With its melting sea ice, greening landscapes and rapidly warming temperatures, the Arctic region is rife with examples of how climate change is impacting the planet. Arctic sea ice, in particular, is a significant climate indicator, as its downward trends in thickness and extent – the area of the ocean covered by at least some ice – show how warmer ocean and air temperatures affect seasonal ice.
Climate change’s impacts on the opposite pole, however, are a bit less clear. While the Arctic has seen steady declines in sea ice thickness and extent in recent decades, Antarctica’s sea ice has remained stable: In fact, for several decades, sea ice extent near the South Pole crept upwards, though this trend abruptly stopped in 2014. This dichotomy between the two poles – the clear ways climate change is impacting the Arctic sea ice, and the less-obvious ways it’s impacting Antarctica’s – is something scientists are actively working to understand. In the meantime, Antarctica’s relative lack of sea ice melt has long been used by doubters of climate change as proof that climate change isn’t occurring, or isn’t as severe as scientists say it is.
Of course, as the IPCC noted in 2021, “human influence on the climate system is now an established fact,” with “sustained changes” documented “in all major elements of the climate system: the atmosphere, land, cryosphere, biosphere and ocean.” Still, Antarctic sea ice extent remains an interesting question for climate scientists. We spoke with Antarctic and Arctic experts to help explain what’s going on at the two poles, and what questions still need to be answered.
Arctic trends in sea ice melt
Arctic sea ice grows and shrinks with the seasons, melting from around March to September and reforming in the cold winter months. Scientists measure sea ice extent and thickness using satellites, and satellite records have shown that, apart from this seasonal shift, Arctic sea ice extent has steadily declined over the last four decades. Paleoclimate records, such as ice cores, show that the current decline is the most severe in at least 1,500 years. And it’s not just the extent of the sea ice that’s shrinking, making the Arctic look physically smaller on a map: the region is also losing its older, thicker ice. When the satellite record began in 1979, average August ice thickness in the Arctic was over 2 meters; in August 2020, average ice thickness was just above 1 meter. As Parkinson notes in a 2019 study in Proceedings of the National Academy of Sciences:
“The decreases have accelerated since the 1990s and have been part of a consistent suite of changes in the Arctic, including rising atmospheric temperatures, melting land ice, thawing permafrost, longer growing seasons, increased coastal erosion, and warming oceans. Overall, it has been a consistent picture solidly in line with the expectations of the warming climate predicted from increases in greenhouse gases. In particular, modeled sea ice predictions showed marked Arctic sea ice decreases, and the actual decreases even exceeded what the models predicted.”
Figure 1 – In 2021, Arctic sea ice was 12th lowest on record. Credit: NASA Earth Observatory images by Joshua Stevens, using data from the National Snow and Ice Data Center. [Source: NASA]
Antarctic trends in sea ice melt
Ice in Antarctica, however, has not seen this consistent decline. For the first 35 years of satellite records, Antarctic sea ice extent increased by about 1 percent per decade. While the Arctic experienced record lows in sea ice extent between 2012 and 2014, the Antarctic saw record winter highs.
Why this period of growth? Scientists have several theories – which we’ll address in the next sections – but they haven’t quite zeroed in on a conclusion.
“Lots of work was done to work out why this positive trend existed, and there still isn’t a definite single answer,” said Caroline Holmes, Polar Climate Scientist at the British Antarctic Survey.
In 2014, however, this growth trend in Antarctic sea ice extent reversed – and did so dramatically. As Parkinson notes, “since then, the decreases have been so great that the yearly averages for 2017 and 2018 are the lowest in the entire 1979–2018 record, essentially wiping out the 35 years of overall ice extent increases in just a few years.”
It’s also important to note that, while the Antarctic did see growth in sea ice extent for multiple decades, the magnitude of that growth was far smaller than the magnitude of the Arctic sea ice loss during that same time period.
“Looking at the trend in annual mean ice extent over the full period we have satellite data for (1979 to the end of 2021), the trends in the Antarctic are nearly ten times smaller than those in the Arctic (as well as being positive, not negative),” Holmes said. “This is a fair comparison, as the average ice extent over those periods is similar.”
Differences in geography
As Holmes noted, scientists are still working to figure out exactly why Antarctic sea ice extent grew slowly for more than three decades, only to start declining in 2014. Several of the theories, however, have to do with geography. Antarctica is an ice-covered continent surrounded by ocean – the opposite of the Arctic, an ice-covered ocean surrounded by land. That land surrounding the Arctic Ocean limits the amount of ice that can grow there – meanwhile, sea ice can grow unimpeded from the Antarctic coastline.
“Overall that means that Antarctic sea ice cover varies more over a year than Arctic sea ice,” Holmes said.
Antarctic sea ice is so vast, it isn’t usually considered to be a single entity – instead, scientists look separately at the sub-regions of the Antarctic, including the Ross Sea sea ice, Amundsen sea ice, and Weddell sea ice.
Figure 2 – Map of Antarctica. Credit: National Snow and Ice Data Center
“They don’t ‘talk’ to each other,” says François Massonnet, a Postdoctoral Research Fellow at the Université catholique de Louvain in Belgium, of the different regions of Antarctic sea ice. “They’re completely different. That’s why it’s dangerous – although tempting – to consider sea ice extent as one indicator of global change in the Southern Hemisphere, because it can be the result of compensating dynamics in different regions across the Antarctic.”
Thin ice and strong winds
Antarctic sea ice tends to be thin – about a meter or so on average, But it’s covered by a thick layer of snow, making it well-insulated from changes in the atmosphere. That means ocean temperatures, rather than air temperatures, tend to have more of an impact on Antarctic sea ice – and that future warming of the Southern Ocean could mean increased melting of Antarctic sea ice.
The Arctic, meanwhile, is home to thick sea ice but a thin insulating layer of snow, making it particularly vulnerable to warming air temperatures. And whereas rising ocean and land temperatures are largely to blame for Arctic sea ice loss – a trend that’s amplified as sun-reflecting ice retreats, replaced by dark, sun-absorbing water – winds and ocean currents have a bigger impact on sea ice extent in Antarctica.[6,7]
“You have very strong winds in Antarctica,” Massonnet said. “The winds slide towards the ocean, and they can push the ice away from the coast locally.” These gusts can impact the growth of sea ice: winds encircling the continent can also push the ice pack northwards, leaving more open sea along the coast to freeze into ice. Holmes also noted another wind-related theory: that wind pattern changes due to shifts in the ozone hole have led to sea ice increases. Winds in the Arctic, meanwhile, don’t have as much impact on ice growth there, thanks to the region’s constrained geography.
“If I had to really summarize everything…I would say that Arctic sea ice is driven by thermodynamic processes – so anything related to changing air temperature – while the Antarctic is mostly driven – or partly driven – by dynamic processes: in particular, winds and changes in atmospheric circulation,” Massonnet said.
Another theory for Antarctic sea ice trends? Melting glaciers.
“As ice sheets melt, they naturally discharge fresh water into the salty ocean,” Holmes said. “This fresh water is lighter than the ocean water, and so naturally floats near the surface, leaving a cold, fresh layer at the ocean surface, in which it is easier for sea ice to form. Therefore, an increase in this discharge from ice sheets could have been affecting trends in sea ice.”
There are a few other possible explanations for the trend in Antarctic sea ice. It’s possible that the Antarctic region is in the midst of a long-term decline in sea ice, but because of strong natural variabilities in the climate system, scientists may not be able to see the decline for another several decades. It could also be that there isn’t a strong trend at all in Antarctica, and that scientific observations so far are just statistical noise. Lastly, the Antarctic sea ice observations could be impacted by data mismatches, since different satellite sensors are used throughout the year to monitor changes.
“And of course, the worst would be that we have all of [these factors] that combine to create this increase in Antarctic sea ice,” Massonnet said.
What’s next for these regions?
Global climate models predict the Arctic could see its first ice-free summer mid-century. Peng et. al finds that, under a medium emissions scenario, models predict the first ice-free summer to arrive by the 2050s, while in a high-emissions scenario, the Arctic will be ice free by the early 2040s. However, that study notes that “climate models may be collectively underestimating the rate of change in sea ice extent,” and that it’s possible, when combining model projections with observed sea ice changes over the last few decades, that the Arctic could be ice-free in summer as early as the 2030s. Claims that the Arctic will be ice-free even earlier, however, have not been supported.
Figure 3 – Model simulations of Arctic sea ice extent for September (1900-2100) based on observed concentrations of heat-trapping gases and particles (through 2005) and four scenarios. Colored lines for RCP scenarios are model averages (CMIP5) and lighter shades of the line colors denote ranges among models for each scenario. Dotted gray line and gray shading denotes average and range of the historical simulations through 2005. The thick black line shows observed data for 1953-2012. (Figure source: adapted from Stroeve et al. 2012). [Source: globalchange.gov]
Projecting changes in Antarctica, however, is trickier. Models have predicted, accurately, a decline in Arctic sea ice, but those same models also historically predicted a decline in Antarctic sea ice – a decline that hasn’t been observed. So scientists are working to figure out why there’s this difference between observed changes and modeled changes in Antarctica.
“No scientist would bet even one penny on where the ice will be in ten years in the Antarctic,” Massonnet said. “Although I can tell you the Arctic will continue to melt.”
What questions remain
Massonnet notes that there’s a data gap between Antarctica and the Arctic. Antarctica is a remote, harsh environment, and the sea ice nearly disappears every summer, making taking measurements difficult. Satellites can measure the height of the ice, but scientists still need to make assumptions about how much of that height is ice and how much is snow. The Arctic, meanwhile, has seen scientific surveys for the last few decades, thanks to its relatively moderate environment.
“We have very poor knowledge of what actually happens in the Antarctic, and in particular over sea ice,” Massonnet said.
In order to predict how Antarctic sea ice is going to change in the future, scientists need accurate climate models attuned to Antarctic ice. But those models had predicted that, with an increase in carbon dioxide in the atmosphere, Antarctic sea ice would already have seen steeper declines. To better hone these models, scientists need to understand why they, so far, have differed from the Antarctic’s reality, Holmes said.
“Was reality just a very particular pattern of natural variability, that is, a very unlikely event?” she said. “Or, is there something that the models are collectively ‘missing’ that would enable them to recreate reality? For example, do they recreate wind patterns poorly, does sea ice respond incorrectly to wind patterns, or do errors in how the ocean is recreated matter?”
How Antarctic sea ice is used to fuel climate denial
Some outlets, including social media accounts like PragerU and institutions like the Global Warming Policy Foundation, have used Antarctic sea ice growth to shed doubt on the scientific consensus of climate change. These social media outlets and advocacy groups sometimes zero in on the mismatch between climate models’ projections and the observed sea ice growth in Antarctica.
For instance, social media posts have claimed that Antarctic ice has grown since 2014 and that Arctic ice “is reaching its highest point in 20 years.” On the contrary, Antarctic sea ice extent hit a record minimum in 2017, with the 2020 minimum extent falling below average but still higher than the 2017 record. The Arctic sea ice minimum for 2021 was the twelfth lowest in the satellite record. And both Antarctica and the Arctic are each losing “significant amounts of land based ice as a result of human-caused global warming,” according to NASA.
Another post claims that “NASA now admits that polar ice has increased beyond its 1979 volume and there’s been no significant warming in 18 years.” In reality, Arctic sea ice extent has declined since 1979 and, while Antarctic sea ice extent has not shown the same strong record of decline, polar ice as a whole has not increased. NASA also states that “direct observations made on and above Earth’s surface show the planet’s climate is significantly changing” and that “human activities are the primary driver of those changes.”
“Yes, Arctic sea ice is melting faster than the models expected. But models also predicted that Antarctic sea ice would decrease, yet Antarctic sea ice is increasing,” one video from PragerU states.
But models have proven useful and reliable for many different parts of the climate system; their challenges with Antarctica are not enough to discount that.
“‘Because they cannot explain everything, they can explain nothing’ – that’s a dubious argument for me,” Massonnet said. “We can explain planes, but we don’t have a good theory for turbulence.”
Like everything, the climate is subject to natural variations. But across much of the planet, as climate change progresses, that natural variability tends to get hidden by longer-term trends. Antarctica is one of the few places left where that’s not happening, Holmes said, so it sticks out as a “stronghold for climate denial.”
“The fact that we don’t have a final answer for why the increases in sea ice happened probably also makes it easier to question; even though we know lots of things that probably contributed,” she said.
-  Dai et al. Arctic amplification is caused by sea-ice loss under increasing CO2 (2019). Nature.
-  IPCC (2021) Climate Change 2021: The Physical Science Basis. Contributions of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change.
-  How is Arctic sea ice changing? (2020). National Snow and Ice Data Center.
-  Climate Change: Arctic sea ice summer minimum (2021). Climate.gov.
-  Parkinson (2019). A 40-y record reveals gradual Antarctic sea ice increases followed by decreases at rates far exceeding the rates seen in the Arctic. Proceedings of the National Academy of Sciences.
-  Why is Arctic sea ice decreasing while Antarctic sea ice is not? (2020). The Royal Society.
-  Nghiem (2016). Geophysical constraints on the Antarctic sea ice cover. Remote Sensing of Environment.
-  Peng et. al (2020) What Do Global Climate Models Tell Us about Future Arctic Sea Ice Coverage Changes? Climate.
-  Turner et al. Non-annular atmospheric circulation change induced by stratospheric ozone depletion and its role in the recent increase of Antarctic sea ice extent (2009). Geophysical Research Letters.
-  Auger et al. Southern Ocean in-situ temperature trends over 25 years emerge from interannual variability (2021). Nature Communications.
-  Hausfather et al. Evaluating the Performance of Past Climate Model Projections (2020). Geophysical Research Letters.