What area of exposed rock that extends from the Great Lakes to the Arctic Ocean is one of the richest mineral regions on the planet?

3.2.1

Observed Changes in Sea Ice and Ocean

3.2.1.1

Sea Ice

Sea ice reflects a high proportion of incoming solar radiation back to space, provides thermal insulation between the ocean and atmosphere, influences thermohaline circulation, and provides habitat for ice-associated species. Sea ice characteristics differ between the Arctic and Antarctic. Expansion of winter sea ice in the Arctic is limited by land, and ice circulates within the central Arctic basin, some of which survives the summer melt season to form multi-year ice. Arctic sea ice variability and impacts on communities includes indigenous knowledge and local knowledge from across the circumpolar Arctic (Cross-Chapter Box 3 in Chapter 1). The Antarctic continent is surrounded by sea ice which interacts with adjacent ice shelves; winter season expansion is limited by the influence of the Antarctic Circumpolar Current (ACC).

Since AR5, understanding how observed changes in the Arctic can influence mid-latitude weather has emerged as a societally important topic because hundreds of millions of people can potentially be impacted (Jung et al., 2015187). The early to middle part of the Holocene coincided with substantial decreases in net precipitation that may be due to weakening jet stream winds related to Arctic temperatures (Routson et al., 2019188). There is only low to medium confidence in the current nature of Arctic/mid-latitude weather linkages because conclusions of recent analyses are inconsistent (National Research Council, 2014189; Barnes and Polvani, 2015190; Francis, 2017191). The atmosphere interacts with the ocean and cryosphere through radiation, heat, precipitation and wind, but a full understanding of complex interconnected physical processes is lacking. Arctic forcing on the atmosphere from loss of sea ice and terrestrial snow is increasing, but the potential for Arctic/mid-latitude weather linkages varies for different jet stream patterns (Grotjahn et al., 2016192; Messori et al., 2016193; Overland and Wang, 2018a194). Connectivity is reduced by the influence of chaotic internal natural variability and other tropical and oceanic forcing. Part of the scientific disagreement is due to irregular connections in the Arctic to mid-latitude linkage pathways, both within and between years (Overland and Wang, 2018b195).

Considerable literature exists on the potential for sea ice loss in the Barents and Kara Seas to drive cold episodes in eastern Asia (Kim et al., 2014196; Kretschmer et al., 2016197), while sea ice anomalies in the Chukchi Sea and areas west of Greenland are associated with cold events in eastern North America (Kug et al., 2015198; Ballinger et al., 2018199; Overland and Wang, 2018a200). Such connections, however, are only episodic (Cohen et al., 2018201). While there is evidence of an increase in the frequency of weak polar vortex events (Screen et al., 2018202), studies do not show increases in the number of mid-latitude cold events in observations or model projections (Ayarzaguena and Screen, 2016203; Trenary et al., 2016204). Potential Arctic/mid-latitude interactions have a more regional tropospheric pathway in November to December (Honda et al., 2009205; Chen et al., 2016a206; McKenna et al., 2018207), whereas January to March has a more hemispheric stratospheric pathway involving migration of the polar vortex off of its usual centred location on the North Pole (Cohen et al., 2012208; Nakamura et al., 2016209; Zhang et al., 2018b210). Overall, changes in the stratospheric polar vortex and Northern Annual Mode are not separable from natural variability, and so cannot be attributed to greenhouse gas forced sea ice loss (Screen et al., 2018211).

Only a few studies have focused on the potential impact of Antarctic sea ice changes on the mid-latitude circulation (Kidston et al., 2011212; Raphael et al., 2011213; Bader et al., 2013214; Smith et al., 2017b215; England et al., 2018216); these find that any impacts on the jet stream are strongly dependent on the season and model examined. England et al. (2018)217 suggest that the response of the jet stream to future Antarctic sea ice loss may in fact be less seasonal than the response to Arctic sea ice loss.

3.2.1.2

Ocean Properties

Ocean heat content trend (0–2000 m depth) during 2005–2017 and 1970–2017 for the global ocean and Southern Ocean. Ordinary Least Square (OLS) method is used; units are 1021 J yr-1. Uncertainties denote the 90% confidence interval accounting for the reduction in the degrees of freedom implied by temporal correlations of residuals, as per Section 5.2. Values in curved brackets are percentages of heat gain by the Southern Ocean relative to the global ocean. Data sources are as per Table SM3.1. The mean proportion and its 5–95% confidence interval (1.65 times standard deviation of individual estimates) are in the last column.

3.2.1.3

Ocean Circulation

The major elements of Southern Ocean circulation are assessed in Cross-Chapter Box 7 in Chapter 3; Arctic Ocean circulation is considered here. Arctic processes, such as discharge of freshwater from the Greenland Ice Sheet, have the potential to impact on the formation of the headwaters of the Atlantic Meridional Overturning Circulation (Section 6.7.1), and can impact on the structure and function of the marine ecosystem with implications for commercially-harvested species (Sections 3.2.3, 3.2.4).

Satellite data indicate a general strengthening of the surface geostrophic currents in the Arctic basin (Armitage et al., 2017324). Between 2003 and 2014, the strength of some currents in the Beaufort Gyre approximately doubled (Armitage et al., 2017323). Over 2001–2014, annual Bering Strait volume transport from the Pacific to the Arctic Ocean increased from 0.7×106 m3s–1 to 1.2×106 m3s–1 (Woodgate et al., 2015325). Mesoscale eddies are characterised by horizontal scales of ~10 km in the Arctic, and are important components of the ocean system. Increased wind power input to the Arctic Ocean system can in principle be compensated by the production of eddy kinetic energy; analysis of observations in the Beaufort Gyre region suggest this is about as likely as not (Meneghello et al., 2017326). Data of sufficiently high resolution is limited in the boundary regions of the Arctic Ocean, precluding estimates of eddy variability on a basin-wide scale. In the central basin regions, a statistically significant higher concentration of eddies was sampled in the Canadian Basin compared to the Eurasian Basin between 2003 and 2014; further, a medium correspondence was found between eddy activity in the Beaufort Gyre region and intensified gyre flow (Zhao et al., 2014327; Zhao et al., 2016328).

In contrast to the Southern Ocean (Cross-Chapter Box 7 in Chapter 3), there is comparatively little knowledge on changing Arctic frontal positions and current cores since AR5. An exception is that the Beaufort Gyre expanded to the northwest between 2003 and 2014, contemporaneous with changes in its freshwater accumulation and alterations in wind forcing, resulting in increased proximity to the Chukchi Plateau and Mendeleev Ridge (Armitage et al., 2017329; Regan et al., 2019330) (Section 3.2.1.2.2).

Authors: Michael P. Meredith (UK), Robert Hallberg (US), Alessandro Tagliabue (UK), Andrew Meijers (UK/Australia), Jamie Oliver (UK), Andrew Hogg (Australia)

Horizontal circulation and movement of fronts

The Southern Ocean is disproportionately important in global climate and ecological systems, being the major connection linking the Atlantic, Pacific and Indian Oceans in the global circulation. The horizontal circulation in the circumpolar Southern Ocean is comprised of an eastward-flowing mean current concentrated in a series of sinuous, braided jets exhibiting strong meandering variability and shedding small-scale transient eddies (Figure CB7.1). The mean flow circumnavigates Antarctica as the world’s largest ocean current, the Antarctic Circumpolar Current (ACC), transporting approximately 173.3 ± 10.7 ×106 m3 s–1 (Donohue et al., 2016331) of water eastward in a geostrophic balance set up by the contrasting properties of waters around Antarctica and those inside the subtropical gyres to the north of ACC. This contrast is maintained by a combination of strong westerly winds and ocean heat loss south of the ACC. 

Trends in the atmospheric forcing of the Southern Ocean are dominated by a strengthening of westerly winds in recent decades (Swart et al., 2015a332), but there is no evidence that this enhanced wind stress has significantly altered the ACC transport. While the annual mean value of transport is stable in the instrumental period (Chidichimo et al., 2014333; Koenig et al., 2014334; Donohue et al., 2016335) it is difficult to resolve changes in barotropic transport; overall there is medium confidence that ACC transport is only weakly sensitive to changes in winds. This is consistent with longer-term analyses that find only minimal changes in ACC transport since the last glaciation (McCave et al., 2013336). Theoretical predictions and high-resolution ocean modelling suggest that the weak sensitivity of the ACC to changes in wind stress is a consequence of eddy saturation (Munday et al., 2013337), whereby the time-mean state of the ocean remains close to a marginal condition for eddy instability and hence additional energy input from stronger winds cascades rapidly into the smaller-scale eddy field. Satellite measurements of eddy kinetic energy over the last two decades are consistent with this, showing a statistically significant upward trend in eddy energy in the Pacific and Indian Ocean sectors of the Southern Ocean (Hogg et al., 2015338) (medium confidence). This is supported by eddy-resolving models, which also show a marked regional variability (Patara et al., 2016339), and there is evidence that local hotspots in eddy energy, especially downstream of major topographic features including the Drake Passage, Kerguelen Plateau, Campbell Plateau and the East Pacific Rise, may dominate the regional fields (Thompson and Naveira Garabato, 2014340). 

Working Group I (WGI) of the IPCC’s 5th Assessment Report (AR5) assessed that there was medium confidence that the mean position of the ACC had moved southwards in response to a contraction of the Southern Ocean circumpolar winds. Such movements can in principle have profound effects on marine ecosystems via, e.g., changing habitat ranges for different species (e.g., Cristofari et al., 2018; Meijers et al., 2019341) (Section 3.2.3.2). Since AR5, however, substantial contrary evidence has emerged. While winds have strengthened over the Southern Ocean, reanalysis products show no significant shift in the annual mean latitude of zonal wind jets between 1979–2009 (Swart et al., 2015a342). Similarly, a variety of methods applied to satellite data have found no long-term trend and no statistically significant correlation of ACC position with winds (Gille, 2014343; Chapman, 2017344; Chambers, 2018345). The discrepancy between these studies and those assessed in WGI AR5 appears to be caused by issues associated with using a fixed sea surface height contour as a proxy for frontal position in the presence of strongly eddying fields (Chapman, 2014346) and large-scale increases in sea surface height consistent with mean global trends in sea level rise (Gille, 2014347). The increase in sea surface height is ascribed largely to warming-driven steric expansion in the upper ocean, but the mechanism driving such warming is still uncertain (Gille, 2014348). These recent findings do not preclude more local changes in frontal position, but it is now assessed as unlikely that there has been a statistically significant net southward movement of the mean ACC position over the past 20 years.

Overturning circulation and water mass formation

The Southern Ocean is the key region globally for the upwelling of interior ocean waters to the surface, enabling waters that were last ventilated in the pre-industrial era to interact with the industrial-era atmosphere and the cryosphere. New water masses are produced that sink back into the ocean interior. Such export of both extremely cold and dense Antarctic Bottom Water and the lighter mode and intermediate waters (Figure CB7.1) represents important pathways for surface properties to be sequestered from the atmosphere for decades to millennia. This upwelling and sinking constitutes a two-limbed overturning circulation, by which much of the global deep ocean is renewed. 

The Southern Ocean overturning circulation plays a strong role in mediating climate change via the transfer of heat and carbon (including that of anthropogenic origin) with the atmosphere (Sections 3.2.1.2; 5.2.2.2); it also has an impact on sea ice extent and concentration, with implications for climate via albedo (Section 3.2.1.1). It acts to oxygenate the ocean interior and sequesters nutrients that ultimately end up supporting a significant fraction of primary production in the rest of the world ocean (Section 5.2.2.2). The upwelling waters in the overturning bring heat to the Antarctic shelf seas, with consequences for ice shelves, marine-terminating glaciers and the stability of the Antarctic Ice Sheet (AIS) (Section 3.3.1). The lower limb of this overturning circulation supplies Antarctic Bottom Water that forms the abyssal layer of much of the world ocean (Section 3.2.1.2; 5.2.2.2).

It is challenging to measure the Southern Ocean overturning directly, and misinterpretation of Waugh et al. (2013)349 led to AR5 erroneously reporting the upper cell to have slowed (AR5 WGI, Section 3.6.4). However, additional indirect estimates since AR5 provide support for the increase in the upper ocean overturning proposed by Waugh et al. (2013). Waugh (2014)350 and Ting and Holzer (2017)351 suggest that over the 1990s–2000s water mass ages changed in a manner consistent with an increase in upwelling and overturning. However, inverse analyses suggest that such overturning experiences significant inter-decadal variability in response to wind forcing, with reductions in 2000–2010 relative to 1990–2000 (DeVries et al., 2017352). This variability, combined with the indirect nature of observational estimates, means that there is low confidence in assessments of long-term changes in upper cell overturning.

Available evidence indicates that the volume of Antarctic Bottom Water in the global ocean has decreased (Purkey and Johnson, 2013353; Desbruyeres et al., 2017354) (medium confidence), thinning at a rate of 8.1 m yr−1 since the 1950s (Azaneu et al., 2013355); recently updated analyses confirm this trend to present day (Figure 5.4). This suggests that the production and export of this water mass has probably slowed, though direct observational evidence is difficult to obtain. The large-scale impacts of Antarctic Bottom Water changes include a potential modulation to the strength of the Atlantic Meridional Overturning Circulation (e.g., Patara and Böning, 2014; see also Section 5.2.2.2.1).

Projections

Projections of future trends in the Southern Ocean are dominated by the potential for a continued strengthening of the westerly winds (Bracegirdle et al., 2013356), as well as a combination of warming and increased freshwater input from both increased net precipitation and changes in sea ice export (Downes and Hogg, 2013357). Dynamical considerations and numerical simulations indicate that, if further increases in the westerly winds are sustained, then it is very likely that the eddy field will continue to grow in intensity (Morrison and Hogg, 2013358; Munday et al., 2013359), with potential consequences for the upper-ocean overturning circulation and transport of tracers (Abernathey and Ferreira, 2015360) (including heat, carbon, oxygen and nutrients), and likely that the mean position and strength of the ACC will remain only weakly sensitive to winds.

The considerable Coupled Model Intercomparison Project Phase 5 (CMIP5) inter-model variations in Southern Ocean time-mean circulation projections reported in WGI AR5 (Meijers et al., 2012361; Downes and Hogg, 2013362) remain largely unchanged. Some of the differences in projected changes have been found to be correlated with biases in the various models’ ability to simulate the historical state of the Southern Ocean, such as mixed layer depth (Sallée et al., 2013a363) and westerly wind jet latitude (Bracegirdle et al., 2013364). This suggests that bias reduction against observed historical metrics (Russell et al., 2018365) in future generations of coupled models (e.g., Coupled Model Intercomparison Project Phase 6 (CMIP6)) should lead to improved confidence in aspects of projected Southern Ocean changes. CMIP5 models suggest that the subduction of mode and intermediate water will increase (Sallée et al., 2013b366), which will affect oxygen and nutrient transports, and the overall transport of the Southern Ocean upper overturning cell will increase by up to 20% (Downes and Hogg, 2013367), but model performance is limited by the inability to explicitly resolve eddy processes (Gent, 2016368; Downes et al., 2018369). The formation and export of Antarctic Bottom Water is predicted to continue decreasing (Heuzé et al., 2015370) due to warming and freshening of surface source waters near the continent. These are, however, some of the most poorly represented processes in global models. Further uncertainty derives from increased meltwater from the AIS not being considered in the CMIP5 climate models, despite its potential for significant impact on Southern Ocean dynamics and the global climate, and its potential for positive feedbacks (Bronselaer et al., 2018371). Due to these uncertainties, low confidence is therefore ascribed to the CMIP5-based model projections of future Southern Ocean circulation and water masses.

3.2.2

Projected Changes in Sea Ice and Ocean

3.2.2.1

Sea Ice

The multi-model ensemble of historical simulations from CMIP5 models identify declines in total Arctic sea ice extent and thickness (Sections 3.2.1.1.1; 3.2.1.1.2; Figure 3.3) which agree with observations (Massonnet et al., 2012372; Stroeve et al., 2012a373; Stroeve et al., 2014a374; Stroeve and Notz, 2015375). There is a range in the ability of individual models to simulate observed sea ice thickness spatial patterns and sea ice drift rates (Jahn et al., 2012376; Stroeve et al., 2014a377; Tandon et al., 2018378). Reductions in Arctic sea ice extent scale linearly with both global temperatures and cumulative CO2 emissions in simulations and observations (Notz and Stroeve, 2016379), although aerosols influenced historical sea ice trends (Gagné et al., 2017380). The uncertainty in sea ice sensitivity (ice extent loss per unit of warming) is quite large (Niederdrenk and Notz, 2018381) and the model sensitivity is too low in most CMIP5 models (Rosenblum and Eisenman, 2017382). Emerging evidence suggests, however, that internal variability, including links between the Arctic and lower latitude, strongly influences the ability of models to simulate observed reductions in Arctic sea ice extent (Swart et al., 2015b383; Ding et al., 2018384).

CMIP5 models project continued declines in Arctic sea ice through the end of the century (Figure 3.3) (Notz and Stroeve, 2016385) (high confidence). There is a large spread in the timing of when the Arctic may become ice free in the summer, and for how long during the season (Massonnet et al., 2012386; Stroeve et al., 2012a387; Overland and Wang, 2013388) as a result of natural climate variability (Notz, 2015389; Swart et al., 2015b390; Screen and Deser, 2019391), scenario uncertainty (Stroeve et al., 2012a392; Liu et al., 2013393), and model uncertainties related to sea ice dynamics (Rampal et al., 2011394; Tandon et al., 2018395) and thermodynamics (Massonnet et al., 2018396). Internal climate variability results in an uncertainty of approximately 20 years in the timing of seasonally ice-free conditions (Notz, 2015397; Jahn, 2018398), but the clear link between summer sea ice extent and cumulative CO2 emissions provides a basis for when consistent ice-free conditions may be expected (high confidence). For stabilised global warming of 1.5°C, sea ice in September is likely to be present at end of century with an approximately 1% chance of individual ice-free years (Notz and Stroeve, 2016399; Sanderson et al., 2017400; Jahn, 2018401; Sigmond et al., 2018402); after 10 years of stabilised warming at a 2°C increase, more frequent occurrence of an ice-free summer Arctic is expected (around 10-35%) (Mahlstein and Knutti, 2012403; Jahn et al., 2016404; Notz and Stroeve, 2016405). Model simulations show that a temporary temperature overshoot of a warming target has no lasting impact on ice cover (Armour et al., 2011406; Ridley et al., 2012407; Li et al., 2013408).

CMIP5 models show a wide range of mean states and trends in Antarctic sea ice (Turner et al., 2013409; Shu et al., 2015410). The ensemble mean across multiple models show a decrease in total Antarctic sea ice extent during the satellite era, in contrast to the lack of any observed trend (Figure 3.3; Section 3.2.1.1.1). Interannual sea ice variability in the models is larger than observations (Zunz et al., 2013411), which may mask disparity between models and observations. Internal variability (Polvani and Smith, 2013412; Zunz et al., 2013413), and model sensitivity to warming (Rosenblum and Eisenman, 2017414) are also important sources of uncertainty. During the historical period, regional trends of Antarctic sea ice are not captured by the models, particularly the decrease in the Bellingshausen Sea and the expansion in the Ross Sea (Hobbs et al., 2015415). There is a very wide spread of model responses in the Weddell Sea (Hobbs et al., 2015416; Ivanova et al., 2016417), a region with complex ocean-sea ice interactions that many models do not replicate (de Lavergne et al., 2014).

There is low confidence in projections of Antarctic sea ice because there are multiple anthropogenic forcings (ozone and greenhouse gases) and complicated processes involving the ocean, atmosphere, and adjacent ice sheet (Section 3.2.1.1.). Model deficiencies are related to stratification (Sallée et al., 2013a418), freshening by ice shelf melt water (Bintanja et al., 2015419), atmospheric processes including clouds (Schneider and Reusch, 2015420; Hyder et al., 2018421), and wind and ocean driven processes (Purich et al., 2016a422; Purich et al., 2016b423; Schroeter et al., 2017424; Purich et al., 2018425; Zhang et al., 2018a426). Uncertainty in sea ice projections reduces confidence in projections of Antarctic Ice Sheet surface mass balance because sea ice affects Antarctic temperature and precipitation trends (Bracegirdle et al., 2015427), and impacts projected changes in the Southern Hemisphere westerly jet (Bracegirdle et al., 2018428; England et al., 2018429) with implications for the Southern Ocean overturning circulation (Cross-Chapter Box 7 in Chapter 3).

Arctic Coastal Polynyas

Arctic polynyas (areas of open water surrounded by sea ice) are important because they ventilate the Arctic Ocean. The polynyas induce bottom reaching convection on shallow shelves (Damm et al., 2018430) because the warm and exposed ocean surface creates very high heat fluxes and new sea ice formation during winter, releasing brine and creating dense water (Barber et al., 2012431). On the shallow Siberian shelves, the ocean surface waters are dominated by river runoff which is rich with sediments (Damm et al., 2018432), which end up both in the dense bottom water and in new sea ice (Bauch et al., 2012433; Janout et al., 2015434). This process maintains the Arctic Ocean halocline (Bauch et al., 2011435), which insulates the sea ice cover from the heat of the underlying Atlantic-derived waters.

Polynyas are projected to change in different ways depending on regional ice conditions and ice formation processes. Further reductions in sea ice are projected for Arctic shelf seas which have already lost ice in recent decades (Barnhart et al., 2015436; Onarheim et al., 2018437) so polynyas will cease to exist where seasonal sea ice disappears or evolve to become part of the marginal sea ice zone due to changes in ice dynamics (i.e., the North Water polynya and the Circumpolar Flaw Lead); new or enlarged polynyas could result in regions where thinner ice becomes more effectively advected offshore, or where marine terminating glaciers increase land ice fluxes to the marine system (medium confidence). The reduced survival rate of sea ice in the Transpolar Drift interrupts the transport of sediment-laden ice produced from Siberian shelf polynyas (Krumpen et al., 2019438), with consequences for the associated biogeochemical matter and gas fluxes (Damm et al., 2018439) (medium confidence).

Projected changes to polynyas are important because the spring phytoplankton bloom starts early as the ocean is often well-ventilated and nutrient rich, so the entire biological range from phytoplankton to seabirds to marine mammals thrive in polynya waters (high confidence) (Stirling, 1997440; Arrigo and van Dijken, 2004441; Karnovsky et al., 2009442). Secondary production and upper food web processes are typically adapted to the early availability of energy to the system with arrival of higher trophic species (Asselin et al., 2011443). Because of the abundance of marine food resources including seals, whales and fish in and around polynyas, Arctic peoples have hunted regularly in these areas for thousands of years (Barber and Massom, 2007444). Recent implementation of Inuit-led marine management areas acknowledge the Inuit knowledge of polynyas, and recognise the potential for development of fisheries and other resources in polynya systems, provided these activities minimise harm on the environment and wildlife. The Inuit Circumpolar Council’s Pikialasorsuaq Commission is an example of a proposal to develop an Inuit management area in the North Water Polynya (Cross-Chapter Box 3 in Chapter 1).

Antarctic Coastal Polynyas

The Antarctic continent is surrounded by coastal polynyas, which form from the combined effects of winds and landfast ice in the lee of coastal features that protrude into the westward coastal current (Nihashi and Ohshima, 2015445; Tamura et al., 2016446). Intense ice growth within these polynyas contributes to the production of Antarctic Bottom Water, the densest and most voluminous water mass in the global ocean (Jacobs, 2004447; Nicholls et al., 2008448; Orsi and Wiederwohl, 2009449; Ohshima et al., 2013450). Sea ice production is greatest in Ross and Weddell sea polynyas and around East Antarctica (Drucker et al., 2011451; Nihashi and Ohshima, 2015452; Tamura et al., 2016453) (high confidence).

Antarctic coastal polynyas are biological hot-spots that support high rates of primary production (Ainley et al., 2015454; Arrigo et al., 2015455) due to a combination of both high light (Park et al., 2017456) and high nutrient levels, especially iron (Gerringa et al., 2015457). Basal ice shelf melt is the primary supplier of iron to coastal polynyas (Arrigo and van Dijken, 2015458) although sea ice melt and intrusions of Circumpolar Deep Water are significant in the Ross Sea (McGillicuddy et al., 2015459; Hatta et al., 2017460). As ice shelves retreat, the polynyas created in their wake also increase local primary production: the new polynyas created after the collapse of the Larsen A and B ice shelves are as productive as other Antarctic shelf regions, likely increasing organic matter export and altering marine ecosystem evolution (Cape et al., 2013461). The recent calving of Mertz Glacier Tongue in East Antarctica has altered sea ice and ocean stratification (Fogwill et al., 2016462) such that polynyas there are now twice as productive (Shadwick et al., 2017463). 

The productivity associated with these polynyas is a critical food source for some of the most abundant top predators in Antarctic waters, including penguins, albatross and seals (Raymond et al., 2014464; Malpress et al., 2017465) (Section 3.2.3.2.4). However, only a fraction of the carbon fixed by phytoplankton in coastal polynyas is consumed by upper trophic levels. The rest sinks to the seafloor where it is re-mineralised or sequestered (Shadwick et al., 2017466), or is advected off the shelf (Lee et al., 2017b467). Given the high amount of residual macronutrients in polynya surface waters, there is evidence that future changes in ice shelf melt rates could increase water column productivity (Gerringa et al., 2015468; Rickard and Behrens, 2016469; Kaufman et al., 2017470), influencing Antarctic coastal ecosystems and increasing the ability of continental shelf waters to sequester atmospheric carbon dioxide (Arrigo and van Dijken, 2015471).

The Weddell Polynya

The Weddell Polynya is a large area of open water within the winter ice pack of the Weddell Sea close to the Maud Rise seamount (at approximately 65ºS, 3ºE), and has importance on a global scale for deep water ventilation. The polynya opens intermittently, and remained open from 1974 to 1976, with an area of 0.2–0.3 million km2 (Carsey, 1980472). A similar polynya appeared in spring 2017, with a smaller area in 2016, but did not occur in 2018 (Campbell et al., 2019473; Jena et al., 2019474). Based on these recent events, there is medium confidence in the drivers of Weddell Polynya formation; it forms over deep water and appears connected to sea ice divergence created by ocean eddies (Holland, 2001475) or strong winds (Campbell et al., 2019476; Francis et al., 2019477; Wilson et al., 2019478). Around Maud Rise, the ocean is weakly stratified, and winter sea ice formation causes brine release and the related deepening mixed layer brings warmer deep waters towards the surface. This causes heat loss to the atmosphere above 200 W m–² (Campbell et al., 2019479). These polynya formation processes cause deep ocean convection that releases heat from the deep ocean to the atmosphere (Smedsrud, 2005480), and may contribute to the uptake of anthropogenic carbon (Bernardello et al., 2014481).

In some CMIP5 models, phases of Weddell polynya activity appear for decades or centuries at a time, and then cease for a similar period (Reintges et al., 2017482). The observational era is not sufficiently long to rule out this behaviour. Models indicate that under anthropogenic climate change, surface freshening caused by increased precipitation reduces the occurrence of the Weddell polynya (de Lavergne et al., 2014). There are systematic biases in modelled ocean stratification resulting in low confidence in future Weddell Polynya projections (Reintges et al., 2017483).

3.2.2.2

Physical Oceanography

Consistent with the projected sea ice decline, there is high confidence that the Arctic Ocean will warm significantly towards the end of this century at the surface and in the deeper layers. Most CMIP5 models capture the seasonal changes in surface heat and freshwater fluxes for the present day climate, and show that the excess summer solar heating is used to melt sea ice, in a positive ice albedo feedback (Ding et al., 2016484). Using RCP8.5, Vavrus et al. (2012) found that the Atlantic layer is projected to warm by 2.5°C at around 400 m depth at the end of the century, but only by 0.5°C in the surface mixed layer. Consistent results for lower Atlantic Water layer warming were found by Koenigk and Brodeau (2014)485 for RCP2.5 (0.5°C), RCP4.5 (1.0°C) and RCP8.5 (2.0°C).

Poleward ocean heat transport contributes to Arctic Ocean warming (medium confidence). Comparing 20 CMIP5 simulations for RCP8.5, Nummelin et al. (2017) found a 2°C–6°C range in Arctic amplification of surface air temperature north of 70°N, consistent with increased ocean heat transport. Comparing 26 different CMIP5 simulations for RCP4.5, Burgard and Notz (2017) found that ocean heat transport changes explain the Arctic Ocean multi-model mean warming, but that differences between models are compensated by changes in surface fluxes. Increased ocean heat transport into the Barents Sea beyond 2020 appears as a probable mechanism with continued warming (Koenigk and Brodeau, 2014486; Årthun et al., 2019487). Based on four CMIP5 models, the Barents Sea is projected to become ice-free during winter beyond 2050 under RCP8.5 (Onarheim and Årthun, 2017488), to which the main response is an increased ocean-to-atmosphere heat flux and related surface warming (Smedsrud et al., 2013489). The ocean heat transport increases in all Arctic gateways, but is dominated by the Barents Sea, and when winter sea ice disappears here the heat loss cannot increase further and the excess ocean heat continues into the Arctic Basin (Koenigk and Brodeau, 2014490). 

The surface mixed layer of the Arctic Ocean is expected to freshen in future because an intensified hydrological cycle will increase river runoff (Haine et al., 2015491) (medium confidence). The related increase in stratification has the potential to contribute to the warming of the deep Atlantic Water layer, as upward vertical mixing will be reduced (Nummelin et al., 2016492). There are, however, biases in salinity of ~1 across the Arctic Basin for the present day climate (Ilicak et al., 2016493) in forced global ice-ocean models with configurations comparable to CMIP5, suggesting limited predictive skill for the Arctic freshwater cycle.

CMIP5 projections (Figure 3.3) indicate that observed Southern Ocean warming trends will continue under RCP4.5 and RCP8.5 scenarios, leading to 1°C–3°C warming by 2100 mostly in the upper ocean (Sallée et al., 2013a494). Projections demonstrate a similar distribution of heat storage to historical observations, notably focused in deep pools north of the Subantarctic Front (e.g., Armour et al., 2016495). Antarctic Bottom Water becomes coherently warmer by up to 0.3°C by 2100 across the model ensemble under RCP8.5 (Heuzé et al., 2015496). The upper ocean also becomes considerably fresher (salinity decrease of approximately 0.1) (Sallée et al., 2013b497) with an overall increase in stratification and a shallowing of mixed layers (Sallée et al., 2013a498). Although the sign of model changes appear mostly robust, there is low confidence in magnitude due to the large inter-model spread in projections and significant warm biases in historical water mass properties (Sallée et al., 2013a499) and sea surface temperature, which may be up to 3°C too high in the historical runs (Wang et al., 2014500). Projections of changes in Southern Ocean circulation are discussed in Cross-Chapter Box 7 in Chapter 3.

3.2.2.3

Carbon and Ocean Acidification

The Arctic and Southern Ocean have a systemic vulnerability to aragonite undersaturation (Orr et al., 2005501). For the RCP8.5 scenario, the entire Arctic and Southern Ocean surface waters will very likely be typified by year-around conditions corrosive for aragonite minerals for 2090–2100 (Figure 3.4) (Hauri et al., 2015502; Sasse et al., 2015503), whilst under RCP2.6 the extent of undersaturated waters are reduced markedly. At a basin/circumpolar scale, there is high confidence in these projections due to our robust understanding of the driving mechanisms. However, there is medium confidence for the response of specific locations, due to the need for improved resolution of the local circulation, interactions with sea ice, and other processes that modulate the rate of acidification.

Under RCP8.5, melting ice causes the greatest declining rate of pH and CaCO3 saturation state in the Central Arctic, Canadian Arctic Archipelago and Baffin Bay (Popova et al., 2014504). In the Canada Basin, projections using RCP8.5 show reductions in mean surface pH from approximately 8.1 in 1986–2005 to 7.7 by 2066–2085, and aragonite saturation from 1.52–0.74 during the same period (Steiner et al., 2014505). A shoaling of the aragonite saturation horizon of approximately 1200 m, a large increase in area extent of undersaturated surface waters, and a pH change in the surface water of –0.19 are projected using the SRES A1B scenario (broadly comparable to RCP6.0) in the Nordic Sea from 2000 to 2065 (Skogen et al., 2014506). Under the same scenario, aragonite undersaturation is projected to occur in the bottom waters over the entire Kara Sea shelf by 2040 and over most of the Barents and East Greenland shelves by 2070 due to the accumulation of anthropogenic CO2 (Wallhead et al., 2017507).

Under RCP8.5, the rate of CO2 uptake by the Southern Ocean is projected to increase from the contemporary 0.91 Pg C yr–1 to 2.38(1.65–2.55) Pg C yr–1 by 2100, but the growth in uptake rate will slow and likely stop around 2070 ± 10 corresponding to cumulative CO2 emissions of 1600 Gt C (Kessler and Tjiputra, 2016508; Wang et al., 2016b509). This halt in the increase in the uptake rate of CO2 is linked to the combined feedbacks from well-understood reductions in buffering capacity and warming, as well as the increased upwelling rate of carbon-rich Circumpolar Deep Water (Hauck and Volker, 2015509) (Cross-Chapter Box 7 in Chapter 3). Although there is high agreement amongst models, contemporary biases in the fluxes of CO2 in CMIP5 models in the Southern Ocean (Mongwe et al., 2018510) suggest medium confidence levels for these projections.

Alongside the mean state changes, Southern Ocean aragonite saturation is also affected by the seasonal cycle of carbonate as well as by the impact of reduced buffering capacity (SM3.2.4) on the seasonal cycle of CO2 (Sasse et al., 201511v; McNeil and Sasse, 2016512). This leads to an amplification of the seasonal variability of pCO2 (Hauck and Volker, 2015513; McNeil and Sasse, 2016514; Landschützer et al., 2018515) and the hydrogen ion concentration that accelerates the onset of hypercapnia (i.e., high pCO2 levels; pCO2 > 1000 μatm) to nearly 2 decades (~2085) ahead of anthropogenic CO2 forcing (McNeil and Sasse, 2016516). The seasonal cycles of pH and aragonite saturation will be attenuated (Kwiatkowski and Orr, 2018517) (Section 5.2.2.3), however when the mean state changes are combined with the changes in seasonality, the onset of undersaturation is brought forward by 10–20 years (Table SM3.5). Model projections remain uncertain and affected by the resolution of local ocean physics, which leads to overall medium confidence in the timing of undersaturation and hypercapnia. 

3.2.3

Impacts on Marine Ecosystems

3.2.3.1

Arctic

Climate change has, and is projected to continue to have, significant implications for Arctic marine ecosystems, with consequences at different trophic levels both in the pelagic, benthic, and sympagic (sea ice related) realms (Figure 3.5). Specifically, climate change is projected to alter the distribution and properties of Arctic marine habitats with associated implications for species composition, production and ecosystem structure and function (Frainer et al., 2017518; Kaartvedt and Titelman, 2018519; Moore et al., 2018520). The rate and severity of ecosystem impacts will be spatially heterogeneous and dependent on future emission scenarios.

In the few Arctic regions where data is sufficient to assess trends in biodiversity, the ecosystem level responses appear to be products of multiple interacting physical, chemical and biological processes (Frederiksen, 2017521) (medium confidence). Climate change impacts on vertical fluxes and stratification (Sections 3.2.1.2.3, 3.2.2.2) will contribute to changes in bentho-pelagic-sympagic coupling. For instance, projected climate driven changes in ocean properties and hydrography (Section 3.2.2.2) and the abundance of pelagic grazers (Box 3.4) could alter the export of organic matter to the sea floor with associated impacts on the benthos in some Arctic shelf ecosystems (Moore and Stabeno, 2015522; Stasko et al., 2018523) (low confidence). Projected future reductions in summer sea ice (Section 3.2.1.1), increased stratification in summer, shifting currents and fronts and increased ocean temperatures (Section 3.2.2.2) and ocean acidification (Section 3.2.2.3) are all expected to impact the future production and distribution of several marine fish and invertebrates (high confidence). 

Ocean acidification (Section 3.2.2.3) will affect several key Arctic species (medium confidence). The effects of current and projected levels of acidification have been examined for a broad suite of species groups (bivalves, cephalopods, echinoderms, crustaceans, corals and fishes) and these studies reveal species-specific differences in sensitivity, as well as differences in the scope for, and energetic cost of, adaptation (Luckman et al., 2014524; Howes et al., 2015525; Falkenberg et al., 2018526). 

3.2.3.2

3.2.3.2 Southern Ocean

Marine ecosystem dynamics in the Antarctic region are dominated by the ACC and its frontal systems (Cross-Chapter Box 7 in Chapter 3), subpolar gyres, polar seasonality, the annual advance and retreat of sea ice (Section 3.2.1.1) and the supply of limiting micronutrients for productivity (most commonly iron) (Section 5.2.2.5). Antarctic krill (Euphausia superba) play a central role in Southern Ocean foodwebs as grazers and as prey items for fish, squid, marine mammals and seabirds (Schmidt and Atkinson, 2016683; Trathan and Hill, 2016684) (SM3.2.6). This is due in part to the high abundance and circumpolar distribution of Antarctic krill, although the abundance and importance of this species varies between different regions of the Southern Ocean (Larsen et al., 2014685; Siegel, 2016686; McCormack et al., 2017687). Recent work has characterised the nature of habitat change for Southern Ocean biota at regional and circumpolar scales (Constable et al., 2014688; Gutt et al., 2015689; Constable et al., 2016690; Hunt et al., 2016691; Gutt et al., 2018692), and the direct responses of biota to these changes (Constable et al., 2014693) (summarised in Figure 3.6). These findings indicate that overlapping changes in key ocean and sea ice habitat characteristics (temperature, sea ice cover, iceberg scour, mixed layer depth, aragonite undersaturation; Sections 3.2.1, 3.2.2) will be important in determining future states of Southern Ocean ecosystems (Constable et al., 2014694; Gutt et al., 2015695) (medium confidence). However, there is a need to better characterise the nature and importance of indirect responses to physical change using models and observations. Important advances have also been made since AR5 in (i) identifying key variables to detect and attribute change in Southern Ocean ecosystems, as part of long-term circumpolar modelling designs (Constable et al., 2016696), and (ii) refining methods for using sea ice projections from global climate models in ecological studies and ecosystem models for the Southern Ocean (Cavanagh et al., 2017697).

3.2.4

Impacts on Social-Ecological Systems

3.2.4.1

Fisheries

3.2.4.2

Tourism

Reductions in sea ice have facilitated an increase in marine and cruise tourism opportunities across the Arctic related to an increase in accessibility (Dawson et al., 2014856; Johnston et al., 2017857) (high confidence). While not exclusively ‘polar’, Alaska attracts the highest number of cruise passengers annually at just over one million; Svalbard attracts 40,000–50,000; Greenland 20,000–30,000; and Arctic Canada 3,500–5,000 (Johnston et al., 2017858). Compared to a decade ago, there are more cruises on offer, ships travel further in a single season, larger vessels with more passenger berths are in operation, more purpose-built polar cruise vessels are being constructed, and private pleasure craft are appearing in the Arctic more frequently (Lasserre and Têtu, 2015859; Johnston et al., 2017860; Dawson et al., 2018861). In Antarctica, almost 37,000 (predominantly shipborne) tourists visited in 2016–2017, with 51,707 during 2017–2018; there were 6700 tourists in 1992–1993 (the first year of record) (ATCM, 2018862). Due to accessibility and convenience, these tourism operations are mostly based around the few ice-free areas of Antarctica, concentrated on the AP (Pertierra et al., 2017863).

Canada’s Northwest Passage (southern route), which only saw occasional cruise ship transits in the early 2000s is now reliably accessible during the summer cruising season, and as a result has experienced a doubling and quadrupling of cruise and pleasure craft activity over the past decade (Johnston et al., 2017864; Dawson et al., 2018865). There is high confidence that demand for Arctic cruise tourism will continue to grow over the coming decade (Johnston et al., 2017866). The anticipated implications of future climate change have become a driver for polar tourism. A niche market known as ‘last chance tourism’ has emerged whereby tourists explicitly seek to experience vanishing landscapes or seascapes, and natural and social heritage in the Arctic and Antarctic, before they disappear (Lemelin et al., 2010; Lamers et al., 2013867).

Increases in polar cruise tourism pose risks and opportunities related to development, education, safety (including search and rescue), security within communities and environmental sustainability (Johnston et al., 2012a868; Johnston et al., 2012b869; Stewart et al., 2013870; Dawson et al., 2014871; Lasserre and Têtu, 2015872; Stewart et al., 2015873). In the Arctic, there are also risks and opportunities related to employment, health and well-being, and the commodification of culture (Stewart et al., 2013874; Stewart et al., 2015875). There is high confidence that biodiversity supported by ice-free areas, particularly those on the AP, are vulnerable to the introduction of terrestrial alien species via tourists and scientists (Chown et al., 2012876; Huiskes et al., 2014877; Hughes et al., 2015878; Duffy et al., 2017879; Lee et al., 2017a880) (Box 3.3) as well as to the direct impacts of humans (Pertierra et al., 2017881). The tourism sector relies on a set of regulations that apply to all types of maritime shipping, yet cruise ships intentionally travel off regular shipping corridors and serve a very different purpose than other vessel types, so there is a need for region-specific governance regimes, specialised infrastructure, and focused policy attention (Dawson et al., 2014882; Pashkevich et al., 2015883; Pizzolato et al., 2016884; Johnston et al., 2017885). Private pleasure craft remain almost completely unregulated, and will pose unique risks in the future (Johnston et al., 2017886).

3.2.4.3

Transportation

The Arctic is reliant on marine transportation for the import of food, fuel and other goods. At the same time, the global appetite for maritime trade and commerce through the Arctic (including community re-supply, mining and resource development, tourism, fisheries, cargo, research, and military and icebreaking, etc.) is increasing as the region becomes more accessible because of reduced sea ice cover. There are four potential Arctic international trade routes: the Northwest Passage, the Northern Sea Route, the Arctic Bridge and the Transpolar Sea Route. All of these routes offer significant trade benefits because they provide substantial distance savings compared to traditional routes via the Suez or Panama Canals.

There is high confidence that shipping activity during the Arctic summer increased over the past two decades in regions for which there is information, concurrent with reductions in Arctic sea ice extent and the shift to predominantly seasonal ice cover (Pizzolato et al., 2014887; Eguíluz et al., 2016888; Pizzolato et al., 2016889). Long term datasets over the pan-Arctic are incomplete, but the distance travelled by ships in Arctic Canada nearly tripled between 1990 and 2015 (from ~365,000 to ~920,000 km) (Dawson et al., 2018890). Other non-environmental factors which influence Arctic shipping are natural resource development, regional trade, geopolitics, commodity prices, global economic and social trends, national priorities, tourism demand, ship building technologies and insurance costs (Lasserre and Pelletier, 2011891; Têtu et al., 2015892; Johnston et al., 2017893). Current impacts associated with the observed increase in Arctic shipping include a higher rate of reported accidents per km travelled compared to southern waters (CCA, 2016), increases in vessel noise propagation (Halliday et al., 2017894) and air pollution (Marelle et al., 2016895). Disruptions to cultural and subsistence hunting activities from increased shipping (Huntington et al., 2015896; Olsen et al., 2019897) compound climate-related impacts to people (Sections 3.4.3.3.2, 3.4.3.3.3). 

It is projected that shipping activity will continue to rise across the Arctic as northern routes become increasingly accessible (Stephenson et al., 2011898; Stephenson et al., 2013899; Barnhart et al., 2015900; Melia et al., 2016901), although mitigating economic and operational factors remain uncertain and could influence future traffic volume (Zhang et al., 2016902). The Northern Sea Route is expected to be more viable than other routes because of infrastructure already in place (Milaković et al., 2018903); favourable summer ice conditions in recent years have reduced transit times (Aksenov et al., 2017904). In comparison, the Northwest Passage and Arctic Bridge presently have limited port and marine transportation infrastructure, incomplete soundings and hydrographic charting, challenging sea ice conditions and limited search and rescue capacity; these compound the risks from shipping activity (Stephenson et al., 2013905; Johnston et al., 2017906; Andrews et al., 2018907). 

Future shipping impacts will be regionally diverse considering the unique geographies, sea ice dynamics, infrastructure and service availability and regulatory regimes that exist across different Arctic nations. Considerations include socioeconomic and political implications related to safety (marine and local accidents), security (trafficking, terrorism and local issues), and environmental and cultural sustainability (invasive species, release of biocides, chemicals and other waste, marine mammal strikes, fuel spills, air and underwater noise pollution and impacts to subsistence hunting) (Arctic Council, 2015a908; Halliday et al., 2017909; Hauser et al., 2018910). Black carbon emissions from shipping activity within the Arctic are projected to increase (Arctic Council, 2017911) and are more easily deposited at the surface in the region compared with emissions from lower latitudes (Sand et al., 2013912). Commercial shipping mainly uses heavy fuel oil, with associated emissions of sulphur, nitrogen, metals, hydrocarbons, organic compounds, black carbon and fly ash to the atmosphere during combustion (Turner et al., 2017a913). Mitigation approaches include banning heavy fuel oil as already implemented in Antarctica and the waters around Svalbard, and the use of new technology like scrubbers.

The predominant shipborne activities in Antarctica are fishing, logistic support to land-based stations, and marine research vessels operating for both non-governmental and governmental sectors. Uncertainty in future Antarctic sea ice conditions (Section 3.2.2.1) pose challenges to considering potential impacts on these activities (Chown, 2017914).

Changes in ice sheet mass have been derived repeatedly over the satellite era using complementary methods based on time series of satellite altimetry to measure volume change, ice-flux measurements combined with modelled surface mass balance (SMB) to calculate mass inputs and outputs, and satellite gravimetry to measure regional mass change. Ice sheet changes over earlier periods have also been reconstructed from firn/ice core and geological evidence (SM3.3.1).

It is virtually certain that the Antarctic Peninsula (AP) and West Antarctic Ice Sheet (WAIS) combined have cumulatively lost mass since widespread measurements began in 1992, and that the rate of loss has increased since around the year 2006 and continued post-AR5 (Martín‐Español et al., 2016915; Zwally et al., 2017916; Bamber et al., 2018917; Gardner et al., 2018918; The IMBIE Team, 2018; Rignot et al., 2019919), extending and reinforcing previous findings (IPCC, 2013920) (Figure 3.7, Table 3.3, SM3.3.1.1). From medium evidence, there is high agreement in the sign and medium agreement in the magnitude of both WAIS and AP mass change between the complementary satellite methods (Mémin et al., 2015921; The IMBIE Team, 2018).

Mass balance (Gt yr-1) of the West Antarctic Ice Sheet (WAIS), Antarctic Peninsula (AP), East Antarctic Ice Sheet (EAIS), the combined Antarctic Ice Sheets (AIS) and the Greenland Ice Sheet (GIS) and the total sea level contribution (mm yr-1). 

AIS mass changes are dominated by changes in snowfall and glacier flow. The WAIS and AP loss trends in recent decades are dominated by glacier flow acceleration (also known as dynamic thinning) (very high confidence) (Figure SM3.8). Dynamic thinning losses were –112 ± 12 Gt yr–1 for 2003–2013, largely from the ASE (Figure SM3.8) (Martín‐Español et al., 2016951), which contributed –102 ± 10 Gt yr–1 from 2003 to 2011 (Sutterley et al., 2014952). Total ASE ice discharge increased by 77% since the 1970s (Mouginot et al., 2014953), primarily from acceleration of Pine Island Glacier that began around 1945, Smith, Pope and Kohler glaciers around 1980, and Thwaites Glacier around 2000 (Mouginot et al., 2014954; Konrad et al., 2017955; Smith et al., 2017c956). Dynamic thinning in the ASE and western AP accounted for 88% of the –36 ± 15 Gt yr–1 increase in AIS mass loss from 2008 to 2015 (Gardner et al., 2018957). Glacier acceleration of up to 25% also affected the Getz Ice Shelf margin from 2007 to 2014 (Chuter et al., 2017958).

Reduction or loss of ice shelf buttressing has dominated AIS dynamic thinning (high confidence). Ice shelves buttress 90% of AIS outflow (Depoorter et al., 2013959; Rignot et al., 2014960; Fürst et al., 2016961; Reese et al., 2018962), and ice shelf thinning increased in WAIS by 70% in the decade to 2012, averaged 8% thickness loss from 1994 to 2012 in the ASE (Paolo et al., 2015963), and explains the post-2009 onset of rapid dynamic thinning on the southern-AP Bellingshausen Sea coast (Wouters et al., 2015964; Hogg et al., 2017965; Martin-Español et al., 2017966) (Figure SM3.8). Grounding line retreat, an indicator of thinning, has been observed with high confidence (Rignot et al., 2014967; Christie et al., 2016968; Hogg et al., 2017969; Konrad et al., 2018970; Roberts et al., 2018971). From 2010 to 2016, 22%, 3% and 10% of grounding lines in WAIS, EAIS and the AP respectively retreated at rates faster than 25 m yr−1 (the average pace since the Last Glacial Maximum; Konrad et al., 2018), with highest rates along the Amundsen and Bellingshausen Sea coasts, and around Totten Glacier, Wilkes Land, EAIS (Konrad et al., 2018), where dynamic thinning has occurred at least since 1979 (Roberts et al., 2018; Rignot et al., 2019972). Ice shelf collapse has driven dynamic thinning in the northern AP over recent decades (high confidence) (Seehaus et al., 2015973; Wuite et al., 2015974; Friedl et al., 2018975; Rott et al., 2018976).

ASE ice shelf basal melting, grounding line retreat and dynamic thinning have varied with ocean forcing (medium confidence) (Dutrieux et al., 2014977; Paolo et al., 2015978; Christianson et al., 2016979; Jenkins et al., 2018980) but this variability is superimposed on sustained mass losses compatible with the onset of Marine Ice Sheet Instability (MISI) for several major glaciers (medium confidence) (Favier et al., 2014981; Joughin et al., 2014982; Mouginot et al., 2014983; Rignot et al., 2014984; Christianson et al., 2016985). Whether unstable WAIS retreat has begun or is imminent remains a critical uncertainty (Cross-Chapter Box 8 in Chapter 3).

Mass gains due to increased snowfall have somewhat offset dynamic-thinning losses (high confidence). On the AP, snowfall began to increase in the 1930s, accelerated in the 1990s (Thomas et al., 2015986; Goodwin et al., 2016987), and now offsets sea-level rise by 6.2 ± 1.7 mm per century (Medley and Thomas, 2018988). EAIS and WAIS snowfall increases offset 20th century sea-level rise by 7.7 ± 4.0 mm and 2.8 ± 1.7 mm respectively (Medley and Thomas, 2018989) (medium confidence). AIS snowfall increased by 4 ± 1 then 14 ± 1 Gt per decade over the 19th and 20th centuries, of which EAIS contributed 10% (Thomas et al., 2017b990). Longer records suggest either an AIS snowfall decrease over the last 1000 years (Thomas et al., 2017a991) or a statistically negligible change over the last 800 years (low confidence) (Frezzotti et al., 2013992).

Mass balance contributions from ice sheet basal melting were not described in AR5 (IPCC, 2013993) and the sensitivity of the AIS subglacial hydrological system to climate change is poorly understood. Around half of the AIS bed melts (Siegert et al., 2017994), producing ~65 Gt yr–1 of water (Pattyn, 2010995) (low confidence), some of which refreezes (Bell, 2008996) and some accumulates in subglacial lakes with a total volume of 10s of 1000s of km3 (Popov and Masolov, 2007997; Lipenkov et al., 2016998; Siegert, 2017999). This system contributes fresh water and nutrients to the ocean (Section 3.3.3.3) (Fricker et al., 20071000; Siegert et al., 20071001; Carter and Fricker, 20121002; Horgan et al., 20131003; Le Brocq, 20131004; Flament et al., 20141005; Siegert et al., 20161006), and lubricates glacier sliding (e.g., Dow et al., 2018b). Changes in the ice sheet thickness can redistribute subglacial water, affecting drainage pathways and ice flow (Fricker et al., 20161007), but hydrological observations are very scarce.

The Greenland Ice Sheet (GIS) experienced a marked shift to strongly negative mass balance between the early 1990s and mid–2000s (very high confidence) (Shepherd et al., 20121008; Schrama et al., 20141009; Velicogna et al., 20141010; van den Broeke et al., 20161011; Bamber et al., 20181012; King et al., 20181013; Sandberg Sørensen et al., 20181014; WCRP, 20181015). It is extremely likely that the 2002–2011 and 2012–2016 ice losses were greater than in the 1992–2001 period (Bamber et al., 20181016) (Table 3.3, Figure 3.7, SM3.3.1.3). GIS mass balance is characterised by large interannual variability (e.g., van den Broeke et al., 2017) but from 2005 to 2016, GIS was the largest terrestrial contributor to global sea level rise (WCRP, 20181017). 

A geodetic reconstruction of past ice sheet elevations indicates a GIS mass change of –75.1 ± 29.4 Gt yr–1 from 1900 to 1983, –73.8 ± 40.5 Gt yr–1 from 1983 to 2003, and –186.4 ± 18.9 Gt yr–1 from 2003 to 2010, with the losses consistently concentrated along the northwest and southeast coasts, and more locally in the southwest and on the large west coast Jakobshavn Glacier, though intensifying and spreading to the remainder of the coastal ice sheet in the latest period (Kjeldsen et al., 20151018). Palaeo evidence also suggests that the GIS has contributed substantially to sea level rise during past warm intervals (Cross-Chapter Box 8 in Chapter 3).

Ongoing GIS mass loss over recent years has resulted from a combined increase in dynamic thinning and a decrease in SMB. Of these, reduced SMB due to an increase in surface melting and runoff recently came to dominate (high confidence) (Andersen et al., 20151019; Fettweis et al., 20171020; van den Broeke et al., 20171021; King et al., 20181022), accounting for 42% of losses for 2000–2005, 64% for 2005–2009 and 68% for 2009–2012 (Enderlin et al., 20141023) (Figure 3.7). 

The GIS was close to balance in the early years of the 1990s (Hanna et al., 20131024; Khan et al., 20151025), the interior above 2000 m altitude gained mass from 1961 to 1990 (Colgan et al., 20151026) and both coastal and ice sheet sites experienced an increasing precipitation trend from 1890 to 2012 and 1890 to 2000 respectively (Mernild et al., 20151027), but since the early 1990s multiple observations and modelling studies show strong warming and an increase in runoff (very high confidence). High-altitude GIS sites NEEM and Summit warmed by, respectively, 2.7°C ± 0.33°C over the past 30 years (Orsi et al., 20171028) and by 2.7°C ± 0.3°C from 1982 to 2011 (McGrath et al., 20131029), while satellite thermometry showed statistically significant widespread surface warming over northern GIS from 2000 to 2012 (Hall et al., 20131030). The post-1990s period experienced the warmest GIS near-surface summer air temperatures of 1840–2010 (+1.1˚C) (statistically highly significant) (Box, 2013), and ice core analysis found the 2000–2010 decade to be the warmest for around 2000 years (Vinther et al., 20091031; Masson-Delmotte et al., 20121032), and possibly around 7000 years (Lecavalier et al., 20171033). This significant summer warming since the early 1990s increased GIS melt event duration (Mernild et al., 20171034) and intensity to levels exceptional over at least 350 years (Trusel et al., 20181035), and melt frequency to levels unprecedented for at least 470 years (Graeter et al., 20181036). GIS melt intensity for 1994–2013 was two to fivefold the pre-industrial intensity (medium confidence) (Trusel et al., 20181037). In response, GIS meltwater production and runoff increased (Hanna et al., 20121038; Box, 2013; Fettweis et al., 20131039; Tedstone et al., 20151040; van den Broeke et al., 20161041; Fettweis et al., 20171042), resulting in 1994–2013 runoff being 33% higher the 20th century mean and 50% higher than the 18th century (Trusel et al., 20181043), and 80% higher in a western-GIS marginal river catchment in 2003–2014 relative to 1976–2002 (Ahlstrom et al., 20171044).

Only around half of the 1960–2014 surface melt ran off, most of the rest being retained in firn and snow (Steger et al., 20171045), particularly in recently observed firn aquifers in south and west Greenland (Humphrey et al., 20121046; Forster et al., 20131047; Munneke et al., 20141048; Poinar et al., 20171049) that cover up to 5% of GIS (Miège et al., 20161050; Steger et al., 20171051) and stored around one fifth of the meltwater increase since the late 1990s (Noël et al., 2017) (medium confidence). While potential aquifer storage is equivalent to about a quarter of annual GIS melt production (Koenig et al., 20141053; van den Broeke et al., 20161054) and aquifers have spread to higher altitudes (Steger et al., 20171055), their potential to buffer runoff has been reduced by firn densification (Polashenski et al., 20141056), diversion of water to the bed via crevasses (Poinar et al., 20171058), and the formation of ice layers that prevent drainage and promote surface ponding on the firn (Charalampidis et al., 2016) (high confidence). Such ponding lowers the firn albedo, promoting further melting (high confidence) (e.g., Charalampidis et al., 2015), but the extent of bare ice is a fivefold stronger control on melt (Ryan et al., 20191059). Bare ice produced ~78% of runoff from 1960 to 2014, and its extent is expected to increase non-linearly as snow cover retreats to higher, flatter areas of ice sheet (Steger et al., 20171060). This extent is not well reproduced in climate models, however, with biases of –6% to 13% (Ryan et al., 20191061).

The remaining ~40% of non-SMB GIS mass loss from 1991 to 2015 has resulted from increased ice discharge due to dynamic thinning (high confidence) (Enderlin et al., 20141062; van den Broeke et al., 20161063; King et al., 20181064) (Figure 3.7). From 2000 to 2016, dynamic thinning of 89% of GIS outlet glaciers accounted for –682 ± 31 Gt mass change, of which 92% came from the northwest and southeast GIS (King et al., 20181065). Half came from only four glaciers (Jakobshavn Isbræ, Kangerdlugssuaq, Koge Bugt, and Ikertivaq South) (Enderlin et al., 20141066). Glacier thinning has decreased glacier discharge, however, reducing the dynamic contribution to GIS mass loss (e.g., from 58% from 2000 to 2005 to 32% between 2009 and 2012; Enderlin et al., 2014). Furthermore, there is now high confidence that for most of the GIS, increased surface melt has not led to sustained increases in glacier flux on annual timescales because subglacial drainage networks have evolved to drain away the additional water inputs (e.g., Sole et al., 2013; Tedstone et al., 20151068; Stevens et al., 20161069; Nienow et al., 20171070; King et al., 2018).

There is medium agreement but limited evidence of anthropogenic forcing of AIS mass balance through both SMB and glacier dynamics (low confidence). Partitioning between natural and human drivers of atmospheric and ocean circulation changes remains very uncertain. Partitioning is challenging because, along with the effects of greenhouse gas increases and stratospheric ozone depletion (Waugh et al., 20151161; England et al., 20161162; Li et al., 2016a1163), atmospheric and ocean variability in the areas of greatest AIS mass change are affected by a complex chain of processes (e.g., Fyke et al., 2018; Zhang et al., 2018a1164) that exhibit considerable natural variability and have multiple interacting links to sea surface conditions in the Pacific (Schneider et al., 20151165; England et al., 20161166; Raphael et al., 20161167; Clem et al., 20171168; Steig et al., 20171169; Paolo et al., 20181170) and Atlantic (Li et al., 20141171), with additional local feedbacks (e.g., Stammerjohn et al., 2012; Goosse and Zunz, 20141172). Recent AP warming and consequent ice shelf collapses have evidence of a link to anthropogenic ozone and greenhouse gas forcing via the SAM (e.g., Marshall, 2004; Shindell, 20041173; Arblaster and Meehl, 20061174; Marshall et al., 20061175; Abram et al., 20141176) and to anthropogenic Atlantic sea surface warming via the Atlantic Multidecadal Oscillation (e.g., Li et al., 2014). This warming was highly unusual over the last 1000 years but not unprecedented, and along with subsequent cooling is within the bounds of the large natural decadal-scale climate variability in this region (Mulvaney et al., 20121177; Turner et al., 20161178). More broadly over the AP and coastal WAIS where dynamic mass losses are concentrated, natural variability in atmospheric and ocean forcing appear to dominate observed mass balance (medium confidence) (Smith and Polvani, 20171179; Jenkins et al., 20181180).

Evidence exists for an anthropogenic role in the atmospheric circulation (NAO) changes that have driven GIS mass loss (Section 3.3.1.5.2) (medium confidence), although this awaits formal attribution testing (e.g., Easterling et al., 2016). Arctic amplification of anthropogenic warming (e.g., Serreze et al., 2009) affects atmospheric circulation (Francis and Vavrus, 20151181; Mann et al., 20171182) and has reduced sea ice extent (Section 3.2.1.1.1), feeding back to exacerbate both warming and NAO changes (Screen and Simmonds, 20101183) that impact GIS mass balance. Negative-NAO wind patterns increased GIS melt observed in a 40-year runoff signal (Ahlstrom et al., 20171184), and an increase in melting beginning in the mid-1800s closely followed the onset of industrial era Arctic warming and emerged beyond the range of natural variability in the last few decades (Graeter et al., 20181185; Trusel et al., 20181186) (Section 3.3.1.4).

Section 4.2 assesses the sea level impacts from observed and projected changes in ice sheets.

Chapter 3 assesses changes in polar glaciers in the Canadian and Russian Arctic, Svalbard, Greenland and Antarctica, independent of the Greenland and Antarctic ice sheets (Figure 3.8). Glaciers in all other regions including Alaska, Scandinavia and Iceland are assessed in Chapter 2. 

Changes in the mass of Arctic glaciers for the ‘present day’ (2006–2015) are assessed using a combination of satellite observations and direct measurements (Figure 3.8; Appendix 2.A, Table 1). During this period, glacier mass loss was largest in the periphery of Greenland (–47 ± 16 Gt yr–1), followed by Arctic Canada North (-39 ± 8 Gt yr–1), Arctic Canada South (–33 ± 9 Gt yr–1), the Russian Arctic (–15 ± 12 Gt yr–1) and Svalbard and Jan Mayen (–9 ± 5 Gt yr–1). When combined with the Arctic regions covered in Chapter 2 (Alaska, the Yukon territory of Canada, Iceland and Scandinavia), Arctic glaciers as a whole lost mass at a rate of –213 ± 29 Gt yr–1, a sea level contribution of 0.59 ± 0.08 mm yr–1 (high confidence). Overall during this period, Arctic glaciers caused a similar amount of sea level rise to the GIS (Section 3.3.1.3), but their rate of mass loss per unit area was larger (Bolch et al., 20131187). 

There is limited evidence (high agreement) that the current rate of glacier mass loss is larger than at any time during the past 4000 years (Fisher et al., 20121188; Zdanowicz et al., 20121189). Further back in time during the early to mid- Holocene, pre-historic glacial deposits, ice core records, and numerical modelling evidence shows that many Arctic glaciers were at various stages similar to or smaller than present (Gilbert et al., 20171190; Zekollari et al., 20171191), experienced greater melt rates (Lecavalier et al., 20171192), or may have disappeared altogether (Solomina et al., 20151193) (medium confidence). This evidence, however, does not provide a complete assessment of the rates and magnitudes of past glacier mass loss.

Atmospheric circulation changes (Box et al., 20181194) have led to pan-Arctic variability in glacier mass balance (high confidence), including different rates of retreat between eastern and western glaciers in Greenland’s periphery (Bjørk et al., 20181195), and a high rate of surface melt in the Canadian Arctic (Gardner et al., 20131196; Van Wychen et al., 2016; Millan et al., 20171197) through persistently high summer air temperatures (Bezeau et al., 20141198; McLeod and Mote, 20161199). Atmospheric circulation anomalies from 2007 to 2012 associated with glacier mass loss are also linked to enhanced GIS melt (Section 3.3.1.4) and Arctic sea ice loss (Section 3.2.1.1), and exceed by a factor of two the interannual variability in daily mean pressure (sea level and 500 hPa) of the Arctic region over the 1871–2014 period (Belleflamme et al., 20151203) (Section 3.3.1.6).

Increased surface melt on Arctic glaciers has led to a positive feedback from lowered surface albedo, causing further melt (Box et al., 20121204), and in Svalbard, mean glacier albedo has reduced between 1979 and 2015 (Möller and Möller, 20171205). Across the Arctic, increased surface melt and subsequent ice-layer formation via refreezing within snow and firn also reduces the ability of glaciers to store meltwater, increasing runoff (Zdanowicz et al., 20121206; Gascon et al., 2013a1207; Gascon et al., 2013b1208; Noël et al., 20171209; Noël et al., 20181210).

Between the 1990s and 2017, tidewater glaciers have exhibited regional patterns in glacier dynamics; glaciers in Arctic Canada have largely decelerated, while glaciers in Svalbard and the Russian Arctic have accelerated (Van Wychen et al., 2016; Strozzi et al., 2017). Annual retreat rates of tidewater glaciers in Svalbard and the Russian Arctic for 2000–2010, have increased by a factor 2 and 2.5 respectively, between 1992 and 2000 (Carr et al., 2017). Acceleration due to surging (an internal dynamic instability) of a few key glaciers has dominated dynamic ice discharge on time-scales of years to decades (Van Wychen et al., 2014; Dunse et al., 2015). 

The recent acceleration and surge behaviour of polythermal glaciers in Svalbard and the Russian Arctic is caused by destabilisation of the marine termini due to increased surface melt, and changes in basal temperature, lubrication and weakening of subglacial sediments (Dunse et al., 2015; Sevestre et al., 2018; Willis et al., 2018) or terminus thinning and response to warmer ocean temperatures (McMillan et al., 2014a) (low confidence). Iceberg calving rates in Svalbard are linked to ocean temperatures which control rates of submarine melt (Luckman et al., 2015; Vallot et al., 2018) (medium confidence). Rapid disintegration of ice shelves in the Canadian and Russian Arctic continues and has led to acceleration and thinning in tributary-glacier basins (high confidence) (Willis et al., 2015; Copland and Mueller, 2017).

Little information is available on Holocene and historic changes in glaciers in Antarctica (separate from the ice sheet), and on sub-Antarctic islands (Hodgson et al., 2014). Mass changes of glaciers in these regions between 2006 and 2015 (–90 ± 860 Gt yr–1) have low confidence as they are based on a single data compilation with large uncertainties in the Antarctic region (Zemp et al., 2019) (Figure 3.8). Limited evidence with high agreement from individual glaciers suggests that regional variability in glacier mass changes may be linked to changes in the large-scale Southern Hemisphere atmospheric circulation (Section 3.3.1.5.2). On islands adjacent to the AP, glaciers experienced retreat and mass loss during the mid to late 20th century, but since around 2009 there has been a reduction in mass loss rate or a return to slightly positive balance (Navarro et al., 2017; Oliva et al., 2017). Reduced mass loss has been linked to increased winter snow accumulation and decreased summer melt at these locations, associated with recent deepening of the circumpolar pressure trough (Oliva et al., 2017). Conversely, on the sub-Antarctic Kerguelen Islands, increased glacier mass loss (Verfaillie et al., 2015) may be due to reduced snow accumulation rather than increased air temperature as a result of southward migration of storm tracks (Favier et al., 2016). 

Projections of all glaciers, including those in polar regions, are covered in Cross-Chapter Box 6 in Chapter 2.

Chapter 4 assesses the sea level impacts from observed and projected changes in ice sheets (Section 3.3.1) and polar glaciers (Section 3.3.2), including uncertainties related to marine ice sheets (Cross-Chapter Box 8 in Chapter 3).

The major large-scale impacts of freshwater release from Greenland on ocean circulation relate to the potential modulation/inhibition of the formation of water masses that represent the headwaters of the Atlantic Meridional Overturning Circulation. The timescales and likelihood of such effects are assessed separately in Chapter 6 (Section 6.7). Freshwater release also affects local circulation within fjords through two principle mechanisms; subglacial release from tidewater glaciers enhances buoyancy driven circulation, whereas runoff from land-terminating glaciers contributes to surface layer freshening and estuarine circulation (Straneo and Cenedese, 20151229). There is limited evidence that freshening occurred between 2003 and 2015 in North East Greenland fjords and coastal waters (Sejr et al., 20171230).

For Antarctica, freshwater input to the ocean from the ice sheet is divided approximately equally between melting of calved icebergs and of ice shelves in situ (Depoorter et al., 20131231; Rignot et al., 20141232). There is high confidence that the input of ice shelf meltwater has increased in the Amundsen and Bellingshausen Seas since the 1990s, but low confidence in trends in other sectors (Paolo et al., 20151233). 

Freshwater injected from the AIS affect water mass circulation and transformation, though sea ice dominates upper ocean properties away from the Antarctic ice shelves (Abernathey et al., 20161234; Haumann et al., 20161235). Over the ice shelf regions, where dense waters sink and flood the global ocean abyss, the role of glacial freshwater input is clearer. From 1980 to 2012, the salinity of Antarctic Bottom Water reduced by an amount equivalent to 73 ± 26 Gt y–1 of freshwater added, around half the estimated increase in freshwater input by Antarctic glacial discharge up to that time (Purkey and Johnson, 20131236). In some places, notably the Indian-Australian sector, Antarctic Bottom Water freshening may be accelerating (Menezes et al., 20171237). There is medium confidence in an overall freshening trend and low confidence that this is accelerating, given the sparsity of information and significant interannual variability in Antarctic Bottom Water properties at other export locations (Meijers et al., 20161238).

For the Southern Ocean, there is limited evidence for stratification changes in the post-AR5 period, and low confidence in how stratification changes are affecting sea ice and basal ice shelf melt. An increase in stratification caused by release of freshwater from the AIS was invoked as a mechanism to suppress vertical heat flux and permit an increase in sea ice extent (Bintanja et al., 20131239; Bronselaer et al., 20181240; Purich et al., 20181241), though some studies conclude that glacial freshwater input is insufficient to cause a significant sea ice expansion (Swart and Fyfe, 20131242; Pauling et al., 20171243) (Section 3.2.1.1). In contrast, where warm water intrusions drive melting within ice shelf cavities, a significant entrained heat flux to the surface can exist and increase stratification and potentially reduce sea ice extent (Jourdain et al., 20171244; Merino et al., 20181245). It has been argued that freshening from glacial melt can enhance basal melting of ice shelves by reducing dense water production and modulating oceanic heat flow into ice shelf cavities (Silvano et al., 20181246).

For Greenland and Svalbard, there is limited evidence with high agreement that the retreat of marine-terminating glaciers will alter food supply to higher trophic levels of marine food webs (Meire et al., 20171292; Milner et al., 20171293). The consequences of changes in glacial systems on marine ecosystems are often mediated via the fjordic environments that fringe the edge of the ice sheets, for example changing physical-chemical conditions have affected the benthic ecosystems of Arctic fjords (Bourgeois et al., 20161294). The amplification of nutrient fluxes caused by enhanced upwelling at calving fronts (Meire et al., 20171295), combined with high carbon/nutrient burial and recycling rates (Wehrmann et al., 20131296; Smith et al., 20151297), plays an important role in sustaining high productivity of the Arctic fjord ecosystems of Greenland and Svalbard (Lydersen et al., 20141298). Glacier retreat, causing glaciers to shift from being marine-terminating to land-terminating, can reduce the productivity in coastal areas off Greenland with potentially large ecological implications, also negatively affecting production of commercially harvested fish (Meire et al., 20171299). There is also evidence that marine-terminating glaciers are important feeding areas for marine mammals and seabirds at Greenland (Laidre et al., 20161300) and Svalbard (Lydersen et al., 20141301).

For Antarctica, there is high agreement based on medium evidence that ice shelf retreat or collapse is leading to new marine habitats and to biological colonisation (Gutt et al., 20111302; Fillinger et al., 20131303; Trathan et al., 20131304; Hauquier et al., 20161305; Ingels et al., 20181306). The loss of ice shelves and retreat of coastal glaciers around the AP in the last 50 years has exposed at least 2.4 × 104 km2 of new open water. These newly-revealed habitats have allowed new phytoplankton blooms to be produced resulting in new marine zooplankton and seabed communities (Gutt et al., 20111307; Fillinger et al., 20131308; Trathan et al., 20131309; Hauquier et al., 20161310) (Section 3.2.3.2.1), and have resulted in enhanced carbon uptake by coastal marine ecosystems (medium confidence), although quantitative estimates of biological carbon uptake are highly variable (Trathan et al., 20131311; Barnes et al., 20181312). Newly available habitat on coastlines may also provide breeding or haul out sites for land-based predators such as penguins and seals (Trathan et al., 20131313) (low confidence). Fjords that have been studied in the subpolar western AP are hotspots of abundance and biodiversity of benthic macro-organisms (Grange and Smith, 20131314) and there is evidence that glacier retreat in these environments can impact the structure and function of benthic communities (Moon et al., 20151315; Sahade et al., 20151316) (low confidence).

Authors: Robert M. DeConto (USA), Alexey Ekaykin (Russian Federation), Andrew Mackintosh (Australia), Roderik van de Wal (Netherlands), Jeremy Bassis (USA)

Over the last century, glaciers were the main contributors to increasing ocean water mass (Section 4.2.1.2). However, most terrestrial frozen water is stored in Antarctic and Greenland ice sheets, and future changes in their dynamics and mass balance will cause sea level rise over the 21st century and beyond (Section 4.2.3).

About a third of the Antarctic Ice Sheet (AIS) is ‘marine ice sheet’, i.e., rests on bedrock below sea level (Figure 4.5), with most of the ice sheet margin terminating directly in the ocean. These features make the overlying ice sheet vulnerable to dynamical instabilities with the potential to cause rapid ice loss; so-called Marine Ice Sheet and Marine Ice Cliff instabilities, as discussed below.

In many places around the AIS margin, the seaward-flowing ice forms floating ice shelves (Figure CB8.1). Ice shelves in contact with bathymetric features on the sea floor or confined within embayments provide back stress (buttressing) that impedes the seaward flow of the upstream ice and thereby stabilises the ice sheet. The ice shelves are thus a key factor controlling AIS dynamics. Almost all Antarctic ice shelves provide substantial buttressing (Fürst et al., 20161316) but some are currently thinning at an increasing rate (Khazendar et al., 20161317). Today, thinning and retreat of ice shelves is associated primarily with ocean driven basal melt that, in turn, promotes iceberg calving (Section 3.3.1.2).

Accumulation and percolation of surface melt and rain water also impact ice shelves by lowering albedo, deepening surface crevasses, and causing flexural stresses that can lead to hydrofracturing and ice shelf collapse (Macayeal and Sergienko, 20131319). In some cases supraglacial (i.e., flowing on the glacier surface) rivers might diminish destabilising impact of surface melt by removing meltwater before it ponds on the ice shelf surface (Bell et al., 20171320). In summary, both ocean forcing and surface melt affect ice shelf mechanical stability (high confidence), but the precise importance of the different mechanisms remains poorly understood and observed.

The future dynamic response of the AIS to warming will largely be determined by changes in ice shelves, because their thinning or collapse will reduce their buttressing capacity, leading to an acceleration of the grounded ice and to thinning of the ice margin. In turn, this thinning can initiate grounding line retreat (Konrad et al., 20181321). If the grounding line is located on bedrock sloping downwards toward the ice sheet interior (retrograde slope), initial retreat can trigger a positive feedback, due to non-linear response of the seaward ice flow to the grounding line thickness change. As a result, progressively more ice will flow into the ocean (Figure CB8.1a). This self-sustaining process is known as Marine Ice Sheet Instability (MISI). The onset and persistence of MISI is dependent on several factors in addition to overall bed slope, including the details of the bed geometry and conditions, ice shelf pinning points, lateral shear from the walls, self-gravitation effects on local sea level and isostatic adjustment. Hence, long-term retreat on every retrograde sloped bed is not necessarily unstoppable (Gomez et al., 20151322).

The MISI process might be particularly important in West Antarctica, where most of the ice sheet is grounded on bedrock below sea level (Figure 4.5). Since AR5, there is growing observational and modelling evidence that accelerated retreat may be underway in several major Amundsen Sea outlets, including Thwaites, Pine Island, Smith, and Kohler glaciers (e.g., Rignot et al., 2014) supporting the MISI hypothesis, although observed grounding line retreat on retrograde slope is not definitive proof that MISI is underway.

It has been shown recently (Barletta et al., 20181323) that the Amundsen Sea Embayment (ASE) experiences unexpectedly fast bedrock uplift (up to 41 mm yr-1, due to mantle viscosity much lower than the global average) as an adjustment to reduced ice mass loading, which could help stabilise grounding line retreat. 

One of the largest outlets of the East Antarctic Ice Sheet (EAIS), Totten glacier, has also been retreating and thinning in recent decades (Li et al., 2015b1324). Totten’s current behaviour suggests that East Antarctica could become a substantial contributor to future sea level rise, as it has been in the previous warm periods (Aitken et al., 20161325). It is not clear, however, if the changes observed recently are a linear response to increased ocean forcing (Section 3.3.1.2), or an indication that MISI has commenced (Roberts et al., 20181326).

The disappearance of ice shelves may allow the formation of ice cliffs, which may be inherently unstable if they are tall enough (subaerial cliff height between 100 and 285 m) to generate stresses that exceed the strength of the ice (Parizek et al., 20191327). This ice cliff failure can lead to ice sheet retreat via a process called marine ice cliff instability (MICI; Figure CB8.1b), that has been hypothesised to cause partial collapse of the West Antarctic Ice Sheet (WAIS) within a few centuries (Pollard et al., 20151328; DeConto and Pollard, 20161329).

Limited evidence is available to confirm the importance of MICI. In Antarctica, marine-terminating ice margins with the grounding lines thick enough to produce unstable ice cliffs are currently buttressed by ice shelves, with a possible exception of Crane glacier on the Antarctic Peninsula (Section 4.2.3.1.2). 

Overall, there is low agreement on the exact MICI mechanism and limited evidence of its occurrence in the present or the past. Thus the potential of MICI to impact the future sea level remains very uncertain (Edwards et al., 20191330).

Limited evidence from geological records and ice sheet modelling suggests that parts of AIS experienced rapid (i.e., on centennial time-scale) retreat likely due to ice sheet instability processes between 20,000 and 9,000 years ago (Golledge et al., 20141331; Weber et al., 20141332; Small et al., 20191333). Both the WAIS (including Pine Island glacier) and EAIS also experienced rapid thinning and grounding line retreat during the early to mid-Holocene (Jones et al., 2015b1334; Wise et al., 20171335). In the Ross Sea, grounding lines may have retreated several hundred kilometers inland and then re-advanced to their present-day positions due to bedrock uplift after ice mass removal (Kingslake et al., 20181336), thus supporting the stabilising role of glacial isostatic adjustment on ice sheets (Barletta et al., 20181337). These past rapid changes have likely been driven by the incursion of Circumpolar Deep Water onto the Antarctic continental shelf (Section 3.3.1.5.1) (Golledge et al., 20141338; Hillenbrand et al., 20171339) and MISI (Jones et al., 2015b1340). Limited evidence of past MICI in Antarctica is provided by deep iceberg plough marks on the sea-floor (Wise et al., 20171341).

The ability of models to simulate the processes controlling MISI has improved since AR5 (Pattyn, 20181342), but significant discrepancies in projections remain (Section 4.2.3.2) due to poor understanding of mechanisms and lack of observational data on bed topography, isostatic rebound rates, etc. to constrain the models. Inclusion of MICI in one ice sheet model has improved its ability to match (albeit uncertain) geological sea level targets in the Pliocene (Pollard et al., 20151343) and Last Interglacial (DeConto and Pollard, 20161344), although the MICI solution may not be unique (Aitken et al., 20161345) (Section 4.2.3.1.2).

The Greenland Ice Sheet (GIS) has limited direct access to the ocean through relatively narrow subglacial troughs (Morlighem et al., 20171346), and most of the bedrock at the ice sheet margin is above sea level (Figure 4.5). However, since AR5 it has been argued that several Greenland outlet glaciers (Petermann, Kangerdlugssuaq, Jakobshavn Isbræ, Helheim, Zachariæ Isstrøm) and North-East Greenland Ice Stream may contribute more than expected to future sea level rise (Mouginot et al., 20151347). It has also been shown that Greenland was nearly ice free for extensive episodic periods during the Pleistocene, suggesting a sensitivity to deglaciation under climates similar to or slightly warmer than present (Schaefer et al., 20161348).

A MICI-style behaviour is seen today in Greenland at the termini of Jakobshavn and Helheim glaciers (Parizek et al., 20191349), but calving of these narrow outlets is controlled by a combination of ductile and brittle processes, which might not be representative examples of much wider Antarctic outlet glaciers, like Thwaites. 

Overall, this assessment finds that unstable retreat and thinning of some Antarctic glaciers, and to a lesser extent Greenland outlet glaciers, may be underway. However, the timescale and future rate of these processes is not well known, casting deep uncertainty on projections of the sea level contributions from the AIS (Cross-Chapter Box 5 in Chapter 1, Section 4.2.3.1). 

Terrestrial snow cover is a defining characteristic of the Arctic land surface for up to nine months each year, with changes influencing the surface energy budget, ground thermal regime and freshwater budget. Snow cover also interacts with vegetation, influences biogeochemical activity and affects habitats and species, with consequences for ecosystem services. Arctic land areas are almost always completely snow covered in winter, so the transition seasons of autumn and spring are key when characterising variability and change.

There is increasing awareness of the influence of a changing climate on freshwater systems across the Arctic, and associated impacts on hydrological, biogeophysical and ecological processes (Prowse et al., 20151456; Walvoord and Kurylyk, 20161457), and northern populations (Takakura, 20181458) (Section 3.4.3.3.1). Assessing these impacts requires consideration of complex interconnected processes, many of which are incompletely observed. The increasing imprint of human development, such as flow regulation on major northerly flowing rivers adds complexity to the determination of climate-driven changes.

Historical simulations from CMIP5 models tend to underestimate observed reductions in spring snow cover extent due to uncertainty in the parameterisation of snow processes (Essery, 20131518; Thackeray et al., 20141519), challenges in simulating snow-albedo feedback (Qu and Hall, 20141520; Fletcher et al., 20151521; Li et al., 2016b1522), unrealistic temperature sensitivity (Brutel-Vuilmet et al., 20131523; Mudryk et al., 20171524), and biases in climatological spring snow cover (Thackeray et al., 20161525). The role of precipitation biases is not well understood (Thackeray et al., 20161526).

Reductions in Arctic snow cover duration are projected by the CMIP5 multi-model ensemble due to later snow onset in the autumn and earlier snow melt in spring (Brown et al., 20171527) driven by increased surface temperature over essentially all Arctic land areas (Hartmann et al., 2013). There is high confidence that projected snow cover declines are proportional to the amount of future warming in each model realisation (Thackeray et al., 20161528; Mudryk et al., 20171529). Projections to mid-century are primarily dependent on natural variability and model dependent uncertainties rather than the choice of forcing scenario (Hodson et al., 20131530). By end of century, however, differences between scenarios emerge. Under RCP2.6 and RCP4.5, Arctic snow cover duration stabilises at 5–10% reduction (compared to a 1986–2005 reference period); under RCP8.5, snow cover duration declines reach –15 to –25% (Brown et al., 20171531) (Figure 3.10) (high confidence).

Positive Arctic snow water equivalent changes emerge across the eastern Eurasian Arctic by mid-century for both RCP4.5 and RCP8.5 (Brown et al., 20171532) (medium confidence). Projected snow water equivalent increases across the North American Arctic are only modest, emerge later in the century, and only under RCP8.5 (Brown et al., 20171533). These projected increases are due to enhanced snowfall (Krasting et al., 20131534) from a more moisture-rich Arctic atmosphere coupled with winter season temperatures that remain sufficiently low for precipitation to fall as snow. There is low confidence in changes to snow properties such as density and stratigraphy (relevant for understanding the impacts of changes to Arctic snow on ecosystems) which are not resolved directly by climate model simulations, but require detailed snow physics models.

Circumpolar- or global-scale models represent permafrost degradation in response to warming scenarios as increases in thaw depth only. The CMIP5 models project with high confidence that thaw depth will increase and areal extent of near-surface permafrost will decrease substantially (Koven et al., 20131535; Slater and Lawrence, 20131536) (Figure 3.10). However, there is only medium confidence in the magnitude of these changes due to at least a five-fold range of estimated present day near-surface permafrost area (<5 – >25 x 106 km2) by these models. This was caused by a wide range of model sensitivity in permafrost area to air temperature change, resulting in a large range of projected near-surface permafrost loss by 2100: 2–66% for RCP2.6 (24 ± 16%; likely range), 15–87% under RCP4.5 and 30–99% (69 ± 20%; likely range) under RCP8.5. A more recent analysis of near-surface permafrost trends from a subset of models that self-identified as structurally representing the permafrost region had a significantly smaller range of estimated present day near-surface permafrost area (13.1–19.3 x 106 km2; mean ± SD, 14.1 ± 3.5 x 106 km2) (McGuire et al., 20181537). This subset of models also showed large reductions of near-surface permafrost area, averaging a 90% loss (12.7 ± 5.1×106 km2) of permafrost area by 2300 for RCP8.5 and 29% loss (4.1 ± 0.6×106 km2) for RCP4.5, with much of that long-term loss already occurring by 2100.

Pulse disturbances are not included in the permafrost projections described above, and there is high confidence that fire and abrupt thaw will accelerate change in permafrost relative to climate effects alone, if the rates of these disturbances increase. The observed trend of increasing fire is projected to continue for the rest of the century across most of the tundra and boreal region for many climate scenarios, with the boreal region projected to have the greatest increase in total area burned (Balshi et al., 20091538; Kloster et al., 20121539; Wotton et al., 20171540). Due to vegetation-climate interactions, there is only medium confidence in projections of future area burned. As fire activity increases, flammable vegetation, such as the black spruce forest that dominates boreal Alaska, is projected to decline as it is replaced by low-flammability deciduous forest (Johnstone et al., 20111541; Pastick et al., 20171542). In other regions such as western Canada, by contrast, black spruce could be replaced by the even more flammable jack pine, creating regional-scale feedbacks that increase the spread of fire on the landscape (Héon et al., 20141543). A regional process-model study of Alaska projected annual median area burned during the 21st century to be 1.3-1.7 times higher compared with the historical average  (Pastick et al., 20171544). Fire also appears to be expanding as a novel disturbance into tundra and forest-tundra boundary regions previously protected by a cool, moist climate (Jones et al., 20091545; Hu et al., 20101546; Hu et al., 20151547) (medium confidence). Annual tundra area burned in Alaska is projected to double under RCP6.0 from a historic rate of 270 km2 yr-1 to 500–610 km2 yr-1 over the 21st century (Hu et al., 20151548). A statistical approach projected a fourfold increase in the 30-year probability of fire occurrence in the forest-tundra boundary by 2100 (Young et al., 20171549). In contrast to fire, there has not yet been a comprehensive circumpolar projection of how abrupt thaw rates may change in the future, but one component of abrupt thaw, change in abrupt thaw lake area, has been projected to increase to increase by 53% under RCP8.5 (Walter Anthony et al., 20181550) above the 1.4 x 106 km2 of small lakes and ponds that currently exist in the permafrost region (Muster et al., 20171551). As a result, there is low confidence in the ability to assess the magnitude by which abrupt thaw across the entire landscape will affect regional permafrost, even though this mechanism for rapid change appears critically important for projecting future change (Kokelj et al., 20171552). 

Climate model simulations project a warmer and wetter Arctic (Krasting et al., 20131553), with increased specific humidity due to enhanced evaporation (Laîné et al., 20141554), and moisture flux convergence increases into the Arctic (Skific and Francis, 20131555). Increased cold-season precipitation is projected across the Arctic by CMIP5 models (Lique et al., 20161556) due to increased moisture flux convergence from outside the Arctic (Zhang et al., 20121557) and enhanced moisture availability from reduced sea ice cover (Bintanja and Selten, 20141558) (high confidence). Increases in precipitation extremes are also projected over northern watersheds (Kharin et al., 20131559; Sillmann et al., 20131560), while rain on snow events are expected to increase (Hansen et al., 20141561). A net increased ratio of precipitation minus evaporation is projected, resulting in increased freshwater flux from the land surface to the Arctic Ocean, projected to be 30% above current values by 2100 under RCP4.5 (Haine et al., 20151562) (Figure 3.10). This is consistent with CMIP5 model projections of increased discharge from Arctic watersheds (van Vliet et al., 20131563; Gelfan et al., 20161564; MacDonald et al., 20181565). The water temperature of this increased discharge is projected to be approximately 1°C warmer than current conditions, increasing the heat flux to Arctic Ocean (van Vliet et al., 20131566).

Lake ice phenology is sensitive to projected changes in surface temperature (Sharma et al., 20191567). Lake ice models project an earlier spring break-up of between 10–25 days by mid-century (compared with 1961–1990), and up to a 15-day delay in the freeze-up for lakes in the North American Arctic, with more extreme reductions for coastal regions (Brown and Duguay, 20111568; Dibike et al., 20111569; Prowse et al., 20111570) (medium confidence). Mean maximum ice thickness is projected to decrease by 10–50 cm over the same period (Brown and Duguay, 20111571). High-latitude warming is projected to drive earlier river ice break-up in spring due to both decreasing ice strength, and earlier onset of peak discharge (Cooley and Pavelsky, 20161572). Complex interplay between hydrology and hydraulics in controlling spring flooding and ice jam events complicate projections of these events (Prowse et al., 20101573; Prowse et al., 20111574).

In polar regions climate-induced changes in terrestrial, ocean and sea ice environments, together with human introduction of non-native species, have expanded the range of some temperate species and contracted the range of some polar fish and ice-associated species (Section 3.2.3.2; Duffy et al., 20171752) (high confidence for detection, medium confidence for attribution). In some cases, spatial shifts in distribution have also been influenced by fluctuations in population abundance linked to climate-induced impacts on reproductive success (Section 3.2.3). These changes have the potential to alter biodiversity in polar marine and terrestrial ecosystems (Frenot et al., 20051753; Frederiksen, 20171754; McCarthy et al., 20191755) (medium confidence).

Ongoing climate change induced reductions in suitable habitat for Arctic sea ice affiliated endemic marine mammals is an escalating threat (Section 3.2.3.1) (high confidence). This is further complicated by the northward expansion of the summer ranges of a variety of temperate whale species, documented recently in both the Pacific and Atlantic sides of the Arctic (Brower et al., 20171756; Storrie et al., 20181757) and increasing pressure from anthropogenic activities. Also, over the recent decade a northward shift in benthic species, with subsequent changes in community composition has been detected in both the northern Bering Sea (Grebmeier, 20121758), off Western Greenland (Renaud et al., 20151759) and the Barents Sea (Kortsch et al., 20121760) (medium confidence). At the same time as these northward expansions or shifts, a number of populations of species as different as polar bear and Arctic char show range contraction or population declines (Winfield et al., 20101761; Bromaghin et al., 20151762; Laidre et al., 20181763).

In the Arctic a number of fish species have changed their spatial distribution substantially over the recent decades (high confidence). The most pronounced recent range expansion into the Arctic may be that of the summer feeding distribution of the temperate Atlantic mackerel (Scomber scombrus) in the Nordic Seas. From 1997 to 2016 the total area occupied by this large stock expanded from 0.4 to 2.5 million km2 and the centre-of-gravity of distribution shifted westward by 1650 km and northward by 400 km (Olafsdottir et al., 2019), far into Icelandic and Greenland waters and even up to Svalbard (Berge et al., 20151764; Jansen et al., 20161765; Nøttestad et al., 20161766). This range expansion was linked both to a pronounced increase in stock size and warming of the ocean (Berge et al., 20151767; Olafsdottir et al., 20191768) (high confidence). Under RCP4.5 and RCP8.5 further range expansions of mackerel are projected in Greenland waters (Jansen et al., 20161769) (medium confidence). However, further northwards expansion of planktivorous species may generally be restricted by them not being adapted to lack of primary production during winter (Sundby et al., 20161770). Range shifts have also been observed in the Bering Sea since 1993, with warm bottom temperatures being associated with range contractions of Arctic species, and range expansions of sub-arctic species, with responses dependent on species specific vulnerably (Alabia et al., 20181771; Stevenson and Lauth, 20181772).

In the Barents Sea, major expansions in distribution over the recent years to decades have been well documented for both individual species and whole biological communities (high confidence). New information strengthens findings reported in WGII AR5 of ecologically and commercially important fish stocks having extended their habitats markedly to the north and east, concomitant to increased sea temperature and retreating sea ice. This includes capelin (Ingvaldsen and Gjøsæter, 20131773), Atlantic cod (Kjesbu et al., 20141774) and haddock (Landa et al., 20141775). Of even greater importance is novel evidence of distinct distributional changes at the community level (Fossheim et al., 20151776; Kortsch et al., 20151777; Frainer et al., 20171778) (Box 3.4 Figure 1). Until recently, the northern Barents Sea was dominated by small-sized, slow-growing fish species with specialised diets, mostly living in close association with the sea floor. Simultaneous with rising sea temperatures and retreating sea ice, these Arctic fishes are being replaced by boreal, fast-growing, large-bodied generalist fish moving in from the south. These large, migratory predators take advantage of increased production while the Arctic fish species suffer from higher competition and predation and are retracting northwards and eastwards. Consequently, climate change is inducing structural change over large spatial scales, leading to a borealisation (‘Atlantification’) of the European Arctic biological communities (Fossheim et al., 20151779; Kortsch et al., 20151780; Frainer et al., 20171781) (medium confidence). 

There is evidence based on population genetics that the ecosystem off Northeast Greenland could also become populated by a larger proportion of boreal species with ocean warming. Andrews et al. (2019)1782 show that Atlantic cod, beaked redfish (Sebastes mentella), and deep-sea shrimp (Pandalus borealis) recently found on the Northeast Greenland shelf originate from the quite distant Barents Sea, and suggested that pelagic offspring were dispersed via advection across the Fram Strait.

Physical barriers to range expansions into the high Arctic interior shelf systems and the outflow systems of Eurasia and the Canadian Archipelago will continue to govern future expansions of fish populations (medium confidence). The limited available information on marine fish from other Arctic shelf regions reveals a latitudinal cline in the abundance of commercially harvestable fish species. For instance, there is evidence of latitudinal partitioning between the four dominant mid-water species (Polar cod, saffron cod (Eleginus gracilis), capelin, and Pacific herring (Clupea pallasii)) in the Chuckchi and Northern Bering Sea, with Polar cod being most abundant to the north (De Robertis et al., 2017). These latitudinal gradients suggest that future range expansions of fish populations will continue to be governed by a combination of physical factors affecting overwintering success and the availability, quality and quantity of prey (medium confidence).

In Antarctic marine systems, there is evidence of recent climate-related range shifts in the southwest Atlantic and West Antarctic Peninsula for penguin species (Pygoscelis papua and P. antarctica) and for Antarctic krill (Euphausia superba), but mesozooplankton communities do not appear to have changed or shifted in response to ocean warming (Section 3.2.3.2). Recent evidence suggests that the ACC and its associated fronts and thermal gradients may be more permeable to biological dispersal than previously thought, with storm-forced surface waves and ocean eddies enhancing oceanographic connectivity for drift particles in surface layers of the Southern Ocean (Fraser et al., 20171783; Fraser et al., 20181784) (low confidence), but it is unclear whether this will be an increasingly important pathway under climate change. Greater ship activity in the Southern Ocean may also present a risk for increasing introduction of non-native marine species, with the potential for these species to become invasive with changing environmental conditions (McCarthy et al., 20191785). Current evidence of invasions by shell-crushing crabs on the Antarctic continental slope and shelf remains equivocal (Griffiths et al., 20131786; Aronson et al., 20151787; Smith et al., 2017d1788).

On Arctic land, northward range expansions have been recorded in species from all major taxon groups based both on scientific studies and local observations (high confidence) (CAFF, 2013a1789; AMAP, 2017a1790; AMAP, 2017b1791; AMAP, 20181792). The most recent examples of terrestrial vertebrates expanding northwards include a whole range of mammals in Yakutia, Russia (Safronov, 20161793), moose (Alces alces) into the Arctic region of both northern continents (Tape et al., 20161795) and North American beaver (Castor canadensis) in Alaska (Tape et al., 20181794). In parallel with these expansions, pathogens and pests are also spreading north (CAFF, 2013a1796; Taylor et al., 20151797; Forde et al., 2016b1798; Burke et al., 2017b1799; Kafle et al., 20181800). A widespread change is tundra greening, which in some cases is linked to shifting plant dominance within Arctic plant communities, in particular an increase in woody shrub biomass as conditions become more favorable for them (Myers-Smith et al., 20151801; Bhatt et al., 20171802).

Expansion of subarctic terrestrial species and biological communities into the Arctic and displacing native species is considered a major threat, since unique Arctic species may be less competitive than encroaching subarctic species favoured by changing climatic conditions (CAFF, 2013a1803). Similar displacements may take place within zones of the Arctic when low- and mid-Arctic species expand northward. Here, the most vulnerable species and communities may be in the species-poor, but unique, northernmost sub-zone of the Arctic because species cannot migrate northward as southern species encroach (CAVM Team, 20031804; Walker et al., 20161805; AMAP, 20181806). This ‘Arctic squeeze’ is a combined effect of the fact that the area of the globe increasingly shrinks when moving poleward and that there is nowhere further north on land to go for terrestrial biota at the northern coast. The expected overall result of these shifts and limits will be a loss of biodiversity (CAFF, 2013a1807; CAFF, 2013b1808; AMAP, 20181809) (medium confidence). At the southern limit of the Arctic, thermal hotspots may support high biological productivity, but not necessarily high biodiversity (Walker et al., 20151810) and may even act as advanced bridgeheads for expansion of subarctic species into the true Arctic (medium confidence). At the other end of the Arctic zonal range, a temperature increase of only 1°C –2°C in the northernmost subzone may allow the establishment of woody dwarf shrubs, sedges and other species into bare soil areas that may radically change its appearance and ecological functions (Walker et al., 20151811; Myers‐Smith et al., 20191812) (medium confidence). 

Range expansions also include the threat from alien species brought in by humans to become invasive and outcompete native species. Relatively few invasive alien species are presently well established in the Arctic, but many are thriving in the subarctic and may expand as a result of climate change (CAFF, 2013a1813; CAFF, 2013b[). Examples of this include: American mink (Neovison vison) and Nootka lupin (Lupinus nootkatensis) in Arctic western Eurasia, Greenland and Iceland that are already causing severe problems to native fauna and flora (CAFF and PAME, 2017).

Alien species are a major driver of terrestrial biodiversity change also in the Antarctic region (Frenot et al., 20051814; Chown et al., 20121815; McClelland et al., 20171816). The Protocol on Environmental Protection to the Antarctic Treaty restricts the introduction of non-native species to Antarctica as do the management authorities of sub-Antarctic islands (De Villiers et al., 2006). Despite this, alien species and their propagules continue to be introduced to the Antarctic continent and sub-Antarctic islands (Hughes et al., 20151817). To date, 14 non-native terrestrial species have colonised the Antarctic Treaty area (excluding sub-Antarctic islands) (Hughes et al., 20151818), while the number in the sub-Antarctic is much higher (on the order of 200 species) (Frenot et al., 20051819) (low confidence). Species distribution models for terrestrial invasive species indicate that climate does not currently constitute a barrier for the establishment of invasive species on all subantarctic islands, and that the AP region will be the most vulnerable location on the Antarctic continent to invasive species establishment under RCP8.5 (Duffy et al., 20171820). Thus, for continental Antarctica, existing climatic barriers to alien species establishment will weaken as warming continues across the region (medium confidence). An increase in the ice-free area linked to glacier retreat in Antarctica is expected to increase the area available for new terrestrial ecosystems (Lee et al., 2017a1821). Along with growing number of visitors, this is expected to increase in the establishment probability of terrestrial alien species (Chown et al., 20121822; Hughes et al., 20151823) (medium confidence). 

The Arctic is home to over four million people, with large regional variation in population distribution and demographics (Heleniak, 20141824). ‘Connection with nature’ is a defining feature of Arctic identity for indigenous communities (Schweitzer et al., 2014) because the lands, waters and ice that surround communities evoke a sense of home, freedom and belonging, and are crucial for culture, life and survival (Cunsolo Willox et al., 20121825; Durkalec et al., 20151826). Climate-driven environmental changes are affecting local ecosystems and influencing travel, hunting, fishing and gathering practises. This has implications for people’s livelihoods, cultural practices, economies and self-determination.

Human responses to climate change in the Arctic and Antarctica are shaped by their unique physical, ecological, social, cultural and political conditions. Extreme climatic conditions, remoteness from densely populated regions, limited human mobility, short seasons of biological productivity, high costs in monitoring and research, sovereignty claims to lands and waters by southern-based governments, a rich diversity of indigenous cultures and institutional arrangements that in some cases recognise indigenous rights and support regional and international cooperation in governance are among the many factors that impede and or facilitate adaptation.

The social and cultural differences are an especially noteworthy factor in assessing polar responses. Approximately four million people currently reside in the Arctic with about three quarters residing in urban areas, and approximately 10% being Indigenous (AHDR, 2014). Regions of the Arctic differ widely in population, ranging from 94% of Iceland’s population living in urban environments to 68% of Nunavut’s population living in rural areas. And while there has been a general movement to greater urbanisation in the Arctic (AHDR, 2014), that trend is not true for all regions (Heleniak, 2014). About 4400 people reside in Antarctic in the summer and about 1100 in the winter, predominantly based at research stations of which approximately 40 are occupied year-round (The World Factbook, 2016).

For most Arctic Indigenous peoples, human responses to climate change are viewed as a matter of cultural survival (Greaves, 2016) (Cross-Chapter Box 3 in Chapter 1). However, Indigenous people are not homogenous in their perspectives. While in some cases Indigenous people are negatively impacted by sectoral activities such as mining and oil and gas development (Nymand and Fondahl, 2014), in other cases they benefit financially (Shadian, 2014), setting up dilemmas and potential internal conflicts (Huskey, 2018; Southcott and Natcher, 2018) (high confidence). Geopolitical complexities also confound responses.

Together these conditions make for complexity and uncertainty in human decision making, be it at the household and community levels to the international level. Adding to uncertainty in human choice related to climate change is the interaction of climate with other forces for change, such as globalisation and land and sea-use change. These interactions necessitate that responses to climate change consider cumulative effects as well as context-specific pathways for building resilience (Nymand and Fondahl, 2014; ARR, 2016). 

The sections below assess human responses to climate change in polar regions by examining various sectors of human-environment activity (i.e., social-ecological subsystems), reviewing their respective systems of governance related to climate change, and considering possible resilience pathways. Table 3.4 summarises the consequences, interacting drivers, responses and assets for responding to climate change by social-ecological subsystems (i.e., sectors) of Arctic and Antarctic regions. An area of response not elaborated in this assessment is geoengineered sea ice remediation to support local-to-regional ecosystem restoration and which may also affect climate via albedo changes. There is an emerging body of literature on this topic (e.g., Berdahl et al., 2014; Desch et al., 2017; Field et al., 2018), which at present is too limited to allow assessing dimensions of feasibility, benefits and risks, and governance.

Responses addressing changes in the abundance and distribution of fish resources (Section 3.2.4.1) differ by region. In some polar regions, strategies of adaptive governance, biodiversity conservation, scenario planning and the precautionary approach are in use (NPFMC, 2018). Further development of coordinated monitoring programs (Cahalan et al., 2014; Ganz et al., 2018), data sharing, social learning and decision support tools that alert managers to climate change impacts on species and ecosystems would allow for appropriate and timely responses including changes in overall fishing capacity, individual stock quotas, shifts between different target species, opening/closure of different geographic areas and balance between different fishing fleets (Busch et al., 2016; NPFMC, 2019; see Section 3.5.4). Scenario planning, adaptive management and similar efforts will contribute to the resilience and conservation of these social-ecological systems (medium confidence).

Five Arctic States, known as ‘Arctic 5’ (Canada, Denmark, Norway, Russia and the United States) have sovereign rights for exploring and exploiting resources within their 200 nautical mile Exclusive Economic Zones (EEZs) in the High Arctic and manage their resources within their own regulatory measures. A review of future harvest in the European Arctic (Haug et al., 2017) points towards high probability of increased northern movement of several commercial fish species (Section 3.3.3.1, Box 3.4), but only to the shelf slope for the demersal species. This shift suggests increased northern fishing activity, but within the EEZs and present management regimes (Haug et al., 2017) (medium confidence).

In 2009, a new Marine Resources Act entered into force for Norway’s EEZ. This act applies to all wild living marine resources, and states that its purpose is to ensure sustainable and economically profitable management of resources. Conservation of biodiversity is described as an integral part of its sustainable fisheries management and it is mandatory to apply “an ecosystem approach, taking into account habitats and biodiversity” (Gullestad et al., 2017). In addition to national management, the Joint Norwegian-Russian Fisheries Commission provides cooperative management of the most important fish stocks in the Barents and Norwegian Seas. The stipulation of the total quota for the various joint fish stocks is a key element, as is more long-term precautionary harvesting strategies, better allowing for responses to climate change (medium confidence). A scenario-based approach to identify management strategies that are effective under changing climate conditions is being explored for the Barents Sea (Planque et al., 2019).

In the US Arctic, an adaptive management approach has been introduced that utilises future ecological scenarios to develop strategies for mitigating the future risks and impacts of climate change (NPFMC, 2018). The fisheries of the southeastern Bering Sea are managed through a complex suite of regulations that includes catch shares (Ono et al., 2017), habitat protections, restrictions on forage fish, bycatch constraints (DiCosimo et al., 2015) and community development quotas. This intricate regulatory framework has inherent risks and benefits to fishers and industry by limiting flexibility (Anderson et al., 2017b). To address these challenges, the North Pacific Fishery Management Council recently adopted a Fishery Ecosystem Plan that includes a multi-model climate change action module (Punt et al., 2016; Holsman et al., 2017; Zador et al., 2017; Holsman et al., 2019). Despite this complex ecosystem-based approach to fisheries management, it may not be possible to prevent projected declines of some high value species at high rates of global warming (Ianelli et al., 2016).

In the US portion of the Chukchi and Beaufort Seas EEZ, fishing is prohibited until sufficient information is obtained to sustainably manage the resource (Wilson and Ormseth, 2009). In the Canadian sector of the Beaufort Sea, commercial fisheries are currently only small-scale and locally operated. However, with decreasing ice cover and potential interest in expanding fisheries, the Inuvialuit subsistence fishers of the western Canadian Arctic developed a new proactive ecosystem-based Fisheries Management Framework (Ayles et al., 2016). Also in Western Canada, the commercial fishery for Arctic char (Salvenius alpinus) in Cambridge Bay is co-managed by local Inuit organisations and Fisheries and Oceans Canada (DFO, 2014).

The high seas region of the CAO is per definition outside of any nation’s EEZ. Recent actions of the international community show that a precautionary approach to considerations of CAO fisheries has been adopted (high confidence) and that expansion of commercial fisheries into the CAO will be constrained until sufficient information is obtained to manage the fisheries according to an ecosystem approach to fisheries management (high confidence). The Arctic 5 officially adopted the precautionary approach to fishing in 2015 by signing the Oslo Declaration concerning the prevention of unregulated fishing in the CAO. The declaration established a moratorium to limit potential expansion of CAO commercial fishing until sufficient information, also on climate change impacts, is available to manage it sustainably. The Arctic 5 and several other nations subsequently agreed to a treaty (the Central Arctic Ocean Fisheries Agreement) that imposed a 16-year moratorium on commercial fishing in the CAO.

CCAMLR is responsible for the conservation of marine resources south of the Antarctic Polar Front (CCAMLR, 1982), and has ecosystem-based fisheries management embedded within its convention (Constable, 2011). This includes the CCAMLR Ecosystem Monitoring Program, which aims to monitor important land-based predators of krill to detect the effects of the krill fishery on the ecosystem. Currently, there is no formal mechanism for choosing which data are needed in a management procedure for krill or how to include such data. However, this information will be important in enabling CCAMLR fisheries management to respond to the effects of climate change on krill and krill predators in the future. 

Commercial fisheries management responses to climate change impacts in the Southern Ocean may need to address the displacement of fishing effort due to poleward shifts in species distribution (Pecl et al., 2017) (Box 3.4) (low confidence). Fisheries in the Southern Ocean are relatively mobile and are potentially able to respond to range shifts in target species, which is in contrast to small-scale coastal fisheries in other regions. Management responses will also need to adapt to the effects of future changes in sea ice extent and duration on the spatial distribution of fishing operations (ATCM, 2017; Jabour, 2017) (Section 3.2.4).

Herders’ responses to climate change have varied by region and respective herding practices, and in some cases are constrained by limited access to pastures (Klokov, 2012; Forbes et al., 2016; Uboni et al., 2016; Mallory and Boyce, 2017). These conditions are exacerbated in some cases by high numbers of predators (Lavrillier and Gabyshev, 2018). In Fennoscandia, husbandry practices of reindeer by some (mostly Sami) include supplemental feeding, which provide a buffer for unfavourable conditions. In Alaska, reindeer herding is primarily free range, where herders must manage herd movements in the event of icing events and the potential loss of reindeer because the movements of caribou herds (wild reindeer), both of which are partially driven by climate. For Nenets of the Yamal, Russia, resilience in herding has been facilitated through herders’ own agency and, to some extent, the willingness of the gas industry to observe non-binding guidelines that provide for herders’ continued use of traditional migrations routes (Forbes et al., 2015). In response to climate change (i.e., icing events and early spring runoffs blocking migration), the only way of avoiding high deer mortality is to change migration routes or take deer to other pastures. In practice, however, the full set of challenges has meant more Yamal herders opting out of the traditional collective migration partially or entirely to manage their herds privately. The reason to have private herds is one of adaptive advantage; smaller, privately owned herds are nimbler in the face of rapid changes in land cover and the expansion of infrastructure (Forbes, 2013). The same logic has more recently been applied by some herders in the wake of recent rain-on-snow events (Section 3.4.3.2.2) (Forbes et al., 2016). In all these regions, restrictions affecting the movement of reindeer to pastures are expected to negatively interact with the effects of climate and affect the future sustainability of herding systems (high confidence). 

The growth of the polar tourism market is, in part, a response to climate change, as travellers seek ‘last-chance’ opportunities, which, in turn, is creating new challenges in governance (Section 3.2.4.2). Polar-class expedition cruise vessels are now, for the first time, being purposefully built for recreational Arctic sea travel. The anticipated near- and long-term growth of Arctic tourism, especially with small vessels (yachts) (Johnston et al., 2017), points to a deficiency in current regulations and policies to address human safety, environmental risks and cultural impacts. Industry growth also points to the need for operators, governments, destination communities and others to identify and evaluate adaptation strategies, such as disaster relief management plans, updated navigation technologies for vessels, codes of conduct for visitors and improved maps (Pizzolato et al., 20162030) and to respond to perceptions of tourism by residents of local destinations (Kaján, 20142031; Stokke and Haukeland, 20172032). Efforts were initiated with stakeholders in Arctic Canada to identify strategies that would lower risks (Pizzolato et al., 20162033); a next step to lower risks and build resilience is to further develop those strategies (AMAP, 2017a2034; AMAP, 2017b2035; AMAP, 20182036). Opportunities for tourism vessels in the Arctic to contribute to international research activities (‘ships of opportunity’) may improve sovereignty claims in some regions, contribute to science and enhance education of the public (Stewart et al., 20132037; Arctic Council, 2015a2038; Stewart et al., 20152039; de la Barre et al., 2016).

Tourism activities in the Antarctic are conducted in accordance with the Protocol on Environmental Protection to the Antarctic Treaty, which establishes general environmental principles, environmental assessment requirements, a scheme of establishing protected areas and restrictions on waste disposal. Site-specific management tools are in place. While there are varying views amongst Antarctic Treaty Parties on the best management regulations for Antarctic tourism, these Parties continue to work to manage tourism activity, including growth in numbers of visitors. In addition to the Protocol, mandatory measures have been agreed to manage aspects of tourism activity. Industry self-regulation supplements these requirements, coordinated by the International Association of Antarctica Tour Operators, which has worked with Antarctic Treaty Consultative Parties to manage changes in operations and their impact on ice-free areas (ATCM, 20162040). 

Reducing and avoiding the impacts of climate change on infrastructure will require special attention to engineering, land use planning, maintenance operations, local culture and private and public budgeting (AMAP, 2017a2054; AMAP, 2017b2055; AMAP, 20182056). In some cases, relocation of human settlements will be required, necessitating more formal methods of assessing relocation needs and identifying sources of funding to support relocations (Cross-Chapter Box 9) (high confidence).

A discussion of the relocation of Alaska’s coastal villages is found in Cross-Chapter Box 9. Alaskan coastal communities are not the only settlements potentially requiring relocation. Subsidence due to thawing permafrost and river and delta erosion makes other rural communities of Alaska and Russia vulnerable, potentially requiring relocation in the future (Bronen, 20152057; Romero Manrique et al., 20182058). These situations raise issues of environmental justice and human rights (Bronen, 20172059), and illustrate the limits of incremental adaptation when transformation change is needed (Kates et al., 20122060). In other cases, cultural resources in the form of historic infrastructure are being threatened and require mitigation (Radosavljevic et al., 20152061). Responsibility for funding has been a key issue in the relocation process (Iverson, 20132062) as well as the overall role of government and local communities in relocation planning (Marino, 20122063; Romero Manrique et al., 20182064). The Alaska Denali Commission, an independent federal agency designed to provide critical utilities, infrastructure and economic support throughout Alaska, is now serving as the lead coordinating organisation for Alaska village relocations and managing federal funding allocations. Several efforts have also been undertaken to provide assessment frameworks and protocols for settlement relocation as an adaptive resource (Bronen, 20152065; Ristroph, 20172066).

While there has been discussion of future ‘climigration’ in rural Alaska (Bronen and Chapin, 20132067; Matthews and Potts, 20182068), a study of Alaska rural villages threated by climate change showed no outmigration response (Hamilton et al., 20162069). Several factors explain the lack of outmigration, including an unwillingness to move, attachment to place, people’s inability to relocate, the effectiveness of alternative ways of achieving acceptable outcomes and methods of buffering through subsidies (Huntington et al., 20182070) (medium confidence).

The current pan-Arctic trend of urbanisation (AHDR, 20142071), suggests that climate change responses related to infrastructure in towns and cities of the North will require significant adaptation in designs and increases in spending (Streletskiy et al., 20122072). These costs do not include costs related to flooding and other stressors associated with warming or additional costs of commercial and industrial operations. Engineers in countries with permafrost are actively working to adapt the design of structures to degrading permafrost conditions (Dore et al., 20162073) and the effects of a warming climate, for example the Cold Climate Housing Research Center of Alaska.

An analysis of the costs of total damages from climate change to public infrastructure in Alaska show the financial benefits of proactive adaptation (Melvin et al., 20172074) (Figure 3.12). In addition to global carbon emission mitigation, hardening and redesigning of infrastructure can reduce costs of future climate-related impacts. For example, retrofitting and redesigning of infrastructure in order to handle increased precipitation and warmer temperatures can reduce climate-related costs by 50%, from USD 5.5 to 2.9 billion under RCP8.5 by 2100. The cost savings of retrofitting and redesigning infrastructure is even higher than the savings from carbon mitigation, where impact costs are estimated at USD 4.2 billion under RCP4.5 by 2100. Engineering adaptation provide proportionally similar cost savings no matter which emission scenario was used.

Increases in Arctic marine transportation create impacts and risks for ecosystems and people, such as an increased likelihood of accidents, the introduction of invasive species, oil spills, waste discharges, detrimental impacts on animals, habitat and subsistence activities (Sections 3.2.4.3, 3.4.3.3.2). There has been a rise in geopolitical debate regarding national and international level regulations and policies, and maritime infrastructure to support Arctic shipping development (Heininen and Finger, 2017; AMAP, 2018; Drewniak et al., 2018; Nilsson and Christensen, 2019). Without further action leading to adequate implementation of well-developed management plans and region-specific regulations, anticipated future increases in Arctic shipping will pose a greater risk to people and ecosystems (high confidence).

The International Maritime Organization (IMO) has responsibility for the safety and security of shipping and the prevention of marine and atmospheric pollution by ships, including in the Arctic and Antarctic. There are a number of mechanisms standardising regulation and governance, such as the International Convention for the Prevention of Pollution from Ships; the International Convention for the Safety of Life at Sea; the International Convention on Standards of Training, and the Certification and Watchkeeping for Seafarers, and the newly implemented International Code for Ships Operating in Polar Waters, or Polar Code (IMO, 2017).

The Polar Code of 2017 sets new standards for vessels travelling in polar areas to mitigate environmental damage and improve safety (IMO, 2017). The Polar Code, however, currently excludes fishing vessels and vessels on government service, thereby excluding many shipping activities, particularly in the Antarctic region (IMO, 2017). Many ships travelling these waters will therefore continue to pose risks to the environment and to themselves, as they are not regulated under the Polar Code (high confidence). The Polar Code does not enhance enforcement capabilities or include environmental protection provisions to address a number of particular polar region-specific risks such as black carbon, ballast water and heavy fuel oil transport and use in the Arctic (Anderson, 2012; Sakhuja, 2014; IMO, 2017). However, both Russian and Canadian legislation provide the possibility for stricter shipping provisions in ice-covered waters. The IMO has prohibited the use of heavy fuel oil in the Antarctic.

States can individually or cooperatively pursue the establishment of Special Areas and Particularly Sensitive Sea Areas at the IMO with a view to protect ecologically unique or vulnerable and economically or culturally important areas in national and international waters from risks and impacts of shipping, including through routing, discharge and equipment measures. Continued, and in some areas, greater, international cooperation on shipping governance can facilitate addressing emerging climate change issues (Arctic Council, 2015a; ARR, 2016; PEW Charitable Trust, 2016; Chénier et al., 2017; IMO, 2017) (high confidence). Cooperation of the member states of the Arctic Council resulted in the 2011 Agreement on Cooperation on Aeronautical and Maritime Search and Rescue in the Arctic and in the 2013 Agreement on Cooperation on Marine Oil Pollution Preparedness and Response in the Arctic. These agreements can, if adequately implemented, reduce risks from increased Arctic shipping (medium confidence), however, developing more effective measures is needed as preparedness and response gaps still exist, for example, for the central Arctic Ocean.

Industry has responded to the increase in shipping activity by investing in development of shipping designs for travel in mixed-ice environments (Stephenson et al., 2011; Stephenson et al., 2013). These increases in investments are occurring in spite of the limited total savings when comparing shorter travel to increased CO2 emissions (Lindstad et al., 2016). In anticipation of spills, research in several regions has explored oil spill response viability and new methods of oil spill response for the Arctic environment (Bullock et al., 2017; Dilliplaine, 2017; Holst-Andersen et al., 2017; Lewis and Prince, 2018) (medium confidence). A comparative risk assessment for spills has been developed for the Arctic waters (Robinson et al., 2017) and Statoil has developed and uses risk assessment decision-support tools for environmental management, together with environmental monitoring (Utvik and Jahre-Nilsen, 2016). These tools facilitate the assessment of Arctic oil-spill response capability, ice detection in low visibility, improved management of sea ice and icebergs, and numerical modelling of icing and snow as risk mitigation. 

At present health adaptation to climate change is generally under-represented in policies, planning, and programming (AHDR, 20142092). For instance, all initiatives of the Fifth National Communications of Annex I parties to the UNFCCC affect health vulnerability, however, only 15% of initiatives had an explicit human health component described (Lesnikowski et al., 20112093). The Arctic is no exception to this global trend. Despite the substantial health risks associated with climate change in the Arctic, health adaptation responses remain sparse and piecemeal (Lesnikowski et al., 20112094; Panic and Ford, 20132095; Ford et al., 2014b2096; Loboda, 20142097), with the health sector substantially under-represented in adaptation initiatives compared to other sectors (Pearce et al., 20112098; Ford et al., 2014b2099; National Research Council, 20152100). Furthermore, the geographic distribution of publicly available documentation on adaptation initiatives is skewed in the Arctic, with more than three-quarters coming from Canada and USA (Ford et al., 2014a2101; Loboda, 20142102). 

Many Arctic health adaptation efforts by governments have been groundwork actions, focused increasing awareness of the health impacts of climate change and conducting vulnerability assessments (Lesnikowski et al., 20112103; Panic and Ford, 20132104; Austin et al., 20152105). For instance, in Canada this effort has included training, information resources, frameworks, general outreach and education and dissemination of information to decision makers (Austin et al., 20152106). Finland’s national adaptation strategy outlines various anticipatory and reactive measures for numerous sectors, including health (Gagnon-Lebrun and Agrawala, 20072107). In Alaska, the Arctic Investigations Program responds to infectious disease via advancing molecular diagnostics, integrating data from electronic health records and environmental observing networks, as well as improving access to in-home water and sanitation services. Furthermore, circumpolar efforts are also underway, including a circumpolar working group with experts from public health to assess climate-sensitive infectious diseases, and to identify initiatives that reduce the risks of disease (Parkinson et al., 20142108). Importantly, health adaptation is occurring at the local scale in the Arctic (Ford et al., 2014a2109; Ford et al., 2014b2110). Adaptation at the local-scale is broad, ranging from community freezers to increase food security, to community-based monitoring programs to detect and respond to climate health events, to Elders mentoring youth in cultural activities to promote mental health when people are ‘stuck’ in the communities due to unsafe travel conditions (Pearce et al., 20102111; Brubaker et al., 20112112; Harper et al., 20122113; Brubaker et al., 20132114; Douglas et al., 20142115; Austin et al., 20152116; Bunce et al., 20162117; Cunsolo et al., 20172118) (high confidence). Several regional and national-level initiatives on food security (ICC, 2012), as well as research reporting high levels of household food insecurity (Kofinas et al., 20162119; Watts et al., 20172120) have prompted greater concerns for climate change (Loring et al., 20132121; Beaumier et al., 20152122; Islam and Berkes, 20162123). A new initiative to operationalise One Health concepts and approaches under the AC’s Sustainable Development Working Group has gained momentum since 2015 (Ruscio et al., 20152124). One Health approaches seek to link human, animal, and environmental health, using interdisciplinary and participatory methods that can draw on indigenous knowledge and local knowledge (Dudley et al., 20152125). Thus far, the initiative has supported new regional-to-international networks, and proposals for its expansion. In the future, the ability to manage, respond, and adapt to climate-related health challenges will be a defining issue for the health sector in the Arctic (Ford et al., 20102126; Durkalec et al., 20152127) (medium confidence). 

Table 3.4: Response of key human sectors /systems to climate change in polar regions. Table 3.4 summarises the consequences, interacting drivers, responses, and assets of climate change responses by select human sectors (i.e., social-ecological systems) of Arctic and Antarctic regions. Also noted are anticipated future conditions and level of certainty and other drivers of change that may interact with climate and affect outcomes. Implications to world demands on natural resources, innovation and development of technologies, population trends and economic growth are likely to affect all systems, as is the Paris Agreement (AMAP, 2017b2128). In several cases, drivers of change interacting with climate change are regionally specific and not easily captured. In many cases there is limited information on human responses to climate change in the Russian Arctic.

Responses to climate change at and across local, regional and national levels occur directly and indirectly through a broad range of governance activities, such as land and sea use planning and regulations, economic development strategies, tax incentives for use of alternative energy technologies, permitting processes, resource management and national security. Increasingly, climate change is considered in environmental assessments and proposals for resource planning of polar regions. 

A comprehensive literature review of 157 discrete cases of Arctic adaptation initiatives by Ford et al. (2014b) found that adaptation is primarily local and motivated by reducing risks and their related vulnerabilities (high confidence). Several elements for successful climate change adaptation planning at the local level have previously been identified: formal analytical models need to be relevant to the concerns and needs of stakeholders, experts should be made sensitive to community perspectives, information should be packaged and communicated in ways that are accessible to non-experts and processes of engagement that foster creative problem solving be used. Furthermore, success of local government involvement in adaptation planning has been closely linked to transnational municipal networks that foster social learning and in which local governments assume a role as key players (Sheppard et al., 2011; Fünfgeld, 2015) (medium confidence). While transnational networks can be a catalyst for action and promoting innovation, there remain outstanding challenges in measuring the effectiveness of these networks (Fünfgeld, 2015). 

Adaptation through formal institutions by Indigenous people is enabled through self-government, land claims and co-management institutions (Baird et al., 2016; Huet et al., 2017). However, organisational capacity is often a limiting factor in involvement (AHDR, 2014; Ford et al., 2014b; Forbes et al., 2015) (high confidence). Interactions across scales are also dependent on the extent to which various stakeholders are perceived as legitimate in their perceptions and recommendations, an issue related to the use of local knowledge and indigenous knowledge in governance (Cross-Chapter Box 4 in Chapter 1) (AHDR, 2014; Ford et al., 2014b; Forbes et al., 2015) (high confidence). 

At a more regional level, Alaska’s ‘Climate Action for Alaska’ was reconstituted in 2017 and is now actively linking local concerns with state-level policies and funding, as well as setting targets for future reductions in the state’s carbon emission. The role of cross-scale boundary organisations in climate change adaptation planning, and how central government initiatives can ultimately translate into ‘hybrid’ forms of adaptation at the local level that allow for actions that are sensitive to local communities has proven important in Norway (Dannevig and Aall, 2015). 

At the national level, Norway, Sweden and Finland have engaged in the European Climate Adaptation Platform (Climate-ADAPT), a partnership that aims to support Europe in adapting to climate change by helping users to access and share data and information on expected climate change in Europe, current and future vulnerability of regions and sectors, national and transnational adaptation strategies and actions, adaptation case studies and potential adaptation options, and tools that support adaptation planning. Level of participation by country and the extent to which national level efforts are linked with regional and local adaptation varies. The Canadian government’s actions on climate change have been among the most extensive of the Arctic nations, including funding of ArcticNet, a Network of Centres of Excellence, and consideration of climate change by The Northern Contaminants and Nutrition North Canada programs. 

The way states and institutions manage international cooperation on environmental governance is changing in response to climate change in the polar regions. Rather than treating regional impacts of climate change and their governance in isolation (i.e., purely with a regional lens), the need to cooperate in a global multi-regulatory fashion across several levels of governance is increasingly realised (Stokke, 20092141; Cassotta et al., 20162142) (medium confidence).

In both polar regions, cooperative approaches to regional governance have been developed to allow for the participation of non-state actors. In some cases, regimes allow for a substantial level of participation by specific groups of the civil society (Jabour, 20172143; Keil and Knecht, 20172144). For example, in the Antarctic Treaty System, the Antarctic Treaty Parties included the Scientific Committee on Antarctic Research into their Protocol on Environmental Protection to the Antarctic Treaty. In the Arctic, the status of Permanent Participants has enabled the effective participation of Indigenous Peoples in the work of the Council (Pincus and Ali, 20162145). Climate change has contributed to modifying the balance between the interests of state and non-state actors, leading to changing forms of cooperation (Young, 20162146). While such changes and modifications occur in both the Arctic and Antarctic, the role of states has remained present in all the regimes and sectors of human responses (Young, 20162147; Jabour, 20172148).

Addressing the risks of climate change impacts in polar regions also requires linking levels of governance and sector governance across local to global scales, considering impacts and human adaptation (Stokke, 20092149; Berkman and Vylegzhanin, 20102150; Tuori, 20112151; Young, 20112152; Koivurova, 20132153; Prior, 20132154; Shibata, 20152155; Young, 20162156) (high confidence). Despite established cooperation in international polar region governance, several authors have come to the conclusion that the current international legal framework is inadequate when applying a precautionary approach at the regional level (medium confidence). For example, several studies have shown that the Convention on the Protection of the Marine Environment of the North East Atlantic, which applies only to the North East Atlantic, and that provides a framework for implementation of the United Nations Convention on the Law of the Sea (UNCLOS) and the Convention on Biological Diversity (CBD), are insufficient to deal with risks when applying a precautionary approach (Jakobsen, 20142157; Hossain, 20152158). 

In the Arctic, responses to climate change do not only lead to international governance cooperation but also to competition in access to natural resources, especially hydrocarbons. With ice retreating and thinning, and improved access to natural resources, coastal states are increasingly recurring to the option to invoke Article 76 of the UNCLOS (Art. 76 UNCLOS; Verschuuren, 20132159) and seek to demonstrate with scientific data, submitted to the Commission on the Limits of Continental Shelf, and within a set timeline, that their continental shelf is extended. In that case they can enjoy sovereign rights beyond the EEZ. It is very unlikely that this new trend from states to refer to Article 76 will lead to future (military) conflicts (Berkman and Vylegzhanin, 20132160; Kullerud et al., 20132161; Stokke, 20132162; Verschuuren, 20132163), although the issue cannot be totally dismissed (Kraska, 20112164; Åtland, 20132165; Huebert, 20132166; Cassotta et al., 20152167; Barret, 20162168; Cassotta et al., 20162169). 

In the Antarctic, cooperation in general does occur via UNCLOS, the Convention for the Safety of Life at Sea and the Convention for the Prevention of Pollution from Ships and the Polar Code. Global environmental and climate regimes that are implemented and managed through regional regimes (such as the Kyoto Protocol or the Paris Agreement) are also relevant for the Antarctic Treaty and its Protocol on Environmental Protection, which requires, amongst other issues, a minimisation of adverse environmental impacts. Cooperation in the Antarctic also occurs through the the Convention for the Conservation of Antarctic Marine Living Resources. Climate change and its consequences for the marine environment are a central issue for this Convention because it challenges ways to regulate and manage fisheries and designate and manage Marine Protected Areas. Nevertheless, CCAMLR has not agreed to any climate change program and at its most recent meeting, there was again no agreement to do so (Brooks et al. (2018)2170, CCAMLR Report on the 37th Meeting of the Commission, CCAMLR (2018)).

Summary of the assessment of practices, tools and strategies that can contribute to climate resilient pathways. Practices are shown with the potential extent of their contribution to resilience building, considering also seven general strategies (Biggs et al., 20122207; Quinlan et al., 20162208; Cross-Chapter Box 2 in Chapter 1). Also shown is the current level of their application in polar regions and key conditions facilitating implementation.

This section presents the status of practices, tools and strategies currently employed in the Arctic and or Antarctica that can potentially contribute to climate resilient pathways. Seven general strategies for building resilience have been recognised: i) maintain diversity and redundancy, ii) manage connectivity, iii) manage slow variables and feedbacks, iv) foster an understanding of social-ecological systems as complex adaptive systems, v) encourage learning and experimentation, vi) broaden participation, and vii) promote polycentric governance systems (Biggs et al., 2012; Quinlan et al., 2016) (Cross-Chapter Box 2 in Chapter 1).

The practices listed below are not inclusive of the many resilience-building efforts underway in the polar regions. Those described are well represented in the literature and have shown sufficient utility to merit further use (ARR, 2016; AMAP, 2017a; AMAP, 2017b; AMAP, 2018) (high confidence). Some require more refinement while others are well developed. The following sections assess the extent to which these practices operationalise resilience-building through knowledge co-production, the linking of knowledge with decision making, and implementation of resilience-based ecosystem management, considering also their application level and key facilitating conditions; a summary is presented in Table 3.5.

The co-production of knowledge and transdisciplinary research are currently contributing to the understanding of polar climate change through the use of a diversity of cultural, geographic, and disciplinary perspectives that provide a holistic framing of problems and possible solutions (Miller and Wyborn, 2018; Robards et al., 2018) (high confidence).

Several factors are important in successful knowledge co-production, including use of social-ecological frameworks, engagement of a broad set actors with diverse epistemological orientations, a ‘team science’ approach to studies, strong leadership, attention to process (vs. only products) and mutual respect for cultural differences (Meadow et al., 2015; National Research Council, 2015; Petrov et al., 2016) (high confidence). Knowledge co-production involving Indigenous peoples comes with its own set of challenges (Armitage et al., 2011; Robards et al., 2018). While advancements have been made, the practice of knowledge co-production would benefit from further experimentation and innovation in methodologies and better training of researchers (van der Hel, 2016; Vlasova and Volkov, 2016; Berkes, 2017) (medium confidence). Three aspects of knowledge co-production are highlighted below.

Renewable resource management and biodiversity conservation that seek to maintain resources in historic levels and reduce uncertainty before taking action remains the dominant paradigm in polar regions (Chapin III et al., 20092289; Forbes et al., 20152290). The effectiveness of this approach, however, is increasingly challenged as the ranges and populations of species and state of ecosystems are being affected by climate change (Chapin III et al., 20102291; Chapin III et al., 20152292). Three practices that build and maintain social-ecological resilience in the face of climate change include Adaptive Ecosystem Governance, Spatial Planning for Biodiversity, and Linking Management of Ecosystem Services with Human Livelihoods.