Editing IPCC:AR6/WGI/Chapter-9 (section)

== 9.3 Sea Ice ==

<div id="9.3.1" class="h2-container"></div>

<span id="arctic-sea-ice"></span>
=== 9.3.1 Arctic Sea Ice ===

<div id="h2-15-siblings" class="h2-siblings"></div>

<div id="9.3.1.1" class="h3-container"></div>

<span id="arctic-sea-ice-coverage"></span>
==== 9.3.1.1 Arctic Sea Ice Coverage ====

<div id="h3-15-siblings" class="h3-siblings"></div>

The observed decrease of Arctic sea ice area is a key indicator of large-scale climate change ( [[IPCC:Wg1:Chapter:Chapter-2#2.3.2.1.1|Section 2.3.2.1.1]] , Cross-Chapter Box 2.2). The SROCC ( [[#Meredith--2019|Meredith et al., 2019]] ) assesses that sea ice extent, which is the total area of all grid cells with at least 15% sea ice concentration, has declined since 1979 in each month of the year ( ''very high confidence'' ). In contrast to SROCC, we assess changes in sea ice area (the actual area of the ocean covered by sea ice) rather than sea ice extent, because sea ice area is geophysically more relevant and not grid-dependent ( [[#Notz--2014|Notz, 2014]] ; [[#Ivanova--2016|Ivanova et al., 2016]] ; [[#Notz--2016|Notz et al., 2016]] ; Notz and SIMIP Community, 2020). Arctic sea ice area is calculated based on measurements by passive microwave satellite sensors that provide near-continuous measurements of gridded, pan-Arctic sea ice concentration from 1979 onwards. Irreducible uncertainties in the conversion of thermal microwave brightness temperature to sea ice concentration, and choices in algorithm design, cause uncertainties in observed Arctic sea ice area, which are far smaller than the observed sea ice loss (e.g., [[#Comiso--2017a|Comiso et al., 2017a]] ; [[#Niederdrenk--2018|Niederdrenk and Notz, 2018]] ; [[#Alekseeva--2019|Alekseeva et al., 2019]] ; [[#Kern--2019|Kern et al., 2019]] ; [[#Meier--2019|Meier and Stewart, 2019]] ). Sea ice area has decreased in every month of the year from 1979 to the present ( ''very high confidence'' ) (Figure 9.13). The absolute and the relative ice losses are highest in late summer-early autumn ( ''high confidence'' ) (Figure 9.13). Averaged over the decade 2010–2019, the monthly Arctic sea ice area from August to October has been around 2 million km² (or about 25%) smaller than during 1979–1988 ( ''high confidence'' ) (Figure 9.13).

<div id="_idContainer035" class="Basic-Text-Frame"></div>

[[File:dcce997cfc6bf21f871ba4f79af9b161 IPCC_AR6_WGI_Figure_9_13.png]]

'''Figure''' '''9.13 |''' '''Arctic sea ice historical records and Coupled Model Intercomparison Project Phase 6 (CMIP6) projections. (Left)''' Absolute anomaly of monthly-mean Arctic sea ice area during the period 1979 to 2019 relative to the average monthly-mean Arctic sea ice area during the period 1979 to 2008. '''(Right)''' Sea ice concentration in the Arctic for March and September, which usually are the months of maximum and minimum sea ice area, respectively. First column: Satellite-retrieved mean sea ice concentration during the decade 1979–1988. Second column: Satellite-retrieved mean sea ice concentration during the decade 2010–2019. Third column: Absolute change in sea ice concentration between these two decades, with grid lines indicating non-significant differences. Fourth column: Number of available CMIP6 models that simulate a mean sea ice concentration above 15 % for the decade 2045–2054. The average observational record of sea ice area is derived from the UHH sea ice area product ( [[#Doerr--2021|Doerr et al., 2021]] ), based on the average sea ice concentration of OSISAF/CCI (OSI-450 for 1979–2015, OSI-430b for 2016–2019) ( [[#Lavergne--2019|Lavergne et al., 2019]] ), NASA Team (version 1, 1979–2019) ( [[#Cavalieri--1996|Cavalieri et al., 1996]] ) and Bootstrap (version 3, 1979–2019) ( [[#Comiso--2017|Comiso, 2017]] ) that is also used for the figure panels showing observed sea ice concentration. Further details on data sources and processing are available in the chapter data table (Table 9.SM.9).

The SROCC discussed the regional distribution of Arctic sea ice loss, and the findings remain valid for the updated time series covering 2019 (Figure 9.13). Sea ice loss in winter is strongest in the Barents Sea, while summer losses occur primarily at the summer sea ice region margins, in particular in the East Siberian, Chukchi, Kara and Beaufort Seas ( [[#Frey--2015|Frey et al., 2015]] ; [[#Chen--2016|Chen et al., 2016]] ; [[#Onarheim--2018|Onarheim et al., 2018]] ; [[#Peng--2018|Peng and Meier, 2018]] ; [[#Maksym--2019|Maksym, 2019]] ). In the Bering Sea, expanding winter sea ice cover was observed until 2017 ( [[#Frey--2015|Frey et al., 2015]] ; [[#Onarheim--2018|Onarheim et al., 2018]] ; [[#Peng--2018|Peng and Meier, 2018]] ), but a marked reduction in sea ice concentration has occurred since then ( ''high confidence'' ) ( [[#Stabeno--2019|Stabeno and Bell, 2019]] ).

With respect to seasonal changes in the sea ice cover, the winter sea ice loss causes a decrease in the average sea ice age and fraction of multi-year ice, as assessed by SROCC ( ''very high confidence'' ), and also of the ocean area covered intermittently by sea ice ( [[#Bliss--2019|Bliss et al., 2019]] ). In contrast, the seasonal ice zone (covered by sea ice in winter but not in summer) has expanded regionally ( [[#Bliss--2019|Bliss et al., 2019]] ) and over the whole Arctic ( [[#Steele--2015|Steele and Ermold, 2015]] ), because the loss of summer sea ice area is larger than the loss of winter sea ice area. Arctic sea ice retreat includes an earlier onset of surface melt in spring and a later freeze up in autumn, lengthening the open water season in the seasonal sea ice zone ( [[#Stroeve--2018|Stroeve and Notz, 2018]] ). However, there is ''low agreement'' in quantification of regional trends of melt and freeze onset between different observational products ( [[#Bliss--2017|Bliss et al., 2017]] ; [[#Smith--2019|Smith and Jahn, 2019]] ).

Reconstructions of Arctic sea ice coverage put the satellite period changes into centennial context. Direct observational data coverage ( [[#Walsh--2017|Walsh et al., 2017]] ) and model reconstructions ( [[#Brennan--2020|Brennan et al., 2020]] ) warrant ''high confidence'' that the low Arctic sea ice area of summer 2012 is unprecedented since 1850, and that the summer sea ice loss is significant in all Arctic regions except for the Central Arctic ( [[#Cai--2021|Cai et al., 2021]] ). Direct winter observational data coverage before 1953 is too sparse to reliably assess Arctic sea ice area. Since 1953, the years 2015 to 2018 had the four lowest values of maximum Arctic sea ice area, which usually occurs in March ( ''high confidence'' ) (Figure 2.20). Reconstructions of Arctic sea ice area before 1850 remain sparse, and as in SROCC, there remains ''medium confidence'' that the current sea ice levels in late summer are unique during the past 1 kyr ( [[IPCC:Wg1:Chapter:Chapter-2#2.3.2.1.1|Section 2.3.2.1.1]] ; [[#Kinnard--2011|Kinnard et al., 2011]] ; [[#De%20Vernal--2013b|De Vernal et al., 2013b]] ).

The observed fluctuations and trends of the Arctic sea ice cover arise from a combination of changes in natural external forcing and anthropogenic forcing, internal variability and internal feedbacks (e.g., [[#Notz--2018|Notz and Stroeve, 2018]] ; [[#Halloran--2020|Halloran et al., 2020]] ). New paleo-proxy techniques indicate regional sea ice changes over epochs and millennia and allow possible drivers to be assessed. Biomarker IP25 ( [[#Belt--2007|Belt et al., 2007]] ) together with other sedimentary biomarkers ( [[#Belt--2018|Belt, 2018]] ) provide local temporal information on seasonal sea ice coverage, permanent sea ice coverage and ice-free waters, with occasional ambiguous contrasting results ( [[#Belt--2019|Belt, 2019]] ). These records and other proposed paleo proxies, including bromine in ice cores ( [[#Spolaor--2016|Spolaor et al., 2016]] ), dinocyst assemblages (e.g., [[#De%20Vernal--2013b|De Vernal et al., 2013b]] ) and driftwood (e.g., [[#Funder--2011|Funder et al., 2011]] ), provide evidence of sea ice fluctuations that exceed internal variability ( ''high confidence'' ).

The inferred sea ice fluctuations over millennia can be related to Northern Hemisphere temperature evolution and give rise to Arctic-wide fluctuations in sea ice coverage in the paleorecord ( [[IPCC:Wg1:Chapter:Chapter-2#2.3.2.1.1|Section 2.3.2.1.1]] ). On a regional scale, fluctuations include decreased sea ice cover during the Allerød warm period (14.7–12.9 ka) in the Laptev ( [[#Hörner--2016|Hörner et al., 2016]] ) and Bering Seas ( [[#Méheust--2018|Méheust et al., 2018]] ); an extensive sea ice cover during the Younger Dryas (around 12 ka) in the Bering ( [[#Méheust--2018|Méheust et al., 2018]] ), Kara ( [[#Hörner--2018|Hörner et al., 2018]] ), Laptev ( [[#Hörner--2016|Hörner et al., 2016]] ) and Barents ( [[#Belt--2015|Belt et al., 2015]] ) Seas, and at the Yermak Plateau ( [[#Kremer--2018|Kremer et al., 2018]] ); little sea ice during the early Holocene, when Northern Hemisphere summer insolation was higher than today (8000 to 9000 years before present), in the North Icelandic Shelf area ( [[#Cabedo-Sanz--2016|Cabedo-Sanz et al., 2016]] ; [[#Xiao--2017|Xiao et al., 2017]] ), Sea of Okhotsk ( [[#Lo--2018|Lo et al., 2018]] ), Canadian Arctic ( [[#Spolaor--2016|Spolaor et al., 2016]] ), Barents ( [[#Berben--2017|Berben et al., 2017]] ), Bering ( [[#Méheust--2018|Méheust et al., 2018]] ), and Chukchi ( [[#Stein--2017|Stein et al., 2017]] ) Seas, at the Yermak Plateau ( [[#Kremer--2018|Kremer et al., 2018]] ) and north of Greenland ( [[#Funder--2011|Funder et al., 2011]] ); increasing sea ice cover throughout much of the middle and late Holocene around Svalbard ( [[#Knies--2017|Knies et al., 2017]] ), in the North Icelandic Shelf area ( [[#Cabedo-Sanz--2016|Cabedo-Sanz et al., 2016]] ; [[#Harning--2019|Harning et al., 2019]] ; [[#Halloran--2020|Halloran et al., 2020]] ), north of Greenland ( [[#Funder--2011|Funder et al., 2011]] ), and in the Western Greenland ( [[#Kolling--2018|Kolling et al., 2018]] ), Barents ( [[#Belt--2015|Belt et al., 2015]] ; [[#Berben--2017|Berben et al., 2017]] ), Chukchi ( [[#De%20Vernal--2013a|De Vernal et al., 2013a]] ; [[#Stein--2017|Stein et al., 2017]] ) and Laptev ( [[#Hörner--2016|Hörner et al., 2016]] ) Seas. The consistent, Arctic-wide changes give ''high confidence'' in millennial-scale co-variability of the sea ice cover with temperature fluctuation.

The SROCC assessed that approximately half of the satellite-observed Arctic summer sea ice loss is driven by increased concentrations of atmospheric greenhouse gases ( ''medium confidence'' ). Recent attribution studies now allow the strengthened assessment that it is ''very likely'' that more than half of the observed Arctic sea ice loss in summer is anthropogenic ( [[IPCC:Wg1:Chapter:Chapter-3#3.4.1.1|Section 3.4.1.1]] ). This assessment is confirmed by process-based analyses of Arctic sea ice loss not assessed by SROCC. Similar to the paleorecord, the satellite record of Arctic sea ice area from 1979 onwards is strongly and linearly correlated with global mean temperature on decadal and longer time scales (Figures 9.14a,e) (e.g., [[#Gregory--2002|Gregory et al., 2002]] ; [[#Rosenblum--2017|Rosenblum and Eisenman, 2017]] ). The correlation holds across all months with R <sup>2</sup> ranging from 0.61 to 0.81 ( [[#Niederdrenk--2018|Niederdrenk and Notz, 2018]] ). However, in contrast to paleorecords, sea ice fluctuations during the satellite period are only weakly correlated with Northern Hemisphere insolation ( [[#Notz--2012|Notz and Marotzke, 2012]] ); modern Northern Hemisphere sea ice area is more strongly correlated with atmospheric carbon dioxide (CO <sub>2</sub> ) concentration ( [[#Johannessen--2008|Johannessen, 2008]] ; [[#Notz--2012|Notz and Marotzke, 2012]] ) and cumulative anthropogenic CO <sub>2</sub> emissions (Figures 9.14b,f; [[#Zickfeld--2012|Zickfeld et al., 2012]] ; [[#Herrington--2014|Herrington and Zickfeld, 2014]] ; [[#Notz--2016|Notz and Stroeve, 2016]] ). The R <sup>2</sup> values of the correlation between sea ice area and cumulative CO <sub>2</sub> emissions range across all months from 0.76 to 0.92 ( [[#Stroeve--2018|Stroeve and Notz, 2018]] ). In summary, there is ''high confidence'' that satellite-observed Arctic sea ice area is strongly correlated with global mean temperature, CO <sub>2</sub> concentration and cumulative anthropogenic CO <sub>2</sub> emissions.

In addition to changes in the external forcing, internal variability substantially affects Arctic sea ice, evidenced from both paleorecords (e.g., [[#Chan--2017|Chan et al., 2017]] ; [[#Hörner--2017|Hörner et al., 2017]] ; [[#Kolling--2018|Kolling et al., 2018]] ) and satellites after 1979 (e.g., [[#Notz--2018|Notz and Stroeve, 2018]] ; [[#Roberts--2020|Roberts et al., 2020]] ) ( ''high confidence'' ). Most of the internal variability on annual time scales is related to atmospheric temperature fluctuations, for example linked to cyclone activities ( [[#Wernli--2018|Wernli and Papritz, 2018]] ; [[#Olonscheck--2019|Olonscheck et al., 2019]] ), while multi-decadal internal variability is primarily related to changes in oceanic heat transport ( [[#Zhang--2015|Zhang, 2015]] ; [[#Halloran--2020|Halloran et al., 2020]] ). These mechanisms are represented in current climate models ( [[#Olonscheck--2019|Olonscheck et al., 2019]] ; [[#Halloran--2020|Halloran et al., 2020]] ), but the resulting internal variability of September sea ice area in CMIP5 and CMIP6 models, as given by the ensemble mean standard deviation Σ <sub>SIA,Sep</sub> = 0.5 million km² ( [[#Olonscheck--2017|Olonscheck and Notz, 2017]] ; Notz and SIMIP Community, 2020), exceeds the estimated internal variability for the period 1850 to 1979 from both reanalyses ( Σ <sub>SIA,Sep</sub> = 0.3 million km <sup>2</sup> ) and direct observational reconstructions ( Σ <sub>SIA,Sep</sub> = 0.2 million km <sup>2</sup> ) ( ''medium confidence'' because of limited reliability of longer-term sea ice reconstructions) ( [[#Brennan--2020|Brennan et al., 2020]] ). Internal variability has been estimated to have contributed 30 to 50% of the observed Arctic summer sea ice loss since 1979 ( [[#Kay--2011|Kay et al., 2011]] ; [[#Stroeve--2012|Stroeve et al., 2012]] ; [[#Ding--2017|Ding et al., 2017]] , 2019; [[#England--2019|England et al., 2019]] ). However, this estimate from models might be biased towards internal over forced variability because of the models’ high internal variability and because the CMIP5 simulated September sea ice sensitivity to forcing is lower than observed, even if internal variability is taken into account ( [[#Notz--2016|Notz and Stroeve, 2016]] ; [[#Rosenblum--2017|Rosenblum and Eisenman, 2017]] ). Most CMIP6 models fail to simulate the observed sensitivity of sea ice loss to CO <sub>2</sub> emissions (as a proxy for time) and to temperature simultaneously. However, they better capture the observed sensitivity of sea ice loss to CO <sub>2</sub> emissions than CMIP5 models ( [[IPCC:Wg1:Chapter:Chapter-3#3.4.1|Section 3.4.1]] ; Figure 9.14h; Notz and SIMIP Community, 2020).

<div id="_idContainer037" class="Basic-Text-Frame"></div>

[[File:45de42f507af8cfcb07c7937366bf85b IPCC_AR6_WGI_Figure_9_14.png]]

'''Figure 9.14''' '''|''' '''Monthly mean March (a–d) and September (e–h) sea ice area as a function of global surface air temperature (GSAT) anomaly (a, e); cumulative anthropogenic CO''' <sub>2</sub> '''emissions (b, f); year (c, g) in Coupled Model Intercomparison Project Phase 6 (CMIP6) model simulations (shading, ensemble mean as bold line) and in observations (black dots).''' Panels (d) and (h) show the sensitivity of sea ice loss to anthropogenic CO <sub>2</sub> emissions as a function of the modelled sensitivity of GSAT to anthropogenic CO <sub>2</sub> emissions. In panels (d) and (h), the black dot denotes the observed sensitivity, while the shading around it denotes internal variability as inferred from CMIP6 simulations (after Notz and SIMIP Community, 2020). Further details on data sources and processing are available in the chapter data table (Table 9.SM.9).

The SROCC examined the different atmospheric and oceanic processes that caused the observed sea ice loss, with recent studies providing new evidence for the importance of variations in air temperature ( [[#Olonscheck--2019|Olonscheck et al., 2019]] ; [[#Dahlke--2020|Dahlke et al., 2020]] ), wind patterns ( [[#Graham--2019|Graham et al., 2019]] ), oceanic heat flux ( [[#Docquier--2021|Docquier et al., 2021]] ) and riverine heat influx ( [[#Park--2020|Park et al., 2020]] ). As in SROCC, the relative contribution of each physical cause to the sea ice loss cannot be robustly quantified because of disagreement among models ( [[#Burgard--2017|Burgard and Notz, 2017]] ), sparse observations, and limited understanding of the variation of each factor with global mean temperature. This is addressed by new diagnostics available from CMIP6 simulations, which now allow for more detailed analyses of the drivers of sea ice loss at a process level ( [[#Keen--2021|Keen et al., 2021]] ).

In examining temperature thresholds for the loss of Arctic summer sea ice, the Special Report on Global Warming of 1.5°C (SR1.5; [[#Hoegh-Guldberg--2018|Hoegh-Guldberg et al., 2018]] ) and SROCC assess that a reduction of September mean sea ice area to below 1 million km <sup>2</sup> , practically a sea ice-free Arctic Ocean, is more probable for a global mean warming of 2°C compared to global mean warming of 1.5°C ( ''high confidence'' ). Analyses of CMIP6 simulations (Notz and SIMIP Community, 2020) confirm this result, as they show that, on decadal and longer time scales, Arctic summer sea ice area will remain highly correlated with global mean temperature until the summer sea ice has vanished (Figure 9.14a,e). Quantitatively, existing studies ( [[#Screen--2017|Screen and Williamson, 2017]] ; [[#Jahn--2018|Jahn, 2018]] ; [[#Ridley--2018|Ridley and Blockley, 2018]] ; [[#Sigmond--2018|Sigmond et al., 2018]] ; Notz and SIMIP Community, 2020) also show that, for a warming between 1.5 and 2°C, the Arctic will only be practically sea ice free in September in some years, while at 3°C warming, the Arctic is practically sea ice free in September in most years, with longer practically sea ice-free periods at higher warming levels ( ''medium confidence'' ). However, because of the CMIP5 and CMIP6 models’ generally too low sensitivity of sea ice loss to global warming, there is only ''low confidence'' regarding the specific warming level at which the Arctic Ocean first becomes practically sea ice free ( [[IPCC:Wg1:Chapter:Chapter-4#4.3.2.1|Section 4.3.2.1]] ; Notz and SIMIP Community, 2020).

In contrast, CMIP6 models capture the observed sensitivity of Arctic sea ice area to cumulative anthropogenic CO <sub>2</sub> emissions well, providing ''high confidence'' that the Arctic Ocean will ''likely'' become practically sea ice free in the September mean for the first time for future CO <sub>2</sub> emissions of less than 1000 Gt and before the year 2050 in all SSP scenarios (Notz and SIMIP Community, 2020). This new assessment is consistent with an observation-based projection of a practically sea ice-free Arctic Ocean in September for additional anthropogenic CO <sub>2</sub> emissions of 800 ± 330 GtCO <sub>2</sub> beyond the year 2018 ( [[#Notz--2018|Notz and Stroeve, 2018]] ; [[#Stroeve--2018|Stroeve and Notz, 2018]] ). This estimate may, however, be too high due to neglecting possible future reduction in atmospheric aerosol load that would cause additional warming ( [[#Gagné--2015a|Gagné et al., 2015a]] ; [[#Wang--2018|Wang et al., 2018]] ), and is subject to the same constraints as the carbon budget analysis for global mean temperature (see section 5.5 for details). Based on CMIP6 simulations, it is ''very likely'' that the Arctic Ocean will remain sea ice covered in winter in all scenarios throughout this century (Sections 4.3.2 and 4.4.2).

There is an indication that CMIP6 simulations of Arctic sea ice have improved relative to CMIP5 ( [[IPCC:Wg1:Chapter:Chapter-3#3.4.1.1|Section 3.4.1.1]] ), but detailed evaluation studies exist mainly for CMIP5 models. These studies found that CMIP5 model projections and reanalyses show a large spread of simulated regional Arctic sea ice concentration ( [[#Laliberté--2016|Laliberté et al., 2016]] ; [[#Chevallier--2017|Chevallier et al., 2017]] ), which remains true for CMIP6 models ( [[#Shu--2020|Shu et al., 2020]] ; [[#Wei--2020|Wei et al., 2020]] ). In addition, both CMIP5 and CMIP6 models show a large spread in the simulated seasonal cycle of Arctic sea ice area, with too high a sea ice area in March in the ensemble mean (Notz and SIMIP Community, 2020). The CMIP5 models have also had difficulty simulating realistic landfast sea ice ( [[#Laliberté--2018|Laliberté et al., 2018]] ). These findings imply that both CMIP5 and CMIP6 models do not realistically capture the regional and seasonal processes governing observed Arctic sea ice evolution, causing ''low confidence'' in the models’ projections of future regional sea ice evolution, including updated projections for shipping routes across the Northern Sea Route and Northwest Passage ( [[#Wei--2020|Wei et al., 2020]] ).

The CMIP5 models also have issues with capturing the seasonal cycle of observed changes in Arctic sea ice drift speed, which affects their simulation of regional sea ice concentration patterns. Direct measurements of Arctic sea ice from drift buoys and satellites show that drift speed of Arctic sea ice has increased over the satellite period in all seasons (e.g., [[#Rampal--2009|Rampal et al., 2009]] ; [[#Docquier--2017|Docquier et al., 2017]] ). In summer, CMIP5 models show a slowdown of Arctic sea ice drift rather than the observed acceleration ( [[#Tandon--2018|Tandon et al., 2018]] ). In winter, CMIP5 models generally capture the observed acceleration of Arctic drift speed. The drift acceleration is primarily caused by the decrease in concentration and thickness in the observational record ( [[#Rampal--2009|Rampal et al., 2009]] ; [[#Spreen--2011|Spreen et al., 2011]] ; Olason and [[#Notz--2014|Notz, 2014]] ; [[#Docquier--2017|Docquier et al., 2017]] ) and, for winter, in CMIP5 models ( [[#Tandon--2018|Tandon et al., 2018]] ). Changes in wind speed are less important for the observed large-scale changes ( [[#Spreen--2011|Spreen et al., 2011]] ; [[#Vihma--2012|Vihma et al., 2012]] ; Olason and [[#Notz--2014|Notz, 2014]] ; [[#Docquier--2017|Docquier et al., 2017]] ; [[#Tandon--2018|Tandon et al., 2018]] ). In summary, there is ''high confidence'' that Arctic sea ice drift has accelerated because of the decrease in sea ice concentration and thickness.

The SR1.5 assessed with ''high confidence'' that there is no hysteresis in the loss of Arctic summer sea ice. In addition, there is no tipping point or critical threshold in global mean temperature beyond which the loss of summer sea ice becomes self-accelerating and irreversible ( ''high confidence'' ). This is because stabilizing feedbacks during winter related to increased heat loss through thin ice and thin snow, and increased emission of longwave radiation from open water, dominate over the amplifying ice albedo feedback (see [[IPCC:Wg1:Chapter:Chapter-7#7.4.2|Section 7.4.2]] for details on the individual feedbacks; e.g., [[#Eisenman--2012|Eisenman, 2012]] ; [[#Wagner--2015|Wagner and Eisenman, 2015]] ; [[#Notz--2018|Notz and Stroeve, 2018]] ). Observed and modelled Arctic summer sea ice and global mean temperature are linked with little temporal delay, and the summer sea ice loss is reversible on decadal time scales ( [[#Armour--2011|Armour et al., 2011]] ; [[#Ridley--2012|Ridley et al., 2012]] ; [[#Li--2013|Li et al., 2013]] ; [[#Jahn--2018|Jahn, 2018]] ). The loss of winter sea ice is reversible as well, but the loss of winter sea ice area per degree of warming in CMIP5 and CMIP6 projections increases as the ice retreats from the continental shore lines, because these limit the possible areal fluctuations ( ''high confidence'' ) ( [[IPCC:Wg1:Chapter:Chapter-4#4.3.2.1|Section 4.3.2.1]] ; [[#Bathiany--2016|Bathiany et al., 2016]] , 2020; [[#Meccia--2020|Meccia et al., 2020]] ).

<div id="9.3.1.2" class="h3-container"></div>

<span id="arctic-sea-ice-volume-and-thickness"></span>
==== 9.3.1.2 Arctic Sea Ice Volume and Thickness ====

<div id="h3-16-siblings" class="h3-siblings"></div>

The SROCC assessed with ''very high confidence'' that Arctic sea ice has become thinner over the satellite period from 1979 onwards, and this assessment is confirmed for the updated time series ( [[IPCC:Wg1:Chapter:Chapter-2#2.3.2.1.1|Section 2.3.2.1.1]] ). Sea ice area has also decreased substantially over this period ( [[#9.3.1.1|Section 9.3.1.1]] ), leading to the assessment that Arctic sea ice volume has also decreased with ''very high confidence'' over the satellite period since 1979. There is, however, only ''low confidence'' in quantitative estimates of the sea ice volume loss over this period because of a lack of reliable, long-term, pan-Arctic observations and substantial spread in available reanalyses ( [[#Chevallier--2017|Chevallier et al., 2017]] ). Current best estimates from reanalyses suggest a reduction of September Arctic sea ice volume of 55 to 65% over the period 1979–2010, and of about 72% over the period 1979–2016, with the latter deemed a conservative estimate ( [[#Schweiger--2019|Schweiger et al., 2019]] ).

For the more recent past, ice thickness can be directly estimated from satellite retrievals of sea ice freeboard ( [[#Kwok--2015|Kwok and Cunningham, 2015]] ; [[#Kwok--2018|Kwok, 2018]] ). Based on these retrievals, there is ''medium confidence'' that Arctic sea ice volume has decreased since 2003. There is ''low confidence'' in the amount of decrease over this period and over the CryoSat-2 period from 2011 onwards, primarily because of snow-induced uncertainties in the retrieval algorithms, the shortness of the record, and the small identified trend (e.g., [[#Bunzel--2018|Bunzel et al., 2018]] ; [[#Petty--2018|Petty et al., 2018]] , 2020).

Observations of regional changes in sea ice thickness vary in quality. Analysis of submarine data in the central Arctic Ocean suggests that its sea ice has thinned by about 75 cm compared to the mid-1970s ( [[IPCC:Wg1:Chapter:Chapter-2#2.3.2.1.1|Section 2.3.2.1.1]] ). For smaller regions, data are too sparse to allow for quantitative estimates of long-term trends ( [[#King--2017|King et al., 2017]] ; [[#Rösel--2018|Rösel et al., 2018]] ), but a clear thinning signal over 10 to 20 years has been found for sea ice in the Fram Strait ( [[#Spreen--2020|Spreen et al., 2020]] ), north of Canada ( [[#Haas--2017|Haas et al., 2017]] ) and for landfast ice in the Kongsfjorden/Svalbard Arctic border ( [[#Pavlova--2019|Pavlova et al., 2019]] ). The CMIP5 models and reanalyses fail to capture the observed distribution ( [[#Stroeve--2014|Stroeve et al., 2014]] ; [[#Shu--2015|Shu et al., 2015]] ) and evolution ( [[#Chevallier--2017|Chevallier et al., 2017]] ) of Arctic sea ice thickness. Most CMIP6 models do not capture the observed spatial distribution of sea ice thickness realistically ( [[#Wei--2020|Wei et al., 2020]] ). This leads to ''low confidence'' in estimates of thickness from reanalyses and from CMIP5 and CMIP6 models and in their projections of sea ice volume.

<div id="9.3.2" class="h2-container"></div>

<span id="antarctic-sea-ice"></span>
=== 9.3.2 Antarctic Sea Ice ===

<div id="h2-15-siblings" class="h2-siblings"></div>

<div id="9.3.2.1" class="h3-container"></div>

<span id="antarctic-sea-ice-coverage"></span>
==== 9.3.2.1 Antarctic Sea Ice Coverage ====

<div id="h3-17-siblings" class="h3-siblings"></div>

The SROCC ( [[#Meredith--2019|Meredith et al., 2019]] ) assessed that there was no significant trend in annual mean Antarctic sea ice area over the period of reliable satellite retrievals starting in 1979 ( ''high confidence'' ). The updated time series is consistent with this assessment. It includes a maximum sea ice area in 2014, then a substantial decline until the minimum sea ice area in 2017, and an increase in sea ice area since 2017 (Figures 2.20 and 9.15; [[#Schlosser--2018|Schlosser et al., 2018]] ; [[#Maksym--2019|Maksym, 2019]] ; [[#Parkinson--2019|Parkinson, 2019]] ). As assessed in [[IPCC:Wg1:Chapter:Chapter-2#2.3.2.1.2|Section 2.3.2.1.2]] , the possible significance of the increase in mean Antarctic sea ice area over the shorter period 1979 to 2014 (Figure 2.20; [[#Simmonds--2015|Simmonds, 2015]] ; [[#Comiso--2017b|Comiso et al., 2017b]] ) is unclear. This is because of observational uncertainty (see [[#9.3.1.1|Section 9.3.1.1]] ), large year-to-year fluctuations in all months (Figure 9.15), and limited understanding of the processes and reliability of year-to-year correlation of Antarctic sea ice area ( [[#Yuan--2017|Yuan et al., 2017]] ).

<div id="_idContainer039" class="Basic-Text-Frame"></div>

[[File:acf3192464d6175b170565c4ccacb36f IPCC_AR6_WGI_Figure_9_15.png]]

'''Figure''' '''9.15 |''' '''Antarctic sea ice historical records and Coupled Model Intercomparison Project Phase 6 (CMIP6) projections.''' '''(Left)''' Absolute anomaly of observed monthly mean Antarctic sea ice area during the period 1979–2019 relative to the average monthly mean Antarctic sea ice area during the period 1979–2008. '''(Right)''' Sea ice coverage in the Antarctic as given by the average of the three most widely used satellite-based estimates for September and February, which usually are the months of maximum and minimum sea ice coverage, respectively. First column: Mean sea ice coverage during the decade 1979–1988. Second column: Mean sea ice coverage during the decade 2010–2019. Third column: Absolute change in sea ice concentration between these two decades, with grid lines indicating non-significant differences. Fourth column: Number of available CMIP6 models that simulate a mean sea ice concentration above 15% for the decade 2045–2054. The average observational record of sea ice area is derived from the UHH sea ice area product ( [[#Doerr--2021|Doerr et al., 2021]] ), based on the average sea ice concentration of OSISAF/CCI (OSI-450 for 1979–2015, OSI-430b for 2016–2019) ( [[#Lavergne--2019|Lavergne et al., 2019]] ), NASA Team (version 1, 1979–2019) ( [[#Cavalieri--1996|Cavalieri et al., 1996]] ) and Bootstrap (version 3, 1979–2019) ( [[#Comiso--2017|Comiso, 2017]] ) that is also used for the figure panels showing observed sea ice concentration. Further details on data sources and processing are available in the chapter data table (Table 9.SM.9).

As assessed by SROCC, the evolution of mean Antarctic sea ice area is the result of opposing regional trends ( ''high confidence'' ), with slightly decreasing sea ice cover during the period 1979 to 2019 in the Amundsen and Bellingshausen Seas, particularly during summer, and slightly increasing sea ice cover in the eastern parts of the Weddell and Ross Seas (Figure 9.15). With the exception of the Ross Sea, these trends are not significant, considering the large variability of the time series ( [[#Yuan--2017|Yuan et al., 2017]] ).

The SROCC assessed that the regional trends are closely related to meridional wind trends ( ''high confidence'' ). This is the case as the regional trends in the maximum northward extent of the ice cover (Figure 9.15) are determined by the balance between the northward advection of the ice that is formed in polynyas near the continental margin, and the lateral and subsurface melting through oceanic heat fluxes. The advection of the sea ice is strongly correlated with winds and cyclones ( [[#Schemm--2018|Schemm, 2018]] ; [[#Vichi--2019|Vichi et al., 2019]] ; [[#Alberello--2020|Alberello et al., 2020]] ). Accordingly, the increasing sea ice area in the Ross Sea can be linked to a strengthening of the Amundsen Sea low (e.g., [[#Holland--2017b|Holland et al., 2017b]] , 2018), while other regional sea ice trends in the austral autumn can be linked to changes in westerly winds, cyclone activity and the Southern Annular Mode (SAM) in summer and spring ( [[#Doddridge--2017|Doddridge and Marshall, 2017]] ; [[#Holland--2017a|Holland et al., 2017a]] ; [[#Schemm--2018|Schemm, 2018]] ). In addition to the wind-driven changes, increased near-surface ocean stratification ( [[#9.2.1.3|Section 9.2.1.3]] ) has contributed to the observed increase in sea ice coverage (e.g., [[#Purich--2018|Purich et al., 2018]] ; L. [[#Zhang--2019|]] [[#Zhang--2019|]] [[#Zhang--2019|Zhang et al., 2019]] ) as it tends to cool the surface ocean (Sections 9.2.1.1 and 9.2.3.2). The changes in stratification result partly from surface freshening ( [[#De%20Lavergne--2014|De Lavergne et al., 2014]] ), associated with increased northward sea ice advection ( [[#Haumann--2020|Haumann et al., 2020]] ) and/or melting of the Antarctic ice sheet ( ''medium confidence'' ) (e.g., [[#Haumann--2020|Haumann et al., 2020]] ; [[#Jeong--2020|Jeong et al., 2020]] ; [[#Mackie--2020|Mackie et al., 2020]] ), and amplified by local ice–ocean feedbacks ( [[#Goosse--2014|Goosse and Zunz, 2014]] ; [[#Lecomte--2017|Lecomte et al., 2017]] ; [[#Goosse--2018|Goosse et al., 2018]] ). In the Amundsen Sea, strong ice shelf melting can cause local sea ice melt next to the ice shelf front by entraining warm circumpolar deep water to the ice shelf cavity and surface ocean ( ''medium confidence'' ) (Sections 9.2.3.2 and 9.4.2.2; [[#Jourdain--2017|Jourdain et al., 2017]] ; [[#Merino--2018|Merino et al., 2018]] ). It has also been suggested that the observed regional increase in sea ice coverage since 1979 results from a long-term Southern Ocean surface cooling trend (e.g., [[#Kusahara--2019|Kusahara et al., 2019]] ; [[#Jeong--2020|Jeong et al., 2020]] ) but the importance of this mechanism for the observed sea ice evolution is unclear owing to intricate feedbacks between sea ice change and surface cooling ( [[#Haumann--2020|Haumann et al., 2020]] ). The importance of changing wave activity ( [[#9.6.4.2|Section 9.6.4.2]] ; [[#Kohout--2014|Kohout et al., 2014]] ; [[#Bennetts--2017|Bennetts et al., 2017]] ; [[#Roach--2018b|Roach et al., 2018b]] ) on sea ice is unclear due to limited process understanding. In summary, there is ''high confidence'' that regional Antarctic trends are primarily caused by changes in sea ice drift and decay, with ''medium confidence'' in a dominating role of changing wind pattern. The precise relative contribution of individual drivers remains uncertain because of limited observations, disagreement between models, unresolved processes, and temporal and spatial remote linkages caused by sea ice drift ( [[#9.2.3.2|Section 9.2.3.2]] ; [[#Pope--2017|Pope et al., 2017]] ).

Recent research has confirmed SROCC assessment of atmospheric and oceanic drivers of the sea ice decline from 2014 to 2017, which can be linked to changes in both subsurface ocean heat flux ( [[#Meehl--2019|Meehl et al., 2019]] ; [[#Purich--2019|Purich and England, 2019]] ) and atmospheric circulation, with the latter partly related to teleconnections with the tropics ( [[#Meehl--2019|Meehl et al., 2019]] ; [[#Purich--2019|Purich and England, 2019]] ; G. [[#Wang--2019|]] [[#Wang--2019|]] [[#Wang--2019|Wang et al., 2019]] ). In the Weddell Sea, these changes caused in 2017 the re-emergence of the largest polynya over the Maud Rise since the 1970s ( [[#9.2.3.2|Section 9.2.3.2]] ; [[#Campbell--2019|Campbell et al., 2019]] ; [[#Jena--2019|Jena et al., 2019]] ; [[#Turner--2020|Turner et al., 2020]] ).

The AR5 ( [[#Collins--2013|Collins et al., 2013]] ) and SROCC found ''low confidence'' in future projections of Antarctic sea ice. This includes the projected mitigation of the sea ice loss by stratospheric ozone recovery ( [[#Smith--2012|Smith et al., 2012]] ) and by an increased freshwater input from melting of the Antarctic Ice Sheet ( [[#Bronselaer--2018|Bronselaer et al., 2018]] ). Compared to the interannual variability during the satellite record from 1979 onwards, models simulate too much variability in both CMIP5 ( [[#Zunz--2013|Zunz et al., 2013]] ) and CMIP6 ( [[#Roach--2020|Roach et al., 2020]] ). The seasonal cycle in sea ice coverage is misrepresented in most CMIP5 (e.g., [[#Holmes--2019|Holmes et al., 2019]] ) and CMIP6 models ( [[#Roach--2020|Roach et al., 2020]] ), but the multi-model mean seasonal cycle in CMIP5 and CMIP6 agrees well with observations ( [[#Shu--2015|Shu et al., 2015]] ; [[#Roach--2020|Roach et al., 2020]] ). Most CMIP5 models do not realistically simulate the evolution of Antarctic sea ice volume ( [[#Shu--2015|Shu et al., 2015]] ) and consistently overestimate the amount of low concentration sea ice, and underestimate the amount of high concentration sea ice ( [[#Roach--2018a|Roach et al., 2018a]] ). In contrast, CMIP6 models simulate a more realistic distribution of regional sea ice coverage ( [[#Roach--2020|Roach et al., 2020]] ). Most CMIP5 models poorly represent Antarctic sea ice drift (e.g., [[#Schroeter--2018|Schroeter et al., 2018]] ; [[#Holmes--2019|Holmes et al., 2019]] ), affecting simulated historical trends, with models that simulate a strong sea ice motion showing more variability in sea ice coverage than models with weaker sea ice motion ( [[#Schroeter--2018|Schroeter et al., 2018]] ). Owing to ''limited agreement'' between model simulations and observations, limited reliable observations on a process level, and a lack of process understanding of the substantial spread in CMIP5 and CMIP6 model simulations, there remains ''low confidence'' in existing future projections of Antarctic sea ice decrease and lack of decrease.

The discrepancy between the modelled and observed evolution of Antarctic sea ice has been related by SROCC to deficiencies in modelled stratification, freshening by ice-shelf meltwater, clouds, and other wind- and ocean-driven processes. Recent studies highlight the possible mis-representation of freshwater fluxes from ice shelves ( [[#Jeong--2020|Jeong et al., 2020]] ), and the possible effect of the low resolution of most models ( [[#Sidorenko--2019|Sidorenko et al., 2019]] ), even though lower-resolution models are, in principle, capable of a realistic simulation of the seasonal sea ice budgets in the Southern Ocean ( [[#Holmes--2019|Holmes et al., 2019]] ). The relative importance of these possible reasons for the models’ shortcomings remains unclear (see [[IPCC:Wg1:Chapter:Chapter-3#3.4.1.2|Section 3.4.1.2]] for details).

The analysis and understanding of the long-term evolution of the Antarctic sea ice cover is hindered by the scarcity of observational records before the satellite period, and the scarcity of paleorecords (see [[IPCC:Wg1:Chapter:Chapter-2#2.3.2.1.2|Section 2.3.2.1.2]] for further details). Such long records are particularly relevant given that the Southern Ocean response to external forcing takes longer than the length of the available direct observational record ( [[#Goosse--2001|Goosse and Renssen, 2001]] ; [[#Armour--2016|Armour et al., 2016]] ). There is only ''limited evidence'' for large-scale decadal fluctuations in sea ice coverage caused by large-scale temperature and wind forcing. Sparse direct pre-satellite observations suggest a decrease in sea ice coverage from the 1950s to the 1970s ( [[#Fan--2014|Fan et al., 2014]] ). Paleo-proxy data indicate that, on multi-decadal to multi-centennial time scales, sea ice coverage of the Southern Ocean follows large-scale temperature trends (e.g., [[#Crosta--2018|Crosta et al., 2018]] ; [[#Chadwick--2020|Chadwick et al., 2020]] ; [[#Lamping--2020|Lamping et al., 2020]] ), for example linked to fluctuations in the El Niño–Southern Oscillation and Southern Annular Mode ( [[#Crosta--2021|Crosta et al., 2021]] ), and that during the Last Glacial Maximum, Antarctic sea ice extended to about the polar front latitude in most regions during winter, whereas the extent during summer is less well understood (e.g., [[#Benz--2016|Benz et al., 2016]] ; [[#Xiao--2016|Xiao et al., 2016]] ; [[#Nair--2019|Nair et al., 2019]] ).

Regionally, proxy data from ice cores consistently indicate that the increase of sea ice area in the Ross Sea and the decrease of sea ice area in the Bellingshausen Sea are part of longer centennial trends and exceed internal variability on multi-decadal time scales ( ''medium confidence'' ) (e.g., [[#Thomas--2019|Thomas et al., 2019]] ; [[#Tesi--2020|Tesi et al., 2020]] ). These centennial trends are consistent with simulations from CMIP5 models ( [[#Hobbs--2016b|Hobbs et al., 2016b]] ; J.M. [[#Jones--2016|]] [[#Jones--2016|Jones et al., 2016]] ; [[#Kimura--2017|Kimura et al., 2017]] ).

There is ''low confidence'' in the attribution of the observed changes in Antarctic sea ice area ( [[IPCC:Wg1:Chapter:Chapter-3#3.4.1.2|Section 3.4.1.2]] ). Based on the available evidence, the lack of a negative trend of Antarctic sea ice area, despite substantial global warming in recent decades, has been attributed to internal variability in analyses of the observational record ( [[#Meier--2013|Meier et al., 2013]] ; [[#Gallaher--2014|Gallaher et al., 2014]] ; [[#Gagné--2015b|Gagné et al., 2015b]] ), reconstructions from early observations ( [[#Fan--2014|Fan et al., 2014]] ; [[#Edinburgh--2016|Edinburgh and Day, 2016]] ) and proxy data ( [[#Hobbs--2016b|Hobbs et al., 2016b]] ) in model simulations ( [[#Turner--2013|Turner et al., 2013]] ; [[#Zunz--2013|Zunz et al., 2013]] ; L. [[#Zhang--2019|]] [[#Zhang--2019|]] [[#Zhang--2019|Zhang et al., 2019]] ). Nonetheless, without accurate simulations of observed changes, the possible contribution of anthropogenic forcing to the regional changes in sea ice area remains unclear ( [[#Hosking--2013|Hosking et al., 2013]] ; [[#Turner--2013|Turner et al., 2013]] ; [[#Haumann--2014|Haumann et al., 2014]] ; L. [[#Zhang--2019|]] [[#Zhang--2019|]] [[#Zhang--2019|Zhang et al., 2019]] ).

The attribution of the observed trends in atmospheric and oceanic forcing is also uncertain because of limited observational records and discrepancies between modelled and observed evolution of the sea ice cover. More specifically, there is contrasting evidence for a direct role of stratospheric ozone depletion on the observed changes in atmospheric circulation ( [[#Haumann--2014|Haumann et al., 2014]] ; [[#England--2016|England et al., 2016]] ; [[#Landrum--2017|Landrum et al., 2017]] ). In contrast, there is ''high confidence'' that multi-decadal variations in the tropical Pacific and in the Atlantic affect the Amundsen Sea low ( [[#Li--2014|Li et al., 2014]] ; [[#Kwok--2016|Kwok et al., 2016]] ; [[#Meehl--2016|Meehl et al., 2016]] ; [[#Purich--2016|Purich et al., 2016]] ; [[#Simpkins--2016|Simpkins et al., 2016]] ), while other modes of climate variability (Annex IV) affect, for example, Southern Ocean cyclone activity ( [[#Simpkins--2012|Simpkins et al., 2012]] ; [[#Cerrone--2017|Cerrone et al., 2017]] ; [[#Schemm--2018|Schemm, 2018]] ) ''.''

<div id="9.3.2.2" class="h3-container"></div>

<span id="antarctic-sea-ice-thickness"></span>
==== 9.3.2.2 Antarctic Sea Ice Thickness ====

<div id="h3-18-siblings" class="h3-siblings"></div>

The SROCC assessed that observations are too sparse to reliably estimate long-term trends in Antarctic sea ice thickness. This remains true, and only qualitative statements on prevailing thicknesses are possible. Data from ICESat-1 laser altimetry ( [[#Kurtz--2012|Kurtz and Markus, 2012]] ), from Operation IceBridge ( [[#Kwok--2018|Kwok and Kacimi, 2018]] ), and long-term shipboard observations collected in the Antarctic Sea Ice Processes and Climate (ASPeCt) dataset ( [[#Worby--2008|Worby et al., 2008]] ) suggest that sea ice thicker than 1 m prevails in regions of multi-year ice along the eastern coast of the Antarctic Peninsula in the Weddell Sea, in the high-latitude embayment of the Weddell Sea, and along the coast of the Amundsen Sea, with remaining regions dominated by thinner first-year sea ice ( ''high confidence'' ). Regional patterns in ice thickness are affected by areas of high snow deposition and resulting snow-ice formation ( [[#Massom--2001|Massom et al., 2001]] ; [[#Maksym--2008|Maksym and Markus, 2008]] ), and deformation, ridging, and rafting that regionally cause formation of very thick sea ice ( [[#Massom--2006|Massom et al., 2006]] ; G. [[#Williams--2015|]] [[#Williams--2015|Williams et al., 2015]] ). In addition, near ice shelves a sub-ice platelet layer from supercooled water can significantly increase sea ice thickness ( [[#Hoppmann--2020|Hoppmann et al., 2020]] ; [[#Haas--2021|Haas et al., 2021]] ). Regarding snow thickness, observations are too sparse in space and time to reliably estimate changes across Southern Ocean sea ice ( [[#Webster--2018|Webster et al., 2018]] ).

There is ''low confidence'' in the long-term trend of Antarctic sea ice thickness. Both ASPeCt and ICESat-1 measurements are biased low in regions with thick ice ( [[#Kern--2015|Kern and Spreen, 2015]] ), compared to results from reanalyses ( [[#Massonnet--2013|Massonnet et al., 2013]] ; [[#Haumann--2016|Haumann et al., 2016]] ) and observations with autonomous vehicles under sea ice (G. [[#Williams--2015|]] [[#Williams--2015|Williams et al., 2015]] ). Estimates of sea ice thickness from CryoSat-2 do not substantially reduce uncertainty, primarily because of the unknown snow thickness and radar scattering above the snow–ice interface ( [[#Bunzel--2018|Bunzel et al., 2018]] ; [[#Kwok--2018|Kwok and Kacimi, 2018]] ; [[#Kacimi--2020|Kacimi and Kwok, 2020]] ). Isolated in situ time series show no clear long-term trend in landfast ice thickness in the Weddell Sea ( [[#Arndt--2020|Arndt et al., 2020]] ). Reanalyses suggest overall increasing sea ice thickness and volume between 1980 and 2010 ( [[#Holland--2014|Holland et al., 2014]] ; [[#Zhang--2014|Zhang, 2014]] ; [[#Massonnet--2015|Massonnet et al., 2015]] ), while CMIP5 ( [[#Shu--2015|Shu et al., 2015]] ; [[#Schroeter--2018|Schroeter et al., 2018]] ) and CMIP6 models simulate a decrease in Antarctic sea ice volume over the historical period. Because of this discrepancy, and the unclear reliability of the reanalyses ( [[#Uotila--2019|Uotila et al., 2019]] ), there is ''low confidence'' in CMIP5 and CMIP6 simulated future Antarctic sea ice thickness.

<div id="9.4" class="h1-container"></div>

<span id="ice-sheets-1"></span>