the stratosphere,further influencing the winter precipitation and temperature of Southern China (Yuan et al.,2022).Moreover,the winter-time haze pollution in central-eastern China is closely related to the tropospheric and stratospheric polar vortices in the Southern Hemisphere in August and September (Chang et al.,2020).

Fig.9.The first Principal Component Analysis (PCA) mode of the low-passed (f <0.1) decadal sea level pressure (SLP,a)and 850-hPa geopotential height (GPH850,b) anomalies during 1948—2021 for boreal summer (JJA).[Reprinted from(Fang et al.,2019),reproduced with permission from Springer Nature].

Fig.10.The correlation pattern between the boreal summer(JJA) precipitation over China and the SAM index in May during 1988—2012.The positive (negative) correlation coefficients are shown in solid (dashed) contours.The color shadings reflect the degree of statistical significance using the students’ t-test.[Reprinted from (Dou et al.,2020),reproduced with permission from Springer Nature].

4.3.Polar climate changes contribute to global sea level rise

One of the most important effects of polar climate changes is their impact on the global mean sea level rise.During the last two decades,the melting of ice sheets and the increase of the ocean heat content dominated the trend of the global mean sea level rise (Li et al.,2013;Chen et al.,2014;Yi et al.,2015;Chen et al.,2017;Yi et al.,2017).Recent studies indicated that the GMSL experienced an acceleration from 1993 to 2014,whose ratio increased from 2.2 ±0.3 mm yr—1(1993) to 3.3 ± 0.3 mm yr—1(2014),with the peak value occurring in 2004 and a slight deceleration happening in the recent decade (Chen et al.,2014,2017).Over the past two decades,ice sheet melting over the polar region mainly contributed to the GMSL rise (Chen et al.,2017).An accelerated mass loss has been observed around the Greenland and Antarctic ice sheets (Gao et al.,2015;Zheng et al.,2022).At present levels,the polar ice sheet melting plays a more important role in the acceleration of the global sea level rise,in comparison to that of the seawater expansion caused by the increased ocean heat content (Xie et al.,2016b;Chen et al.,2017).

Over the past two decades,glacial melt around the Greenland ice sheet has been dramatically accelerated.Greenland is one of the largest contributors to the global sea level rise(Chen et al.,2017).From 1993 to 2014,the acceleration of global mean sea level rise has been faster than in previous decades,which can be largely attributed to the freshwater discharge from the Greenland ice sheet,whose contribution increased from 5% of the global sea level change rate (1993)to more than 25% (Chen et al.,2017).Recent studies indicated that the oceanic warm water intrusion dominates the accelerated basal melting of the Jakobshavn Glacier,one of the largest outlet glaciers in Greenland (Wang et al.,2020c).In addition,the glacial melt around Greenland may also be driven by a positive phase of the summer North Atlantic Oscillation (Ruan et al.,2019)

Due to the rapid melting and collapse of the West Antarctic ice shelf,the contribution of the Antarctic ice sheet to sea level rise has increased from 0.08 (—0.10 to 0.27) mm yr—1in 1992—2001 to 0.40 (0.20—0.61) mm yr—1in 2002—2011,whereas the enhanced Antarctic ice sheet mass loss was partially offset by increased snowfall over East Antarctica(Ding et al.,2017).The long-term mass loss of the Antarctic ice sheet is primarily attributed to oceanic basal melting rather than the surface mass balance,noting that the strongest basal melting occurred around the West Antarctic and the Antarctic Peninsula (Gao et al.,2019b;Zhang et al.,2021a).On the other hand,the mass loss of the East Antarctic ice sheet is mainly concentrated near the Poinsett Cape,which is much weaker than that of the West Antarctic ice sheet (Li et al.,2016).Under the background of global warming,the size of the Antarctic ice sheet may continue to decrease,considering that the frequency of the ice calving events continues to increase all around the Antarctic (Li et al.,2016,2020b;Gao et al.,2019b;Qi et al.,2020b,2021).

In addition to these regional effects,the increase in ocean heat content (Chen et al.,2013;Chen and Tung,2018)and the tropical—polar teleconnection may also impact the glacial melt around the Arctic and the Antarctic (Li et al.,2021b),and thus further contribute to the global mean sea level rise.Moreover,it has been reported that the mountain glaciers in Alaska are also melting rapidly due to the recent climatic change,which greatly contributes to the global sea level rise (Jin and Feng,2016).

5.The role of the stratospheric circulation

The stratospheric climate variability is key in the Arctic and Antarctic climate systems.It directly interacts with the variabilities in the troposphere and further mediates the interactions between the polar climate systems and the lower latitudes in both hemispheres,therefore largely impacting the global climate system (Fig.11).

Fig.11.Schematic diagram showing the influence of the stratospheric circulation on the global climate system.The stratospheric circulation interacts with the troposphere in a coupled way.The green arrow shows how the tropospheric circulation affects the stratosphere.Through tropospheric processes,circulation anomalies in the polar,mid-latitudes,and tropics can affect the upward propagation of the Rossby waves,further affecting the polar stratospheric circulation.In addition,stratospheric processes can also affect polar circulations,as shown by the blue arrows.The orange arrows show the influence of the stratospheric circulation on the troposphere.Stratospheric circulation anomalies can directly affect midlatitudes and the polar regions through wave-mean flow interaction,wave reflection processes,and radiative processes and proceed to further affect tropical regions through tropospheric processes.

The stratospheric polar vortex (SPV) is the dominant stratospheric circulation system,usually active from autumn to spring in each hemisphere.Its variability can dramatically affect the tropospheric circulation,including the southern/northern annular mode and the Amundsen Sea Low,etc.Better estimation of the SPV may help to improve forecast skill and future projections of the polar climate system (Rao et al.,2021).

In the Northern Hemisphere,the stratospheric circulation anomalies associated with SPV changes can directly descend into the troposphere through stratosphere—troposphere coupling processes,including wave-mean flow interaction and the reflection of planetary waves (Gong et al.,2018;Wang and Ting,2022).This interaction may significantly affect the climate and weather conditions over the Northern Hemisphere,leading to anomalous cold-air outbreaks (Cheung et al.,2016;Huang et al.,2021a),widespread precipitation (Zhang et al.,2022a),atmospheric environmental conditions (e.g.,PM2.5concentration)(Huang et al.,2021b;Lu et al.,2021),and even tropical deep convection (Wang et al.,2020a).Accordingly,it serves as an important pathway linking the circulations over the Arctic,extratropics,and tropical regions (Ding et al.,2021;Xu et al.,2021b).However,the interaction between the stratospheric and tropospheric circulation anomalies over the Northern Hemisphere largely depends on the circulation condition in both the stratosphere and the troposphere(Yu and Ren,2019;Zhang et al.,2019b;Xu et al.,2022).

Given that the Northern Hemispheric SPV exerts such an important impact on the global climate system,many efforts have been devoted to exploring the possible causes of its variation on different timescales.On the subseasonal timescale,the Madden-Julian Oscillation can significantly affect the SPV through Rossby wave dynamics (Yang et al.,2019a;Ma et al.,2020a).On the interannual timescale,both the equatorial stratospheric Quasi-Biennial Oscillation(QBO) and El Ni?o-Southern Oscillation can influence the SPV by modulating the wave propagation and affecting the intensity of tropospheric wave forcing,respectively (Dai and Tan,2016;Zhou et al.,2018;Zhang et al.,2022f).On decadal timescales,the SST anomalies over the North Atlantic and the central North Pacific may significantly enhance the intensification of the SPV (Hu et al.,2018,2019).Meanwhile,the SPV has shifted towards the Eurasian continent in recent decades,largely due to the recently observed declines of Arctic sea ice,especially around the Barents-Kara Seas (Zhang et al.,2016).

Although the Southern Hemispheric SPV is less volatile compared to its Northern Hemispheric counterpart,recent studies provide evidence showing that it can still exert remarkable impacts on the tropospheric circulation(Shen et al.,2020a;Wang et al.,2020d).In particular,extremely weak SPV events can sometimes occur attributed to the joint effect of the QBO and anomalous tropospheric waves (Rao et al.,2020;Shen et al.,2020b).The anomalous circulation further descends into the troposphere and persists for months,significantly influencing the hemispheric-scale weather condition and the Antarctic sea-ice extent (Wang et al.,2021b;Shen et al.,2022a).

A main driver of the observed high-latitude stratospheric climate change is the variation of the stratospheric ozone concentration,particularly the ozone hole depletion and its recent recovery over the high latitudes of the Southern Hemisphere.In the Southern Hemisphere,the stratospheric ozone concentration has shown preliminary signs of recovery since the early 2000s (Bian et al.,2012;Hu et al.,2022b).Its recovery has been suggested to contribute to stratospheric warming in recent years and may be responsible for the more frequent occurrence of weak SPV events (Xia et al.,2020a;Shen et al.,2022a).In addition,the Antarctic ozone anomaly greatly impacts the Antarctic sea-ice extent via the radiative process and the ice-albedo feedback (Xia et al.,2020b).

In the Northern Hemisphere,the stratospheric ozone anomaly impacts the tropospheric climate variability through two main pathways.The ozone concentration change can directly influence the atmospheric temperature through its radiative forcing (Xia et al.,2021).It may also induce hemispheric-scale circulation change in the troposphere by modulating the SPV (Xie et al.,2016a).The physical processes by which changes in ozone concentration impact the polar atmospheric circulation are complicated.The Arctic sea-ice concentration,the background circulation over the mid-latitudes and tropics,as well as the hydrological cycle may all influence the processes through their adjustment of the local circulation,the atmosphere—ocean interaction,and the meridional overturning circulation (Xie et al.,2017;Hu et al.,2022a;Wang et al.,2022e;Zhang et al.,2022b).Many studies have also attempted to identify possible factors influencing stratospheric polar ozone concentrations on different timescales.It has been found that tropospheric circulation anomalies,including SST anomalies,sea-ice loss,and the MJO,can all contribute to the observed ozone variability(Tian et al.,2017;Zhang et al.,2018a;Yang et al.,2020).

6.Conclusion and Discussion

In the past four decades since the modern satellite era,a series of polar climate changes have been observed.These changes include a rapid Arctic warming that is more than twice as fast as the global warming rate,known as the Arctic Amplification,and zonally asymmetric climate changes over the Antarctic,with the fastest surface warming,seaand land-ice melting all happening around the West rather than the East Antarctic.

These polar climate changes on the multi-decadal scale have been primarily driven by anthropogenic greenhouse gas increases and ozone loss,while the regional features of the polar climate changes and the more frequent extreme events are largely attributed to the tropical—polar teleconnections.On the other hand,polar climate changes significantly impact the Earth’s climate system in several aspects.First,the accelerated melting of the Greenland and Antarctic ice sheets is a primary cause of global-mean sea level rise.In addition,the Arctic warming tends to cool the land areas over the mid-latitude Northern Hemisphere.The Arctic and Antarctic sea-ice changes may also impact the precipitation and large-scale circulation over the tropics and mid-latitudes.The adjustment of the atmospheric circulation in both the troposphere and the stratosphere,as well as the oceanic processes,all contribute to the interactions between the polar climate changes and climate variability over lower latitudes.

An important mission of the polar climate investigation is to improve observational systems around the Arctic and the Antarctic,as many of the polar climate changes and their associated variability last for decades or longer.Thus,long-lasting,high-quality measurements over high latitudes are needed to investigate these variations.However,unlike tropical and mid-latitude regions,observations over polar regions are comparatively difficult to obtain because of the harsh environment.Although the assessment of many variables,including the temperatures of the ocean and atmosphere,sea level,and sea-ice concentrations,have become available during the modern satellite era,most satellites do not fly over the North or South Pole,leaving a big gap in the observation coverage.The lack of polar satellite observations leads to observational and reanalysis datasets with compromised quality and reliability around the Arctic and Antarctic regions,which need to be improved in future investigations.

Considering that the decadal and multi-decadal variability over the tropics and the mid-latitudes,including the PDO and AMO,all contributed to Arctic and Antarctic climate changes,the phase change of these modes of decadal variability may also lead to a phase change of the polar climate changes.For example,the zonally asymmetric Antarctic climate change,with the West warming and East cooling,may potentially switch in the future following the phase reversals of the AMO and the PDO.The interaction between the longterm trend and the decadal/multi-decadal variability may largely contribute to future polar climate changes and consequently influence the predictability and the projection of the Earth’s climate system,which represents an issue that warrants further investigation.

In addition,regarding the interactions between the polar climate and the lower latitude climate variability,we found that more attention was paid to the impact of the Arctic Amplification on lower latitudes and the contributions of the tropics (and mid-latitudes) to the Antarctic climate.The converse effects of the Antarctic on the lower latitudes and the impacts of the tropics on the Arctic climate changes involve multiple interactions between the ocean,the atmosphere,and the cryosphere,which is relatively more complicated and less understood.Such linkages also require further investigation in future studies.

Acknowledgements.X.LI was supported by the National Key Research and Development Program of China (2018YFA 0605703),the National Natural Science Foundation of China (No.41976193 and No.42176243).X.CHEN was supported by the National Key Research and Development Program of China(2019YFC1509100) and the National Science Foundation of China(No.41825012).B.WU was supported by the Major Program of the National Natural Science Foundation of China (41790472),the National Key Basic Research Project of China (2019YFA0607 002),and the National Natural Science Foundation of China(41730959).X.CHENG was funded by the Innovation Group Project of Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai) (Grant No.311021008).M.DING was supported by the National Natural Science Foundation of China(42122047 and 42105036) and the Basic Research Fund of the Chinese Academy of Meteorological Sciences (2021Y021 and 2021Z006).Q.SUN was supported by the National Key R&D Program of China (No.2022YFE0106300).