A major ocean current in the Arctic is faster and more turbulent as a result of rapid sea ice melt, a new study from NASA shows.
The current is part of a delicate Arctic environment that is now flooded with fresh water, an effect of human-caused climate change.
Using 12 years of satellite data, scientists have measured how this circular current, called the Beaufort Gyre, has precariously balanced an influx of unprecedented amounts of cold, fresh water – a change that could alter the currents in the Atlantic Ocean and cool the climate of Western Europe.
The Beaufort Gyre keeps the polar environment in equilibrium by storing fresh water near the surface of the ocean.
Wind blows the gyre in a clockwise direction around the western Arctic Ocean, north of Canada and Alaska, where it naturally collects fresh water from glacial melt, river runoff and precipitation.
This fresh water is important in the Arctic in part because it floats above the warmer, salty water and helps to protect the sea ice from melting, which in turn helps regulate Earth’s climate.
The gyre then slowly releases this fresh water into the Atlantic Ocean over a period of decades, allowing the Atlantic Ocean currents to carry it away in small amounts.
But since the 1990s, the gyre has accumulated a large amount of fresh water—1,920 cubic miles (8,000 cubic kilometers) – or almost twice the volume of Lake Michigan.
The new study, published in Nature Communications, found that the cause of this gain in freshwater concentration is the loss of sea ice in summer and autumn.
This decades-long decline of the Arctic’s summertime sea ice coverhas left the Beaufort Gyre more exposed to the wind, which spins the gyre faster and traps the fresh water in its current.
Persistent westerly winds have also dragged the current in one direction for over 20 years, increasing the speed and size of the clockwise current and preventing the fresh water from leaving the Arctic Ocean.
This decades-long western wind is unusual for the region, where previously, the winds changed direction every five to seven years.
Scientists have been keeping an eye on the Beaufort Gyre in case the wind changes direction again.
If the direction were to change, the wind would reverse the current, pulling it counterclockwise and releasing the water it has accumulated all at once.
“If the Beaufort Gyre were to release the excess fresh water into the Atlantic Ocean, it could potentially slow down its circulation.
And that would have hemisphere-wide implications for the climate, especially in Western Europe,” said Tom Armitage, lead author of the study and polar scientist at NASA’s Jet Propulsion Laboratory in Pasadena, California.
Fresh water released from the Arctic Ocean to the North Atlantic can change the density of surface waters.
Normally, water from the Arctic loses heat and moisture to the atmosphere and sinks to the bottom of the ocean, where it drives water from the north Atlantic Ocean down to the tropics like a conveyor belt.
This important current is called the Atlantic Meridional Overturning Circulation and helps regulate the planet’s climate by carrying heat from the tropically-warmed water to northern latitudes like Europe and North America.
If slowed enough, it could negatively impact marine life and the communities that depend it.
“We don’t expect a shutting down of the Gulf Stream, but we do expect impacts.
That’s why we’re monitoring the Beaufort Gyre so closely,” said Alek Petty, a co-author on the paper and polar scientist at NASA’s Goddard Space Flight Center in Greenbelt, Maryland.
The study also found that, although the Beaufort Gyre is out of balance because of the added energy from the wind, the current expels that excess energy by forming small, circular eddies of water.
While the increased turbulence has helped keep the system balanced, it has the potential to lead to further ice melt because it mixes layers of cold, fresh water with relatively warm, salt water below.
The melting ice could, in turn, lead to changes in how nutrients and organic material in the ocean are mixed, significantly affecting the food chain and wildlife in the Arctic.
The results reveal a delicate balance between wind and ocean as the sea ice pack recedes under climate change.
“What this study is showing is that the loss of sea ice has really important impacts on our climate system that we’re only just discovering,” said Petty.
The Arctic region is warming twice as fast as the global average1, manifested by a decrease in sea ice, snow and glaciers and permafrost degradation relative to their benchmark average states for the period between 1979 and 20052–6.
These changes can accelerate global warming further through a variety of climatic feedbacks. Carbon from thawing permafrost released into the atmosphere results in the permafrost carbon feedback (PCF)7,8.
Decreasing sea ice and land snow covers increase solar absorption in high latitudes, causing the surface albedo feedback (SAF)9,10.
Both feedbacks amplify the anthropogenic signal.
The PCF and SAF represent three of the thirteen main tipping elements the Earth’s climate system identified in recent surveys11–13.
Tipping elements are physical processes acting as positive nonlinear climate and biosphere feedbacks that, after passing a threshold, could irreversibly shift the planetary system to a new warmer state13.
They could cause additional impacts on ecosystems, economies and societies throughout the world.
The risk of triggering the tipping elements is one of the arguments for adopting the ambitious 1.5 °C and 2 °C targets in the Paris Agreement14–16.
Therefore, a rigorous quantitative assessment of the climate tipping elements under different climatic and socio-economic scenarios is required to estimate their impacts and narrow down the uncertainties.
Despite significant advances documented by the IPCC 5th Assessment Report (AR5)6, projections of future climate using general circulation models (GCMs) from the 5th climate model inter-comparison project (CMIP5) do not include the PCF17,18, although several models are set to incorporate the PCF in their next versions as part of CMIP6.
Consequently, most climate policy assessments based on results from the GCMs underestimate the extent of global warming in response to anthropogenic emissions. The SAF, on the other hand, is present in GCM climate projections through the coupling of sea ice and land surface models to atmosphere and ocean models17.
However, existing estimates of the total economic impact of climate change under different policy assumptions using integrated assessment models (IAMs) assume that radiative forcing from the SAF increases linearly with global mean temperature19,20, which is inconsistent with predictions of the GCMs21.
In this paper, we explore nonlinear transitions in the state-dependent PCF and SAF, and estimate the resulting climatic and economic impacts globally.
To perform the analysis, we develop dynamic model emulators of the nonlinear PCF and SAF, which are comparatively simple statistical surrogates of the highly complex physical models.
The emulators are integrated within PAGE-ICE, a new development of the PAGE09 IAM19,20 that includes a number of updates to climate science and economics (Methods, Supplementary Note 1).
The climatic impacts focus on changes in the global mean surface temperature (GMST) and the economic impacts focus on the net present value (NPV) of the total cost associated with future climate change.
We consider a wide range of scenarios: zero emissions after 2020, the 1.5 °C and 2 °C targets for 2100 and the nationally determined contributions (NDCs) from the Paris Agreement, and a business as usual (BaU) scenario.
We also introduce an intermediate 2.5 °C target, which requires more mitigation than is proposed by the NDCs, and an NDCs Partial scenario with a persistent under-delivery on pledges consistent with an estimated long-term effect of the US’s withdrawal from the Paris Agreement.
The scenarios extend out to 2300 to capture the effects of multiple slow physical processes including the PCF and the loss of the winter sea ice under high emissions pathways. While very long horizons like this may appear irrelevant from the point of view of the actual socio-economic processes, the well-established technological, demographic and resource constraints22,23 imply that the range of scenarios is still plausible beyond the 21st century24.
In addition to the PCF and SAF, Arctic feedbacks include carbon emissions from thawing sub-sea permafrost, boreal forest uptake and changes in ocean circulation from the melting of the Greenland ice sheet13,25, which we do not explicitly simulate it in this study.
Emissions from thawing sub-sea permafrost on Arctic shelf are poorly understood in comparison with land permafrost emissions26.
The boreal forest and Greenland ice sheet feedbacks are beyond the scope of this study, along with the non-Arctic tipping elements and other major uncertain elements in the climate system such as the cloud feedback27.
While not modelled directly, many of these effects are included implicitly in the PAGE-ICE IAM through a number of uncertain climate system parameters constrained according to the latest literature (Methods, Supplementary Note 1).
Our results show that the PCF gets progressively stronger in warmer climates, while the SAF weakens. Both feedbacks are characterised by nonlinear equilibrium responses to warming. The PCF also develops state-dependent lagged behaviour.
Compared with zero PCF and constant SAF, which are the legacy values that have been used in climate policy modelling to date, the combined nonlinear PCF and SAF cause statistically significant extra warming globally under the low and medium emissions scenarios.
For high emissions scenarios, the strength of the PCF saturates, and the weakening SAF gradually cancels the warming effect of the PCF; for BaU, this takes place from the second half of the 22nd century onwards.
Nevertheless, under all scenarios, the predominantly warmer future climate associated with the nonlinear PCF and SAF relative to their legacy values translates into marginal increases in the total discounted economic effect of climate change.
These increases, which are significant for all scenarios except for BaU, occur through additional temperature-driven impacts on economy, ecosystems and human health, additional impacts from sea level rise, as well as highly uncertain extra impacts from social discontinuities and climate tipping elements other than the PCF and SAF.
Even with the legacy PCF and SAF, emissions pathways in the range between the 1.5 °C and 2 °C targets lead to the lowest total economic effects of climate change compared to all other scenarios. Considering the nonlinear PCF and SAF makes the pathways towards the lower end of the range covered by the Paris Agreement targets marginally more economically attractive.
Implications for the total economic effect of climate change
The NPV of the total economic effect of climate change, denoted as CNPV, consists of mitigation costs, adaptation costs and climate-related economic impacts aggregated until 2300 and discounted using equity weighting and a pure time preference rate45.
We base the economic impacts due to changing temperatures on a recent macro-econometric analysis of historic temperature shocks on economic growth in multiple countries46. We project the economic impact function derived from this analysis onto the 8 global regions of the PAGE model19 using gridded population-weighted ERA-Interim reanalysis data47 for mean climatological temperatures in the base year, and adapt it to fit with the consumption-only approach for climate impacts in PAGE with no lasting effects on economic growth. Termed the level effects46, this provides a likely lower end estimate for the economic impacts and also allows one to compare directly with the default PAGE09 impact functions48–51, for which the original results for the PCF were derived41. We also carry out updates to the sea level rise driver, discontinuity impacts and mitigation costs according to the latest literature (Methods, Supplementary Note 1). The Zero Emissions scenario provides a hypothetical upper bound for the mitigation costs and includes residual impacts from historic emissions.
First, we calculated CNPV for the global climate-economy system using the base PAGE-ICE model with the legacy Arctic feedbacks, and PAGE-ICE with the nonlinear PCF and SAF representations (Fig. 5a).
In both settings, the mean total economic effects of the 1.5 °C, 2 °C and 2.5 °C scenarios are the lowest of the seven scenarios considered, while the NDC scenarios and, particularly, the BaU scenario have much higher mean total economic effects.
All the distributions have long upper tails representing a possibility of large impacts relative to the means. The tails get elongated for higher emissions scenarios and when the nonlinear PCF and SAF representations are used.
We then calculated the additional economic effect of the nonlinear PCF and SAF relative to the legacy values (Fig. 5b).
The nonlinear PCF leads to statistically significant increases in CNPV at the 5% significance level for all the scenarios considered, especially the NDC and BaU.
The nonlinear correction to the SAF leads to small but statistically significant increases in CNPV for Zero Emissions and 1.5 °C, 2.0 °C and 2.5 °C target scenarios, statistically significant decreases in CNPV for NDCs Partial and BaU, and is not significant for NDCs (all at the 5% level).
When the nonlinear PCF and SAF representations are combined, the statistical mean of the economic effect of climate change increases relative to the base estimate with the legacy PCF and SAF by $16.1 trillion (1 trillion = 1012) for the counterfactual Zero Emissions scenario ($1288 trillion base estimate), $24.8 trillion for 1.5 °C target ($613 trillion base estimate), $33.8 trillion for 2.0 °C target ($613 trillion base estimate), $50.3 trillion for 2.5 °C target ($815 trillion base estimate), $66.9 trillion for NDCs ($1390 trillion base estimate), and by $59.8 trillion for NDCs Partial ($1702 trillion base estimate).
These increases are statistically significant (5% level). We also found marginal but statistically insignificant increases for BaU ($2197 trillion base estimate), which remains the most expensive and least desirable scenario.
The mean economic impact of net additional warming from the nonlinear PCF & SAF peaks at just under $70 trillion (NPV until 2300) for NDCs.
To put this number into context, it exceeds the estimated long-term gains from economic development in the Arctic region through transit shipping routes52 and mineral resource extraction53 under high emissions scenarios by around 10 times, and could also dwarf pan-Arctic damages to infrastructure from thawing permafrost54,55.
The economic losses due to climate warming also tend to be higher in warmer poorer regions such as India and Africa20, which are also less likely to benefit from the economic opportunities associated with a warmer Arctic56.
More information: Thomas W. K. Armitage et al. Enhanced eddy activity in the Beaufort Gyre in response to sea ice loss, Nature Communications (2020). DOI: 10.1038/s41467-020-14449-z
