Sea ice in the Arctic has decreased in size and thickness over recent decades because of climate change. It is melting more during summer than it is freezing during winter. Global warming, which is caused by greenhouse gases, is the reason for the loss of Arctic sea ice. The loss of sea ice has increased rapidly since the early 2000s, with a decrease of 4.7% every ten years (it has declined by more than 50% since the first satellite records). Summer sea ice may disappear completely sometime during the 21st century.
The Arctic region is now the warmest it has been in at least 4,000 years. The time when sea ice melts each year has grown longer by five days every ten years (from 1979 to 2013), mainly because freezing happens later in autumn. The IPCC Sixth Assessment Report (2021) said that Arctic sea ice area may fall below 1 million square kilometers in some Septembers before 2050. In September 2020, the US National Snow and Ice Data Center reported that Arctic sea ice covered 3.74 million square kilometers, the second-smallest area since satellite records began in 1979. Earth lost 28 trillion tonnes of ice between 1994 and 2017, with Arctic sea ice accounting for 7.6 trillion tonnes of this loss. The speed of ice loss has increased by 57% since the 1990s.
The loss of sea ice is a major cause of Arctic amplification, a process where the Arctic warms faster than the rest of the world due to climate change. It is possible that the loss of sea ice also weakens the jet stream, which could lead to more frequent and extreme weather in mid-latitude regions. Shipping routes in the Arctic are now more accessible and will likely become even more common. The disappearance of sea ice and the increased human activity in the Arctic Ocean may harm local wildlife, such as polar bears.
An important factor in understanding the loss of sea ice is the Arctic dipole anomaly. This phenomenon slowed the overall loss of sea ice between 2007 and 2021, but this trend is unlikely to continue. In March 2026, the US National Snow and Ice Data Center reported that Arctic sea ice reached about 14.22 million square kilometers, one of the lowest yearly high points recorded in more than 40 years of satellite monitoring.
Definitions
The Arctic Ocean is the large body of water located around 65 degrees north latitude. Arctic sea ice is the part of the Arctic Ocean that is covered with ice. The Arctic sea ice minimum is the day each year when Arctic sea ice reaches its smallest area, usually at the end of summer, which is typically in September. The Arctic sea ice maximum is the day each year when Arctic sea ice reaches its largest area, usually at the end of the cold season, which is typically in March. Common ways to show Arctic sea ice data include average monthly measurements or graphs that display the annual minimum or maximum area, as seen in the images next to this text.
Sea ice extent is the area where at least 15% of the surface is covered by ice. This measure is more commonly used than simply calculating the total sea ice area. It helps address challenges in identifying whether an area is open water or melted water on top of solid ice, which satellites have difficulty distinguishing. This is especially a problem during summer months.
Observations
A study from 2007 found that the decrease in Arctic sea ice was happening faster than models predicted. A 2011 study suggested that natural changes in the climate could explain this faster decline. A 2012 study, using updated models, also showed that the rate of ice loss was slightly slower than what was actually observed.
Satellite observations show that Arctic sea ice area, extent, and volume have been decreasing for several decades. The amount of sea ice that has survived for many years has dropped significantly. In 1988, about 26% of Arctic sea ice was at least four years old. By 2013, this percentage had fallen to 7%.
In mid-August through late October 2012, scientists measured wave heights of 16 feet (5 meters) during a storm in the Beaufort Sea. This is unusual because sea ice typically prevents wave formation. Waves can break up sea ice, which may create a process that worsens ice loss.
In January 2016, satellite data showed the lowest Arctic sea ice extent recorded since observations began in 1979. The lowest Arctic sea ice extent previously recorded was in 2012, when it covered 1.31 million square miles (3.387 million square kilometers). This replaced the earlier record from September 18, 2007, when ice covered 1.61 million square miles (4.16 million square kilometers). The lowest extent in September 2019 was 1.60 million square miles (4.153 million square kilometers).
A 2018 study found that sea ice thickness has decreased by 66% or 2.0 meters over the past six decades. This change has shifted ice from being permanent to mostly seasonal.
Data from satellite microwave measurements between 1978 and mid-1995 shows that Arctic sea ice extent decreased by 2.7% each decade. Later studies using the same data found that from late October 1978 to the end of 1996, the decrease was 2.9% per decade. From November 1978 to December 2012, Northern Hemisphere sea ice extent decreased by 3.8% ± 0.3% each decade.
Future ice loss
An "ice-free" Arctic Ocean, sometimes called a "blue ocean event" (BOE), is often described as having less than 1 million square kilometers of sea ice. This is because the thick ice near the Canadian Arctic Archipelago is very hard to melt. The IPCC AR5 defines "nearly ice-free conditions" as having less than 10 km of sea ice for at least five years in a row.
Predicting the exact year the Arctic Ocean will become ice-free is difficult because sea ice trends change a lot from year to year. In a study by Overland and Wang (2013), the researchers looked at three methods for predicting future sea ice levels. They found that the average of all models used in 2013 was not as accurate as observations. Only the models that predicted the fastest ice loss matched the observations. However, the researchers warned that these models might not stay accurate, so their estimate that ice-free conditions could appear in the 2040s might not be correct. They suggested using expert judgment along with models to improve predictions. Experts might use two approaches: predicting ice loss based on past trends (which would suggest an ice-free Arctic by 2020) or assuming slower ice loss with occasional large melt seasons (like those in 2007 and 2012), which could delay the ice-free event to 2028 or the 2030s. Because of these differences, scientists have made competing predictions about when the Arctic will become ice-free.
A 2006 study used the Community Climate System Model and predicted nearly ice-free Arctic conditions by 2040.
A 2009 study by Muyin Wang and James E. Overland used real-world data to improve predictions from six CMIP3 climate models. They estimated that the Arctic could be nearly ice-free by September 2037, with a possibility of this happening as early as 2028. In 2012, the same researchers repeated this process with CMIP5 models. They found that under the highest-emission scenario (RCP 8.5), the Arctic could first be ice-free between 14 and 36 years after 2007, with the most likely time being around 2035.
In 2009, a study using 18 CMIP3 models found that under a medium-emission scenario, the Arctic would likely be ice-free just before 2100. In 2012, a different team used CMIP5 models and a moderate-emission scenario (RCP 4.5) and found that while most models predict ice-free conditions after 2100, some models suggest ice-free conditions could occur as early as 2045.
In 2013, a study compared predictions from the best-performing CMIP5 models with all 30 models after adjusting for historical ice data. The results showed strong agreement. Under RCP 8.5, the Arctic is projected to be ice-free by September between 2054 and 2058. Under RCP 4.5, Arctic ice would get very close to the ice-free threshold in the 2060s but would not fully disappear by the end of the century, remaining at about 1.7 million square kilometers.
The IPCC Fifth Assessment Report (2014) said there is a risk of ice-free Arctic summers by 2050 under the highest-emission scenario.
The Third U.S. National Climate Assessment (2014) stated that the Arctic Ocean is expected to be ice-free in summer before mid-century. Models that best match past trends predict nearly ice-free Arctic summers by the 2030s.
The IPCC Sixth Assessment Report (2021) said there is "high confidence" that the Arctic Ocean will likely be practically ice-free in September before 2050 under all scenarios.
A 2021 study found that the best-performing CMIP6 models predict the first ice-free Arctic conditions around 2035 under the SSP5-8.5 scenario, which assumes rapidly increasing greenhouse gas emissions.
By combining results from multiple CMIP6 models, the first ice-free Arctic summer is likely to occur between 2040 and 2072 under the SSP3-7.0 scenario.
Impacts on the physical environment
Arctic sea ice helps keep polar regions cool and plays an important role in Earth’s climate. Its bright surface reflects sunlight during the summer, but when ice melts, the darker ocean below absorbs more heat, which warms the ocean and causes more ice to melt. This process can make it harder for ice to return during the winter. Between 1979 and 2011, the loss of Arctic ice contributed as much to global warming as a quarter of all carbon dioxide emissions during that time. This impact is similar to the total increase in nitrous oxide emissions and nearly half the increase in methane levels since the Industrial Revolution began.
As more Arctic ice is lost, its effect on global warming will grow stronger. Scientists use climate models to predict future warming, and these models include the impact of sea ice loss. They also study how losing ice affects other climate factors, such as changes in air temperature, water vapor, and cloud patterns.
In 2021, the IPCC Sixth Assessment Report stated with high confidence that Arctic summer sea ice loss will not reach a point of no return. This is because other stabilizing factors, like increased heat loss during winter, can help ice return even after a summer without ice. However, if global temperatures rise significantly, ice-free summers may happen more often and earlier in the year. A 2018 study estimated that an ice-free September could occur once every 40 years if global warming stays below 1.5°C, but once every 8 years at 2°C and once every 1.5 years at 3°C.
If global warming becomes very high, Arctic sea ice might not return during winter, leading to an ice-free winter. This could be a permanent change. A 2022 study suggested this might happen at around 6.3°C of global warming, though it could occur as early as 4.5°C or as late as 8.7°C. Such a change would add about 0.6°C of warming in the Arctic region.
Arctic warming happens faster than in other parts of the world, a process called Arctic amplification. This is linked to sea ice loss, as models show strong Arctic warming only occurs when ice is melting. In contrast, Antarctica has not warmed much because its thick ice sheet remains stable. Warming in Antarctica is instead driven by the Southern Ocean absorbing a large amount of heat.
Since the early 2000s, climate models have shown that global warming will shift jet streams toward the poles. Observations from 1979 to 2001 confirmed this, showing the northern jet stream moved northward by about 2 kilometers per year. Scientists also think the jet stream may weaken over time due to Arctic warming. This weakening could lead to more extreme weather, such as colder air escaping from the Arctic and causing longer, more intense weather events.
The Barents Sea is the fastest-warming part of the Arctic, and its ice may disappear permanently if global warming exceeds 1.5°C. Studies have found links between Barents Sea ice loss and changes in weather elsewhere, such as colder winters in Europe and changes in monsoon patterns. For example, reduced ice in the Barents and Kara Seas has been connected to cooler temperatures in parts of Eurasia and changes in snow cover. Other research suggests ice loss in these areas may affect summer rainfall in South China and the ice cover of Lake Qinghai on the Tibetan Plateau.
However, research on these connections is still uncertain, as it depends on complex climate systems. Recent studies show that Barents Sea ice loss can lead to colder Eurasian winters in autumn but warmer winters in winter. As ice loss increases, the risk of extreme cold in Eurasia decreases, but the risk of heatwaves in spring and summer rises.
In 2019, scientists found that reduced sea ice around Greenland in autumn affects snow cover during the Eurasian winter, which can intensify the Korean summer monsoon and indirectly influence the Indian summer monsoon.
A 2021 study showed that losing sea ice in the East Siberian Sea, Chukchi Sea, and Beaufort Sea during autumn can affect spring temperatures in Eurasia. If ice loss in these areas increases by one standard deviation, average spring temperatures in central Russia could drop by nearly 0.8°C, while the chance of cold weather would increase by about a third.
A 2015 study concluded that Arctic sea ice loss increases methane emissions from the Arctic tundra. Methane is a powerful greenhouse gas that contributes to global warming.
Impacts on wildlife
The decrease in Arctic sea ice is allowing humans to reach coastal areas that were once difficult to access. This change is harming land ecosystems and putting many sea animals in danger.
The loss of sea ice is connected to the decline of boreal forests in North America and is expected to increase the frequency and intensity of wildfires in this area. During warm years when sea ice retreats earlier, the amount of plant life produced in the Eastern Bering Sea increased by 40 to 50% due to more phytoplankton blooms.
Polar bears are searching for new food sources because Arctic sea ice is melting earlier and freezing later each year. This means they have less time to hunt seal pups, their usual food, and must spend more time on land hunting other animals. Their new diet is less nutritious, leading to smaller body sizes and fewer offspring, which suggests their population is decreasing. The Arctic refuge is the main place where polar bears live and raise their young. Melting sea ice is causing the loss of species in this area. There are approximately 900 polar bears in the Arctic refuge national conservation area.
As Arctic ice melts, microorganisms create substances that affect ice melting and stability. Some bacteria in ice pores produce polymer-like materials that might help strengthen the ice. Scientists from the University of Washington studying this process believe these polymers could make ice more stable. However, other researchers have found that algae and other microorganisms create substances like cryoconite or pigments that speed up ice melting and encourage the growth of these microorganisms.