Editing IPCC:AR6/SROCC/Chapter-3 (section)

===== 3.2.1.1.1 Extent and concentration =====

The pan-Arctic loss of sea ice cover is a prominent indicator of climate change. Sea ice extent (the total area of the Arctic with at least 15% sea ice concentration) has declined since 1979 in each month of the year ( ''very high confidence'' ) (Barber et al., 2017 <sup>[[#fn:r41|41]]</sup> ; Comiso et al., 2017b <sup>[[#fn:r42|42]]</sup> ; Stroeve and Notz, 2018 <sup>[[#fn:r43|43]]</sup> ) (Figure 3.3) ''.'' Changes are largest in summer and smallest in winter, with the strongest trends in September (1979–2018; summer month with the lowest sea ice cover) of –83,000 km <sup>2</sup> yr <sup>–1</sup> (–12.8% per decade ± 2.3% relative to 1981–2010 mean), and –41,000 km <sup>2</sup> yr <sup>–1</sup> (–2.7% per decade ± 0.5% relative to 1981–2010 mean) for March (1979–2019; winter month with the greatest sea ice cover) (Onarheim et al., 2018 <sup>[[#fn:r44|44]]</sup> ). Regionally, summer ice loss is dominated by reductions in the East Siberian Sea (explains 22% of the September trend), and large declines in the Beaufort, Chukchi, Laptev and Kara seas (Onarheim et al., 2018 <sup>[[#fn:r45|45]]</sup> ). Winter ice loss is dominated by reductions within the Barents Sea, responsible for 27% of the pan-Arctic March sea ice trends (Onarheim and Årthun, 2017 <sup>[[#fn:r46|46]]</sup> ). Summer Arctic sea ice loss since 1979 is unprecedented in 150 years based on historical reconstructions (Walsh et al., 2017 <sup>[[#fn:r47|47]]</sup> ) and more than 1000 years based on palaeoclimate evidence (Polyak et al., 2010 <sup>[[#fn:r48|48]]</sup> ; Kinnard et al., 2011 <sup>[[#fn:r49|49]]</sup> ; Halfar et al., 2013 <sup>[[#fn:r50|50]]</sup> ) ( ''medium confidence'' ).

<span id="figure-3.3"></span>
<!-- START IMG -->
<!-- IMG TITLE -->
'''Figure 3.3'''

<span id="maps-of-linear-trends-in-oc-per-decade-of-arctic-a-c-and-antarctic-e-g-sea-surface-temperature-sst-for-19822017-in-march-a-e-and-september-c-g.-b-d-f-h-same-as-a-c-e-g-but-for-the-linear-trends-of-sea-ice-concentration-in-per-decade.-stippled-regions"></span>
<!-- IMG CAPTION -->
'''Maps of linear trends (in oC per decade) of Arctic (a, c) and Antarctic (e, g) sea surface temperature (SST) for 1982−2017 in March (a, e) and September (c, g). (b, d, f, h) same as (a, c, e, g), but for the linear trends of sea ice concentration (in % per decade). Stippled regions […]'''
<!-- IMG FILE -->
[[File:142cb65f02b265ad385c70283141521f IPCC-SROCC-CH_3_3.jpg]]

Maps of linear trends (in oC per decade) of Arctic (a, c) and Antarctic (e, g) sea surface temperature (SST) for 1982−2017 in March (a, e) and September (c, g). (b, d, f, h) same as (a, c, e, g), but for the linear trends of sea ice concentration (in % per decade). Stippled regions indicate the trends that are statistically insignificant. Dashed circles indicate the Arctic/Antarctic Circle. Beneath each map of linear trend shows the time series of SST (area-averaged north of 40oN/south of 40oS) or sea ice extent in the northern/southern hemisphere. Black, green, blue, orange, and red curves indicate observations, Coupled Model Intercomparison Project Phase 5 (CMIP5) historical simulation, Representative Concentration Pathway (RCP)2.6, RCP4.5, and RCP8.5 projections respectively; shading indicates ± standard deviation of multi-models. SST trend was calculated from Hadley Centre Sea Ice and Sea Surface Temperature data set (Version 1, HadISST1; Rayner, 2003). Sea ice concentration trend was calculated from the NOAA/NSIDC Climate Data Record of Passive Microwave Sea Ice Concentration, Version 3 (https://nsidc.org/data/g02202). The time series of observed SST are averages of HadISST1 and NOAA Optimum Interpolation SST dataset (version 2; Reynolds et al., 2002). The time series of observed sea ice extent are the averages of HadISST, the NOAA/NSIDC Climate Data Record of Passive Microwave Sea Ice Concentration, and the Global sea ice concentration reprocessing dataset from EUMETSAT (http://osisaf.met.no/p/ice/ice_conc_reprocessed.html).

Approximately half of the observed Arctic summer sea ice loss is driven by increased concentrations of atmospheric greenhouse gases, with the remainder attributed to internal climate variability (Kay et al., 2011 <sup>[[#fn:r51|51]]</sup> ; Notz and Marotzke, 2012 <sup>[[#fn:r52|52]]</sup> ) ( ''medium confidence'' ). The sea ice albedo feedback (increased air temperature reduces sea ice cover, allowing more energy to be absorbed at the surface, fostering more melt) is a key driver of sea ice loss (Perovich and Polashenski, 2012 <sup>[[#fn:r53|53]]</sup> ; Stroeve et al., 2012b; Serreze et al., 2016 <sup>[[#fn:r54|54]]</sup> ) and is exacerbated by the transition from perennial to seasonal sea ice (Haine and Martin, 2017 <sup>[[#fn:r55|55]]</sup> ; see Section 3.2.1.1.2). Other drivers include increased warm, moist air intrusions into the Arctic during both winter (Box 3.1) and spring (Boisvert et al., 2016 <sup>[[#fn:r56|56]]</sup> ; Cullather et al., 2016 <sup>[[#fn:r57|57]]</sup> ; Kapsch et al., 2016 <sup>[[#fn:r58|58]]</sup> ; Mortin et al., 2016 <sup>[[#fn:r59|59]]</sup> ; Graham et al., 2017 <sup>[[#fn:r60|60]]</sup> ; Hegyi and Taylor, 2018 <sup>[[#fn:r61|61]]</sup> ), radiative feedbacks associated with cloudiness and humidity (Kapsch et al., 2013 <sup>[[#fn:r62|62]]</sup> ; Pithan and Mauritsen, 2014 <sup>[[#fn:r63|63]]</sup> ; Hegyi and Deng, 2016 <sup>[[#fn:r64|64]]</sup> ; Morrison et al., 2018 <sup>[[#fn:r65|65]]</sup> ), and increased exchanges of sensible and latent heat flux from the ocean to the atmosphere (Serreze et al., 2012 <sup>[[#fn:r66|66]]</sup> ; Taylor et al., 2018 <sup>[[#fn:r67|67]]</sup> ). A lack of complete process understanding limits a more definitive differentiation between anthropogenic versus internal drivers of summer Arctic sea ice loss (Serreze et al., 2016 <sup>[[#fn:r68|68]]</sup> ; Ding et al., 2017 <sup>[[#fn:r69|69]]</sup> ; Meehl et al., 2018 <sup>[[#fn:r70|70]]</sup> ). The unabated reduction in Arctic summer sea ice since AR5 means contributions to additional global radiative forcing (Flanner et al., 2011 <sup>[[#fn:r71|71]]</sup> ) have continued, with estimates of up to an additional 6.4 ± 0.9 W/m <sup>2</sup> of solar energy input to the Arctic Ocean region since 1979 (Pistone et al., 2014 <sup>[[#fn:r72|72]]</sup> ).

Although Arctic ice freeze-up is occurring later (Section 3.2.1.1.3), rapid thermodynamic ice growth occurs over thin ice areas after air temperatures drop below freezing in autumn. Later freeze-up also delays snowfall accumulation on sea ice, leading to a thinner and less insulating snowpack (Section 3.2.1.1.6) (Sturm and Massom, 2016 <sup>[[#fn:r73|73]]</sup> ). These two negative feedbacks help to mitigate sudden and irreversible loss of Arctic sea ice (Armour et al., 2011 <sup>[[#fn:r74|74]]</sup> ).

Total Antarctic sea ice cover exhibits no significant trend over the period of satellite observations (Figure 3.3; 1979–2018) ( ''high confidence'' ) (Ludescher et al., 2018 <sup>[[#fn:r75|75]]</sup> ). A significant positive trend in mean annual ice cover between 1979 and 2015 (Comiso et al., 2017a <sup>[[#fn:r76|76]]</sup> ) has not persisted, due to three consecutive years of below average ice cover (2016–2018) driven by atmospheric and oceanic forcing (Turner et al., 2017b <sup>[[#fn:r77|77]]</sup> ; Kusahara et al., 2018 <sup>[[#fn:r78|78]]</sup> ; Meehl et al., 2019 <sup>[[#fn:r79|79]]</sup> ; Wang et al., 2019 <sup>[[#fn:r80|80]]</sup> ). The overall Antarctic sea ice extent trend is composed of near-compensating regional changes, with rapid ice loss in the Amundsen and Bellingshausen seas counteracted by rapid ice gain in the Weddell and Ross seas (Holland, 2014 <sup>[[#fn:r81|81]]</sup> ) (Figure 3.3). These regional trends are strongly seasonal in character (Holland, 2014 <sup>[[#fn:r82|82]]</sup> ); only the western Ross Sea has a trend that is statistically significant in all seasons, relative to the variance during the period of satellite observations.

Multiple factors contribute to the regionally variable nature of Antarctic sea ice extent trends (Matear et al., 2015 <sup>[[#fn:r83|83]]</sup> ; Hobbs et al., 2016b <sup>[[#fn:r84|84]]</sup> ). Sea ice trends are closely related to meridional wind trends ( ''high confidence'' ) (Holland and Kwok, 2012 <sup>[[#fn:r85|85]]</sup> ; Haumann et al., 2014 <sup>[[#fn:r86|86]]</sup> ): poleward wind trends in the Bellingshausen Sea push sea ice closer to the coast (Holland and Kwok, 2012 <sup>[[#fn:r87|87]]</sup> ) and advect warm air to the sea ice zone (Kusahara et al., 2017 <sup>[[#fn:r88|88]]</sup> ), and the reverse is true over much of the Ross Sea. These meridional wind trends are linked to Pacific variability (Coggins and McDonald, 2015 <sup>[[#fn:r89|89]]</sup> ; Meehl et al., 2016 <sup>[[#fn:r90|90]]</sup> ; Purich et al., 2016b <sup>[[#fn:r91|91]]</sup> ). Ozone depletion may also affect meridional winds (Fogt and Zbacnik, 2014 <sup>[[#fn:r92|92]]</sup> ; England et al., 2016 <sup>[[#fn:r93|93]]</sup> ), but there is ''low confidence'' that this explains observed sea ice trends (Landrum et al., 2017 <sup>[[#fn:r94|94]]</sup> ).

Coupled climate models indicate that anthropogenic warming at the surface is delayed by the Southern Ocean circulation, which transports heat downwards into the deep ocean (Armour et al., 2016 <sup>[[#fn:r95|95]]</sup> ). This overturning circulation (Cross-Chapter Box 7 in Chapter 3), along with differing cloud and lapse rate feedbacks (Goosse et al., 2018 <sup>[[#fn:r96|96]]</sup> ), may explain the weak response of Antarctic sea ice cover to increased atmospheric greenhouse gas concentrations compared to the Arctic ( ''medium confidence'' ). Because Antarctic sea ice extent has remained below climatological values since 2016, there is still potential for longer-term changes to emerge in the Antarctic (Meehl et al., 2019 <sup>[[#fn:r97|97]]</sup> ), similar to the Arctic.

Historical surface observations (Murphy et al., 2014 <sup>[[#fn:r98|98]]</sup> ), reconstructions (Abram et al., 2013b <sup>[[#fn:r99|99]]</sup> ), ship records (de la Mare, 2009; Edinburgh and Day, 2016 <sup>[[#fn:r100|100]]</sup> ), early satellite images (Gallaher et al., 2014 <sup>[[#fn:r101|101]]</sup> ), and model simulations (Gagné et al., 2015 <sup>[[#fn:r102|102]]</sup> ) indicate a decrease in overall Antarctic sea ice cover since the early 1960s which is too modest to be separated from natural variability (Hobbs et al., 2016a <sup>[[#fn:r103|103]]</sup> ) ( ''high confidence'' ).

<!-- END IMG -->
<div id="section-3-2-1-1-sea-ice-block-3"></div>

<span id="errata-figure-3.3"></span>
<!-- START IMG -->
<!-- IMG TITLE -->
'''Errata Figure 3.3'''

<span id="maps-of-linear-trends-in-oc-per-decade-of-arctic-a-c-and-antarctic-e-g-sea-surface-temperature-sst-for-19822017-in-march-a-e-and-september-c-g.-b-d-f-h-same-as-a-c-e-g-but-for-the-linear-trends-of-sea-ice-concentration-in-per-decade.-stippled-regions-1"></span>
<!-- IMG CAPTION -->
'''Maps of linear trends (in oC per decade) of Arctic (a, c) and Antarctic (e, g) sea surface temperature (SST) for 1982−2017 in March (a, e) and September (c, g). (b, d, f, h) same as (a, c, e, g), but for the linear trends of sea ice concentration (in % per decade). Stippled regions […]'''
<!-- IMG FILE -->
[[File:1eb311662e7f1e4f811ad2cb8b9fc956 Figure_Chapter_3_3_errata-3000x2397.jpg]]

Maps of linear trends (in oC per decade) of Arctic (a, c) and Antarctic (e, g) sea surface temperature (SST) for 1982−2017 in March (a, e) and September (c, g). (b, d, f, h) same as (a, c, e, g), but for the linear trends of sea ice concentration (in % per decade). Stippled regions indicate the trends that are statistically insignificant. Dashed circles indicate the Arctic/Antarctic Circle. Beneath each map of linear trend shows the time series of SST (area-averaged north of 40oN/south of 40oS) or sea ice extent in the northern/southern hemisphere. Black, green, blue, orange, and red curves indicate observations, Coupled Model Intercomparison Project Phase 5 (CMIP5) historical simulation, Representative Concentration Pathway (RCP)2.6, RCP4.5, and RCP8.5 projections respectively; shading indicates ± standard deviation of multi-models. SST trend was calculated from Hadley Centre Sea Ice and Sea Surface Temperature data set (Version 1, HadISST1; Rayner, 2003). Sea ice concentration trend was calculated from the NOAA/NSIDC Climate Data Record of Passive Microwave Sea Ice Concentration, Version 3 (https://nsidc.org/data/g02202). The time series of observed SST are averages of HadISST1 and NOAA Optimum Interpolation SST dataset (version 2; Reynolds et al., 2002). The time series of observed sea ice extent are the averages of HadISST, the NOAA/NSIDC Climate Data Record of Passive Microwave Sea Ice Concentration, and the Global sea ice concentration reprocessing dataset from EUMETSAT (http://osisaf.met.no/p/ice/ice_conc_reprocessed.html).

<!-- END IMG -->
<div id="section-3-2-1-1-sea-ice-block-4"></div>

<span id="age-and-thickness"></span>