Warming Temperatures, Disappearing Ice

The Arctic has clearly undergone significant changes in average temperatures and sea ice extent during the last couple of decades but what does the future hold?

Overland et al. (2013) investigated Arctic climate over the next 30 years using recent climate model projections (CMIP5) for mitigation and business-as-usual scenarios. The results suggested that a seasonally nearly ice-free Arctic Ocean is highly likely before 2050. A mitigation emission scenario could lead to a 7°C temperature increase in late Autumn by the end of the century compared to a 13°C rise for business-as-usual. There is a large gap between modelled results and extrapolated sea ice loss from observations. However, what is clear, is that 'it is very likely that the Arctic climate will continue to show major changes over the next decades' with expanded areas of open water for longer periods, permafrost degradation and dramatic ecological consequences.

Difficulties in projections

Projecting future sea ice loss from global climate models comes with major difficulties and uncertainties, and this difference depends on the proposed timescale in question. Firstly, different models hindcast and forecast different results due to variations in formulas (sea ice physics, clouds, radiation, atmospheric and ocean dynamics) and parameterisation. Internal variability also means that results can differ from modelled near-term projections despite similar initial conditions. For the CMIP5 model used, 80% of the 56 ensemble 1979-2011 member trends are smaller than observed.

1966 to 2005 annual mean SAT based on observations (left) and the 36 CMIP5 models ensemble mean (right) (Overland et al., 2013)

In contrast, longer-term projections are heavily dependent on the future emission pathway used:

Monthly Arctic temperature anomalies for emission scenarios RCP4.5 (blue) and RCP8.5 (red) (Overland et al., 2013)

Sept 1900 to 2100 Arctic sea ice extent based on model simulations and compared against historical and observed records (Overland et al., 2013)

An Unseasonally Warm Christmas

Articles about an Arctic heat wave and record lows in sea ice have covered the news over the past month with many climate scientists linking the unseasonable weather to anthropogenically-induced climate change. During November sea ice extent was at a record low (1.95 million km² below the 1981-2010 average) and even declined by 50,000 km² during the middle of the month, an exceptional occurrence. 

While North Pole temperatures have been significantly higher than average and are expected to be up to 20°C higher than previous Christmas Eve temperatures, continental areas (Alaska and Siberia) have been unseasonally cold. This complies with the controversial idea 'Warm Arctic, Cold Continents' in which changes to the Arctic atmosphere (due to warming and sea ice loss) alter large-scale atmospheric patterns and effectively the cold and warm air masses are swapped.  

Daily Arctic sea ice extent for 2012-2016 against the 1981-2010 average (NSIDC, 2016)

Monthly November Arctic sea ice extent from 1978-2016 (NSIDC, 2016)

Arctic Pollution: Confounding Climate Change

After considering the impacts of Arctic climate change, one of the issues I wanted to cover was the extent to which confounding factors play a role in inducing ecosystem responses. Although modern conditions are comparable to the Holocene Thermal Maximum (HTM), the unprecedented nature of ecological shifts recorded during the 20^th century, for example in Arctic lakes, question the involvement of additional factors (Holmgren et al., 2010).

Despite being regarded as pristine wilderness; the Arctic is as vulnerable to pollution as industrial regions of Europe and North America. The long-range transport of atmospheric pollutants (sulphur and nitrogen, black carbon, aerosols, heavy metals and organic compounds) from the industrialised mid-latitudes mean that remoteness from direct anthropogenic activity does not guarantee that Arctic ecosystems will be unaffected.

We have already seen how, in addition to increasing GHG concentrations, aerosols and black carbon have contributed to the recent warming in the Arctic. In this post, I want to investigate how pollution, alongside climate change, can synergistically force ecological change.

Arctic haze is a powerful reminder of the extent of anthropogenically-released atmospheric pollutants

Arctic acidification

Ice cores retrieved from the Arctic reveal enhanced concentrations of the major acidifying pollutants during the late 20^th century. Impacts resulting from acidification have mostly been recorded on the Kola Peninsula, Fennoscandia and in the Taymir region of Russia. Following legislation implemented to counteract acidification of surface waters, SO2 emissions declined by ~67% in Europe from 1980-2000. As a result, the recovery of lakes has been observed in the Barents region, in particular on the Kola Peninsula.

Acidification in the Arctic is at most a subregional issue, impacting poorly buffered areas where acid deposition levels exceed the neutralising capacity of the system (AMAP, 2006). Interaction between pollutants increases the complexity in system response and may mean that ecological changes are not simply due to acidification. Kashulina et al. (2003) showed that the widespread environmental damage from nickel industries on the Kola Peninsula was spatially variable and toxic elements, in addition to SO2, acted as a driving mechanism of vegetation damage. With SO2 the dominant component of emissions and exceeding critical levels, previous studies linked ecosystem degradation in the region to the acidifying impact of SO2. However, evidence of widespread acidification is unconvincing and ecogeochemical mapping indicates a complex interaction between pollutants, with elements of low volatility potentially resulting in greater damage despite being emitted in much lower concentrations. Furthermore, in the study areas the increase in base cations (Na⁺, Ca²⁺, Mg²⁺, K⁺) exceeded the deposition of acid anions (SO4^2-, Cl^-) counteracting acidification.

Acid rain has also led to changes in Arctic terrestrial ecosystems

Nutrient enrichment

Although sulphur has previously been of major concern with its role as a primary acidifying agent, major reductions in sulphur emissions over the last decades and less significant declines in nitrogen has brought further attention to the impact of nitrogen deposition. Terrestrial and aquatic studies have investigated both the acidifying role of nitrogen and more recently concerns have grown over its eutrophying potential in these typically nutrient-limited ecosystems. 

Holmgren et al. (2010) compared recent diatom and geochemical records against data from the early to mid-Holocene for four lakes in western Spitsbergen. Increases in diatom valve and chrysophyte cyst concentrations occurring during the past century, with the most pronouncing alterations in the last few decades, indicate enhanced primary production and taxonomic richness. The recent dramatic shifts in diatom assemblages and significant increase in taxonomic richness contrast the relatively sparse flora recorded during the HTM, pointing to additional explanatory variables. Progressive depletion in nitrogen isotopic composition by ~2% during the last part of the 20^th century coincides with community trends, suggesting both climate change and N deposition are synergistically driving shifts towards new ecological states that have not been recorded in the Holocene.

Relative abundance of diatoms in four lakes in western Spitsbergen, Svalbard, alongside concentrations of diatom valves and chrysophyte cysts (Holmgren et al., 2010)

With the possibility of future accelerations in soil mineralisation, increased nitrogen availability (despite declining atmospheric deposition rates) could further threaten Arctic ecosystems (Hobbie et al., 1999). Similarly to acidification, ecological responses to nitrogen deposition are highly variable because of high spatial heterogeneity in lake characteristics, nutrient limitation, community structure and N supply amongst arctic lakes (Hogan et al., 2014).

Heavy metals and persistent organic pollutants (POPs)

Deriving from mining and fossil fuel combustion, the accumulation of heavy metals in biota and transferal up food chains threatens Arctic ecosystems. Concentrations of heavy metals (Pb, Cd, Cu, Fe, Mn and Zn) exceeding levels in the Earth's upper crust and local rocks have been recorded in cryoconites (aggregates of mineral particles, organic substances and microorganisms on glacial surfaces). This has vital consequences for Arctic ecosystems as contaminants have already been reported in higher-trophic-level species including seals, polar bears and reindeer (Łokas et al., 2016).

Similarly to heavy metals, POPs (organic compounds found in pesticides, industrial chemicals and products of burning that do not break down naturally) bioaccumulate and can have major health implications for wildlife and people (Vorkamp and Rigét, 2014). In response, several global and regional conventions are responsible for the elimination and reduction of POPs and therefore much of the contamination occurring today is from 'legacy' POPs. Temporal trends of some legacy POPs in Arctic biota was examined by Rigét et al. (2010), discovering that many of the  compounds in biota show significant declining trends over the last few decades consistent with the decreasing trends recorded in Arctic air.

Ecological implications of nickel mining in the Russian Arctic town, Nikel

Earth's Changing Wobble

Following an earlier post on the melting of the Greenland Ice Sheet, I wanted to look at its implications for the Earth's spin axis. By altering the distribution of weight, the melting of ice sheets has caused a dramatic shift in polar motion towards the UK at a rate of ~18 cm/year (previously the North Pole was moving away from Canada at ~8 cm/year) since 2000. 

The spin axis of the Earth was drifting toward Canada prior to 2000, however, climate-driven ice loss has shifted the direction towards the east (NASA, 2016)

Overall, the melting of the GIS is responsible for ~40% of the polar movement, while 25% is affected by gains in ice volume in East Antarctica and losses in West Antarctica. Redistribution of water on continents due to melting glaciers is also contributing to 25% of the axial wobble. Although this shift is harmless, it is a powerful reminder of the extent to which humanity is affecting the Earth.

Changes to water storage on the continents (NASA, 2016)

The video below summarises the changes in polar motion:

Ocean Acidification: The Other CO2 Problem

Many of Arctic responses to climate change explored in this blog, such as lake ecology regime shifts, terrestrial ecosystem alterations, melting of the Greenland Ice Sheet, and permafrost thawing, have focused on the warming effect of increased atmospheric CO2 levels as a driver for change. However, marine environments are also under threat from enhanced CO2 uptake in the ocean as a result of elevated atmospheric CO2 concentrations (Sabine et al., 2004). Worldwide, there has been an increase in ocean acidity of ~30% over the last 200 years, a rise that is mirrored in the Arctic Ocean (AMAP, 2013). With its low temperatures, sea ice retreat and large input freshwater input, the biological and chemical impacts of ocean acidification are expected to be most pronounced in the Arctic Ocean (Popova et al., 2014).

To what extent will the Arctic Ocean become acidified?

Aragonite reduction

The reaction of excess CO2 with water forms carbonic acid that acts to lower pH and decrease the concentration of carbonate ions (carbon undersaturation) which are used in the formation of the shells and skeletons of some organisms (e.g. corals, coccolithophorids, foraminifera, pteropods). A seasonal reduction in aragonite (one of two forms of calcium carbonate) has already been observed in Arctic surface waters, including the Canada Basin, Arctic shelves, and western Arctic Ocean. The transferral of anthropogenic CO2 to Arctic surface waters has accelerated because of significant reductions in sea ice. In addition, this increased influx of brackish water has led to further declines in the availability of Ca²⁺, decreasing alkalinity and moving us closer toward an aragonite tipping point (AMAP, 2013).

The inorganic carbon system in the Arctic Ocean (AMAP, 2013)

The role of primary production

Coastal and riverine erosion supply a large volume of organic carbon to the Arctic shelves. In some nearshore locations, the release of some of this terrigenous organic carbon store through oxidation to CO2 is not permitted by seasonal ice, however, metabolism still occurs. With some of the highest primary productivity rates observed in the world ocean, metabolism of organic carbon in Arctic shelf waters releases large amounts of CO2, which lowers pH. Upwelling of this CO2-rich water is likely to increase with reductions in sea ice, with these waters in the western Arctic Ocean then transferred to the Pacific Ocean (AMAP, 2013).

Alterations to the marine N cycle

The response of microbially-driven biogeochemical cycling to enhanced CO2 concentrations was investigated by Tait et al. (2014). Fixation of nitrogen gas by diazotrophs is the largest N source to oceans globally, while denitrification through microbial processes (reduction of oxidised N to N gas) and anammox (anaerobic oxidation of ammonium with nitrate) leads to losses of N from oceans. By altering microbially-induced ammonia oxidation sediment-water N fluxes are influenced by ocean acidification. Reductions in pH lead to enhanced nitrate uptake by sediments and increase ammonium release while decreasing release of nitrite. This could have a significant effect on primary production rates by altering the benthic supply of key nutrients.

Arctic's Ticking Time Bomb

Frozen soils in the Arctic prevent the decomposition of organic material, trapping almost 1700 billion tonnes of carbon (more than twice the amount present in the atmosphere). Climatic warming threatens to transform permafrost soils from a carbon sink to a source as permafrost melting will release methane and CO2, initiating a positive feedback mechanism. While previous models suggested abrupt transferrals of carbon to the atmosphere from melting permafrost, new research points towards more gradual and sustained release (Schuur et al., 2015).

Thawing permafrost

While it's normal for the top 30-100cm of permafrost to thaw during Arctic summers, warming is leading to the gradual deepening of the seasonally thawed soil layer. In addition, the creation of thermokarst landscapes when high ice content permafrost soils thaw leads to soil collapses and upward water seepage. The combined influence of active layer deepening and thermokarst creation initiates enhanced organic carbon erosion and mineralisation into CO2 and methane. However, the role of permafrost thaw through thermokarst initiation has previously not been incorporated into models projecting future greenhouse gas release in arctic regions, underestimating the permafrost carbon feedback. 

Melting of ancient permafrost

Thermokarst lakes

Olefeldt et al. (2016) used existing spatial data on the northern boreal and tundra permafrost region to estimate that thermokarst landscapes cover ~20% and act as a store for up to half of its soil organic carbon. In interpreting the maps and data it must be considered that data for certain landscape characteristics (e.g. ground ice content), although spatially extensive, varied in quality and resolution. Hence, the heterogeneity present in reality is not fully captured. Despite this, the results are an important advancement in evaluating the large-scale impacts of thermokarst expansion as a result of climate change.

Types and extent of thermokarst landscapes in the northern boreal and tundra permafrost region (Olefeldt et al., 2016)

Past extreme warming events

Hyperthermals (extreme warming events) occurred around 52-55.5 million years ago superimposed on a long-term warming trend. The Palaeocene Thermal Maximum (PETM) is the largest of these events, in which global temperature increased by ~5°C in a few thousand years. Simulations indicate that organic carbon decomposition in Arctic permafrost, triggered by orbital forcing (high eccentricity and obliquity), corresponds with the timing of hyperthermals. The rapid recovery in temperature following each event is likely to be a result of the replenishment of organic carbon in permafrost soils. Following the PETM, declining permafrost areal extent resulted in a diminished carbon stock and hence smaller hyperthermals thereafter (DeConto et al., 2012).

The High Arctic as a C sink

With increasing temperatures, deeper soils will develop in Arctic regions (e.g. the High Arctic and polar deserts) that were previously characterised by low productivity and lacked liquid water. This could therefore enhance C sinks in these areas (McGowan et al., 2016). 

In addition to temperature, the amount and frequency of precipitation is an important control of the carbon dynamics through alterations in soil water availability and temperature distribution. Heat is delivered to the permafrost table by percolating water, which influences microbial processes at the interface between the active layer and permafrost. Using measurements of C fluxes and sources in northwest Greenland, Lupascu et al. (2014) demonstrated that while warming alone decreased the C sink strength by up to 55%, warming combined with enhanced precipitation reduced loss of old carbon. Thus, there is a potential for parts of the High Arctic to remain strong carbon sinks, partly counteracting the expansion of C source regions at lower latitudes.

Thresholds and Tipping Points

From the Paris climate agreement coming into force to the election of Trump, this month has seen both significant steps forward and substantial setbacks for global cooperation on tackling climate change. In light of these events, a 5-year comprehensive study on Arctic ecosystems and societies published on Friday, suggesting that we may be headed for uncontrollable changes, has reinforced the necessity for global action.  As already highlighted throughout this blog, significant changes are taking place in the Arctic and there is a clear interconnectedness within and between Arctic biophysical systems, as well as the rest of the world. The Arctic Resilience Report outlines 19 potentially irrevocable tipping points that may already have been passed or will be exceeded soon. 

What is society's capacity to adapt and transform to Arctic climate change?

A 'social-ecological systems' framework was applied, recognising that human and natural systems are closely intertwined and local interactions are embedded within wider regional and global dynamics. Although individual drivers may be predictable, the triggering of multiple, interacting drivers and feedback loops make the eventual outcome difficult to predict. Thus, in an uncertain future understanding thresholds and society resilience is fundamental.

Social-ecological systems interactions across numerous spatial scales (Arctic Council, 2016)

The Greenland Ice Sheet

Similarly to Arctic sea ice, the extent of Greenland Ice Sheet (GIS) melting was above average during 2016, ranking as the 10^th highest in the 38-year satellite record. Summer air pressure was higher than average, a trend observed during the last several years, inducing drier and warmer conditions and surface melting, particularly along the western coast. Areas of darker ice with lower albedo, frequent along the west coast, were exposed acting to further enhance warming and ice sheet melting (NSIDC, 2016). 

(1) Average daily melt area anomaly for Greenland comparing melt area in each year to the 1981-2010 average
(2) Cumulative melt day area for 2015 and 2016 against the highest ice melting extent years (2007, 2010 and 2012) (NSIDC, 2016)

Why is the GIS so important?

The GIS covers most of Greenland (~81%) and at around 1.7 million km² is the second largest ice body, following the Antarctic Ice Sheet. The biggest concern over Greenland is that the complete melting of the ice sheet could cause a 7m+ sea level rise, enough to submerge 1.14 million km² of land with populations of 375 million people. 

However, sea level rise is not the only major consequence of the ice sheet melting. A fast rate of melting would produce large quantities of freshwater in the North Atlantic on top of the heavier salt water. Such an alteration in sea salinity could depress the Gulf Stream and fundamentally alter ocean circulation (Driesschaert et al., 2007). Atmospheric circulation would also be liable to significant change, with declines in heat transfer from equatorial to polar regions because of reduced global temperature difference.

Response to warming

Mass loss of the GIS is occurring through two main mechanisms: integrated surface mass balance decreases (difference between surface accumulation and ablation) and glacial acceleration inducing increased discharge (Fyke et al., 2014).

Furthermore, with the enhanced melting of the GIS, an increase in the formation of supraglacial lakes (meltwater in the surface depressions of an ice sheet) during the melt season has been apparent over the past two decades. The presence of supraglacial lakes leads to a reduction in albedo and, acting as a positive feedback, enhances surface melt (Ignéczi et al., 2016). Water reaching the ice sheet base through the drainage of these lakes transports warmth to the base of the ice sheet, enhancing basal sliding and ice velocity (Johansson et al., 2013).

The video below nicely summarises the mechanisms contributing to the decline of the GIS and the observations that have been recorded over the last few decades:

The carbon cycle

In addition to releasing vast quantities of freshwater, the melting of the GIS is exporting many nutrients, including dissolved silica. Dissolved silica is essential in sustaining diatom communities, which act as a carbon pump in the oceans. Meire et al. (2016) recently investigated the extent of dissolved silica export through the physical and chemical weathering by glaciers. Previously considered inactive in the global silica cycle, meltwater from the GIS was shown to strongly enrich surface waters with silicate. An increased supply of between 20-160% to coastal areas was predicted by the end of this century.

The decay of the GIS may also have carbon cycle feedbacks counteracting the reduction in CO2 caused by potentially higher ocean primary productivity. Ryu and Jacobson (2012) investigated the significance of glacial meltwater in transporting inorganic and organic carbon to the oceans and atmosphere. They found that, although rather modest compared to other sources, rivers draining the GIS were contributing to increased CO2 in the atmosphere. It is likely that soils, vegetation, and microbial metabolism have created this carbon reservoir beneath the GIS. Scenarios inferred by simple models suggest that CO2 released from the GIS could act as a rapid positive feedback mechanism. However, many of the parameters used are presently highly uncertain, for example the size of the CO2 reservoir beneath the ice sheet, rate of chemical weathering, and timing and trend of melting. Uncertainties are also apparent in the scaling of carbon fluxes measured from the Akuliarusiarsuup Kuua River (drains the Isunnguata and Russell Glaciers) to the entire GIS.

Another Record Low

A quick update on Arctic sea ice extent that I talked about at the start of the blog. During the second half of October, Arctic sea ice extent set new daily record lows, averaging 6.4 million km² (2.55 million km² below the 1981-2010 average) because of unusually high air temperatures and raised sea surface temperatures. This month's data is yet another stark indicator of the extent of Arctic sea ice demise. However, a mild winter may not necessarily result in record-lows next summer, as the dominating weather patterns at the time determine the amount of melting ice (NSIDC, 2016).

Daily Arctic sea ice extent for 2012-2016 against the 1981-2010 average (NSIDC, 2016)

Monthly October Arctic sea ice extent from 1979-2016 (NSIDC, 2016)

A Greening Arctic

In addition to aquatic ecosystems, Arctic warming is reshaping the structure and functioning of terrestrial environments across vast spatial and temporal scales. Increased productivity is evident through satellite imagery, whereas photographic records have revealed increased shrub cover and northward boreal forest expansion. Such shifts in terrestrial ecosystems are likely to trigger numerous feedback loops in the global climate system (Normand et al., 2013).

The northward advancement of the boreal forest into Arctic tundra

Before delving into terrestrial ecosystem responses, it is important to understand the distinct biomes that characterise Arctic environments. Arctic tundra covers around 20% of the Earth's surface and typically has low biotic diversity. It is characterised by plants with shallow root systems including low-growing shrubs, mosses, lichens and grasses that can survive in the shallow active soil layer of thawed permafrost during summer months. The Arctic treeline, marking the boundary between taiga and tundra, represents the northern limit of tree growth and is a sensitive indicator of climate variability.

Biomes of the World

Tundra responses

A warming Arctic climate is inducing earlier thawing, greening and longer growing seasons of tundra plant communities and is likely to promote the advancement of tundra into the polar deserts. Shifts in plant assemblages can influence productivity, nutrient cycling, active layer depth, decomposition, snow distribution and surface albedo. Elmendorf et al. (2012) quantified the climatic effects on tundra ecosystems by investigating responses to experimental warming over time periods of up to 20 years at 27 different locations. The results illustrated increased canopy height, amount of dead material and abundance of shrubs and graminoids (herbaceous plants with grass-like features). On the other hand, species diversity and the abundance of lichens and mosses declined in correspondence with warming.

Impacts of experimental warming on (a) community attributes and abundance (b) vegetation height (Elmendorf et al., 2012)

What's also apparent in this study is the spatial and temporal variation in ecosystem responses. Temporally, the effects of rising temperature influence a range of ecosystem processes including photosynthetic rates, soil organic matter, and biogeochemical cycling and thus the direction and extent of climate change responses vary over time. Spatially, the tundra biome covers a range in average summer temperature of more than 10°C and an array of moisture contents from wetlands to polar desert, with variations in nutrients, organic matter content, pH, and herbivore communities. For example, caribou and muskoxen herbivory counteracts the positive effect on shrub expansion that warming has, favouring the growth of graminoids (Post and Pedersen, 2008). There are also regional variations in the structure and composition of species, for example, vascular species are more abundant at warmer sites and the shrub canopy is generally taller and denser.

Despite the comprehensiveness of the study and agenda to investigate spatial variation in ecosystem responses, sites in the Siberian tundra are lacking. This has led to a bias towards Canada, Greenland and northwest Europe, neglecting a significant tundra environment. In addition, a substantial amount of unexplained variation exists regarding effect size, particularly for long-term impacts.

Study sites in the Arctic tundra (Elmendorf et al., 2012)

Arctic treeline advancement

The position of the Arctic Front (the southern boundary of Arctic air) during the summer determines the position of the northern treeline. Therefore, ocean-atmospheric phenomena and teleconnections are important determinants of the northern treeline (Sulphur et al., 2016). The increase in productivity that has been recorded in correlation with Arctic warming is also linked to the range expansion of birch, willow and alder. This shift to taller, darker canopies, with reduced albedo, and elevated evapotranspiration rates, will act as a positive feedback to warming. It is estimated that boreal forest coverage can reduce surface albedo by 25-50% compared to Arctic tundra (Macdonald et al., 2008).

In addition, the areal coverage of northern terrestrial ecosystems is a fundamental component of the global carbon cycle, influencing the Earth’s radiative balance by acting as a sink. Although expansion of woody species that are more biologically productive, creating more litter, significantly increases the carbon content of above-ground biomass, this may be modest when compared to the potential loss of carbon stored in tundra soils. An increased abundance of fungi associated with increased forest cover, potentially enhanced carbon cycling with litter from new plant species, and increased winter soil temperature may enhance the decomposition of older carbon stores (Parker et al., 2015).

The projected contraction of tundra and expansion of taiga will evidently have major impacts on Arctic environments and produce wider global feedbacks, the extent to which will vary on a spatial and temporal scale.

Current and projected (2090-2100) Arctic vegetation (ACIA, 2004)

Melting Ice Behind Carbon Emissions

We're all familiar with the concept of a carbon footprint and how our energy usage contributes to rising CO2 emissions, however, new research has put this accountability into context by calculating individual contribution to Arctic sea ice melt. The researchers estimate that a tonne of CO2 contributes to summer sea ice loss of 3m². This translates to the melting of 24m² of Arctic sea ice for the average UK resident. What's more, this research calls into question the 2°C global warming target, suggesting that the extra 1000 gigatons of CO2 released with 2°C warming would be enough for Arctic summer sea ice to disappear. The 1.5°C target agreed at COP21 would make the survival of Arctic summer sea ice a more realistic prospect.

How much Arctic summer sea ice loss are we causing?

Evolving Ecosystems

Despite the climatic fluctuations characterising past Arctic climate, recent warming over the last 150 years has been unprecedented. It has exceeded warming in other regions and even that experienced during the Pleistocene-Holocene transition (which concurred with large-scale vegetational shifts and faunal extinctions in the Arctic). However, the impacts of climate change on Arctic ecosystems has received relatively little attention compared with tropical, temperate and montane biomes (Post et al., 2009). In looking at the impacts of climate change in the Arctic, I will begin at the local and regional scale, investigating how climate change is affecting lake ecosystem structure and functioning.

Forcings, thresholds and regime shifts

Natural ecosystems respond to a number of external forcings and are sensitive to any changes. However, this response is not necessarily linear and instead alternative stable states may exist, which can be reached when changing conditions are sufficient to force the system past the threshold. What's also interesting is that reversal back to prior conditions may not return the ecosystem to its previous state, instead, the ecosystem must be pushed back beyond another switch point. In addition, gradual changes may not directly cause change but reduce the resilience of the system to deal with perturbation and shifts can then be dramatic and unexpected (Scheffer et al., 2001).

This leads to the question, how have arctic lake ecosystems responded to climatic changes? - Has warming reduced system resilience and have any irreversible shifts occurred?

Ecosystem state at five different conditions showing different degrees of resilience to perturbation (Scheffer et al., 2001)

'Lakes as sentinels of climate change'

The remoteness of Arctic lakes, as well as the sensitivity of the region to warming, makes their ecological components particularly useful indicators of climatic warming. Both directly and indirectly, lake ecosystems respond to climate warming and integrate terrestrial catchment responses, thus acting as 'sentinels' of change (Adrian et al., 2009).

Declines in ice cover have increased light penetration and lengthened the overturn period in spring, increasing turbulence and nutrient cycling and the availability of pelagic habitats (van Donk and Kilham, 1990). As a result, diatom community changes in response to warming have been recorded across Russia, Greenland and Finland, with a distinct shift from benthic taxa (e.g. Achnanthes and Fragilaria) and tychoplanktonic species (e.g. Aulacoseira) towards planktonic genera (e.g. Asterionella and Cyclotella) (Rühland et al., 2008). In addition, small-celled diatoms (e.g. Cyclotella) with high surface area to volume ratios are increasingly favoured as warming strengthens thermal stratification leading to declines in turbulence and upward nutrient transportation during the summer months (Winder et al., 2009).

Impact of climatic warming on stratification, nutrient flux and diatom cell size (Winder and Sommer, 2012)

Hobbs et al. (2010) investigated diatom compositional changes across 52 lakes in Greenland and North America from the Little Ice Age (~1550-1800) to the present day. Diatom β-diversity was implemented to capture the community turnover and the results indicated high rates in the 20^th century that has not previously been seen. This indicates that while natural climatic changes have previously not induced regime shifts, recent environmental factors (climatic warming and nitrogen deposition) have driven dramatic alterations in community dynamics. Lower β-diversity values in regions that have experienced lower rates of warming suggest climatic warming is the primary driver. It is evident that thresholds have been exceeded with regime shifts toward new ecological states. 

Cascading effects

Intensifying thermal stratification and subsiding turbulence will continue to disfavour diatoms, as their cellular requirements and silica frustules mean that their cell size is restricted to 2-4μm. Thus, phytoplankton species adapted to such conditions will become increasingly favoured (Winder et al., 2009). As diatoms contribute 20-25% of primary production global, such a significant decline in abundance could have dramatic ecological implications (Rühland et al., 2015).

Winter is Coming

Before delving into the major impacts of Arctic climate change I just wanted to share some new research that looks at an indirect effect that's a bit closer to home and a conversation favourite, the British weather. New research conducted by Overland et al. (2016) has linked recent cold winters in the UK and US to Arctic warming amplifying the effects of the jet stream's position. This has added to previous studies linking severe mid-latitude cold snaps to when the jet stream is wavy and is vital to improving our understanding of complex atmospheric interactions for predicting long-term winter weather.

Could we be expecting colder winters amid Arctic warming?

Global Teleconnections

As we have seen, Arctic climate has fluctuated over the last few millennia in response to orbital forcing and climatic feedback with recent warming trends significantly diverging from natural cyclicity. Focusing in on temporal scale, it is also apparent that the past century does not follow a steady, uniform temperature rise but instead variations characterise the increase, reflecting the Atlantic Multidecadal Oscillation (AMO). Past surface air temperatures reveal two distinct warming events in northern latitudes: one during the 1920s-40s (the early 20^th century warming event or ETCW) and the ongoing temperature increase that began in the 1980s. As mentioned earlier in the introductory post, the enhanced rate of warming in the Arctic compared to the global average is a result of the phenomena 'Arctic Amplification'.

Annual mean (Oct-Sept) surface air temperature (SAT) at 60-90°N compared to the global average from 1900-2015 (Overland et al., 2015)

Internal variability

The mechanisms forcing ETCW have been widely debated and the relative role of external forcings (greenhouse gases, volcanic and solar activity), which are often considered the main drivers of climatic fluctuations, is still under debate (Zhang et al., 2013). In a study incorporating observations and global climate model simulations, decreased volcanic activity and enhanced solar radiation during the early 20^th century was suggested to be the main cause of ETCW (Suo et al., 2013). However, considerable uncertainty surrounds solar activity reconstructions and volcanic aerosol estimations, which greatly influences model results.

Beitsch et al. (2014) analysed an unperturbed climate simulation covering the last 3000 years and found that internal variability within the Northern Hemisphere is sufficient to produce ETCW-like events. ETCW may, therefore, have been initiated through warming in the North Atlantic and associated feedback through melting ice, enhancing albedo in the Barents Sea and warming the atmosphere above. The fact that ETCW was concentrated at high latitudes in the Atlantic Arctic, in comparison to the more Arctic-wide warming of recent decades, also favours the dominance of internal variability.

Recent warming

In contrast, despite being so distant in both location and climate, changes in sea surface temperatures in the tropics may explain the recent Arctic warming experienced since the 1980s. This is all down to global teleconnections and interactions between the atmosphere, oceans, and cryosphere.

Ding et al. (2014) investigated warming during the past three decades over northeastern Canada and Greenland. Previously, this had been attributed anthropogenic climate change, however, the spatial inconsistency of warming suggests that natural climate variability may play a larger role. The North Atlantic Oscillation (NAO) is a mode of climate variability in which the pressure difference between the Azores high-pressure system and the Icelandic low-pressure shifts. In the positive phase, the pressure zones are strengthened enhancing the difference between them, whereas during the negative phase both systems are weakened. It is thought that alterations in tropical sea surface temperatures can influence the NAO by altering convection in the low latitude troposphere, stimulating atmospheric Rossby waves.

The authors argue that Arctic surface and tropospheric warming is more likely remotely forced and stimulated in the tropics, rather than initiated locally. Declining sea surface temperatures in the tropical Pacific and associated Rossby-wave activity induced a negative NAO phase and thus warming in Greenland and northeastern Canada. With increases in greenhouse gas emissions in the future, external forcing could become increasingly dominant as a determinant of Arctic climate (Bader, 2014).

Annual mean surface and near-surface temperature per decade from 1979-2012 (Bader, 2014)

The role of pollution

Recently, it has been suggested that Arctic climate is also influenced by global pollution trends. Model simulations conducted by Acosta Navarro et al. (2016) illustrate recent warming connected to declines in European sulfur emissions. Following scientific observations of acid rain and resultant pH declines in freshwater environments that raised public awareness and led to the implementation of several clean air policies in Europe and America, sulphur emissions have declined significantly. In addition to their acidifying effects, sulphur dioxide molecules form small sulphate aerosol particles in the atmosphere that effectively scatter light and cool the planet by reflecting some solar radiation back to space. Although controversial, sulphate aerosol particles may also contribute to cooling through cloud formation. 

As air and ocean currents transported to the southern Arctic latitudes pass America and Europe, the warming of the upper ocean and atmosphere at these lower latitudes has strengthened heat transport to the Arctic. The model shows that it is possible that as much as half of the recent warming trend to be a response to reduced aerosol cooling through sulphur reductions. However, this study was based on a single model, which was rerun nine times to produce a statistically significant signal, so the relative contribution of sulphate aerosol particles may have been a lot more or less than 0.5°C.

Following such illustrations of the importance of pollution in contributing to climatic change, the role of black carbon aerosols is increasingly being investigated. Black carbon aerosols absorb solar radiation and reduce albedo once deposited on snow or ice, contributing to net warming. Therefore, by reducing black carbon emissions, a reduction of 0.2 +- 0.17K could be realised by 2050 (Sand et al., 2016).

However, the contribution of sulphate and black carbon aerosols to Arctic climate will only be temporary as they continue to decline into the future, with their net contribution becoming more negligible and undetectable. Ultimately natural variability, internal and external forcings will drive future climatic change, as they have done in the past (Mauritsen, 2016).

Peering into the Past

After examining the recent changes in Arctic sea ice last week I wanted to consider whether such trends are indeed a mark of human-induced climate change or, as climate sceptics argue, part of a cycle of natural variability.

Understanding past climatic trends is fundamental for putting recent warming into context. The problem is direct instrumental records of temperature are often brief and geographically sparse, particularly in high latitude regions. The few records that do exist suggest the Arctic has warmed by ~0.6°C since the early 20^th century, peaking in 1945 (Comiso and Hall, 2014). However, this is not enough to derive the natural cyclic trends and forcing mechanisms characterising Arctic climate. Proxies in palaeo records, including tree rings, ice cores and lake sediments, are valuable and widely applied tools for reconstructing past environmental changes. To account for limitations in individual records, multi-proxy analysis is increasingly being incorporated into palaeo studies to expand spatial and temporal coverage and improve the reliability of reconstructions.

One such study was conducted by McKay and Kaufman (2014) (revised version of original study by Kaufman et al., 2009) who reconstructed Arctic summer temperatures over the last 2000 years at a decadal resolution. The study incorporated proxies from glacier ice, lake sediment and tree rings, extending the climatic record beyond the previous 400-year-long record (Overpeck et al., 1997). The results indicated recent warming that is anomalous when compared to past variability. The beginning of the 2000-year record shows decreasing Arctic summer temperatures at a rate of ~0.2°C per 1000 years until levelling out during the Little Ice Age (16^th to mid-19^th century). This cooling trend corresponds with the shift in the timing of the perihelion (when the Earth is closest to the Sun) from September to January. Despite the continued low summer Arctic insolation induced by precessional forcing, summer temperatures rose to ~1.4°C higher by 2000 than would otherwise be expected.

Comparison between temperature reconstruction of Kaufman et al. (2009) and McKay and Kaufman (2014)

In response to critics, this revision by Mckay and Kaufman is far more spatially extensive than the previous study, although bias remains toward Greenland. Moreover, in validating the temperature reconstruction, uncertainty exists (particularly in the mid-20^th century) between the calculated proxy values and measured instrumental data because of differences in their locations.

Location and archive of proxy temperature records (McKay and Kaufman, 2014)

A more recent study conducted by Florian et al. (2015) reconstructed climate based on algal pigments, stable isotopes (δ¹⁵N and δ¹³C) and diatoms. The use of sedimentary pigments meant that taxonomic groups morphologically absent from sediment cores could be reconstructed and differentiated between. Climatic change can be inferred from these biological proxies as alterations in the extent and persistence of ice cover impact nitrogen cycling and thus ecology. Reconstruction of the proxies revealed that modern conditions are comparable to those during the Holocene Thermal Maximum (HTM). However, it was also shown that confounding factors, such as atmospheric nitrogen deposition, have further altered lake ecology beyond what was recorded during the HTM.

It is evident that arctic environments have experienced significant natural variations in climate throughout the Holocene in response to changes in orbital parameters and climate feedback mechanisms. Although the recent warming trend may initially appear to be part of this natural cyclicity, analysis of past trends from multiple sources suggest that the present-day Arctic climate is responding in a fundamentally different way, largely in response to human-induced climate change.