Several studies indicate that sea ice is shrinking. The surface of the sea ice at the end of summer has declined very rapidly, from 8.5 million km2 during the period 1950-1975 to 5.5 million km2 in 2010. The specialized satellite CryoSat-2 was set in orbit in April 2010 after the failure of the first CryoSat satellite in 2005. It must provide more precise information on the quantities of polar ice.
In the Arctic
(A polar bear leaping between two blocks of melting sea ice on the island of Spitzberg in the Norwegian Svalbard Archipelago. )
Satellite observations show that the sea ice is losing the most surface area in the Arctic Ocean. In addition, thinning of these sea ice, especially around the North Pole, has been observed.
The average ice age over the period 1988-2005 has changed from more than six years to less than three years. The reduction in the average extent of the Arctic sea ice since 1978 is of the order of 2.7% per decade (plus or minus 0.6%), its minimal extent at the end of summer decreasing by 7.4%. decade (plus or minus 2.4%).
The warming in this region is of the order of 2.5 °C (instead of 0.7 °C average on the planet), and the average ice thickness has lost 40% of its value between the periods 1958-1976 and 1993-1997.
2007 marks a minimum of sea ice in summer. In that year, satellite observations show an acceleration of Arctic sea ice melting, with a loss of 20% of the summer sea ice surface in one year.
The observations made during the Tara expedition, a private initiative under the auspices of the European Damocles program (Developping Arctic Modeling and Observing Capabillities for Long-Term Environmental Studies) from September 2006 to December 2007, indicate that the changes initiated in the Arctic ocean are deep and irreversible. In addition, Greenland saw its glaciers shrink by 230 to 80 billion tonnes per year from 2003 to 2005, contributing to 10% of the current 3mm annual sea level rise.
A study dating from 2010 shows an anticorrelation and a bipolar shift between the temperatures of the poles during the twentieth century: when one pole heats up, the other cools, and the phases of heating/cooling succeed each other in cycles of a few dozen years. The link between the two poles would be the Atlantic Ocean. According to the authors, “the recent acceleration of Arctic warming is the result of a positive reinforcement of the warming trend (due to increased greenhouse gases and other possible forcings by the warming phase due to multi-decadal climatic variability (due to fluctuations in the circulation of the Atlantic Ocean)”.
The disappearance of sea ice in summer decreases the albedo of the Arctic, reinforcing the warming of the Arctic Ocean during this season. Part of the accumulated heat is transmitted to the atmosphere during the winter, changing the circulation of the polar winds. These changes would result in arctic air incursions at mid-latitudes that explain the severe winter episodes that hit the United States and Europe during the 2010 to 2012 winters. However, statistics on these recent events are still too few to derive an final conclusion on impact.
In the Guardian, September 17, 2012, Peter Wadhams, director of the Department of Polar Ocean Physics at the University of Cambridge, England, claimed that the Arctic sea ice could have completely disappeared in summer by 2016. Observations 2016-17 confirms that the Arctic is the warmest zone in the world, for reasons that are now clearly anthropogenic.
In Antarctica, the civil satellite imagery available (since 1979) did not show a total decrease in area, unlike the Arctic sea ice. However (especially thanks to vehicles and underwater robots), there are large areas of thinning and a number of exceptional phenomena. Thus, 3,500 km2 of the Larsen B sea ice, fragmented in March 2002, the first crevasses having appeared in 1987, and 5,800 km2 of the Larsen C sea ice broke away from the continent in July 2017. This sea ice helps keep continental ice in place, and its dislocation can lead to an increase in ice flow to the sea. In 2016, two studies show that Antarctica actually loses ice for longer than previously thought, thanks in particular to recently declassified satellite images (made by North American spy satellites).
This sea ice was considered stable for 10,000 years. As of April 2009, the Wilkins plaque, which once totaled 16,000 km2, also broke away. In general, the area of sea ice surrounding the Antarctic continent has been increasing steadily over the last 30 years. Scientists have questioned the reasons for the apparent extension of these Antarctic ice. Among the explanations proposed, according to a Dutch study, the melting of the ice that covers the continent could be at the origin of this extension. Indeed, meltwater would cause a cooling of the sea surface, which would promote the formation of sea ice.
In 2013, the IPCC was cautiously estimating that the melting of Antarctica (70% of the world’s available freshwater supply) would be moderate (unlike that of the Arctic) and contribute only a few centimeters to the level rise from the sea in 2100.
In 2014, a study by NASA and the University of California at Irvine published in May 2014 in the journals Science and Geophysical Research Letters concludes that a part of the West-Antarctic Ice Sheet, melting quickly, appears to be in a irreversible decline state, nothing can stop the glaciers; forty years of observing the behavior of the six largest glaciers in this region of the Amundsen Sea in West Antarctic (Pine Island, Thwaites, Haynes, Smith, Pope and Kohler) indicate that these glaciers “have passed the point of no return”; they already contribute significantly to sea-level rise, releasing almost as much ice in the ocean annually as the entire Greenland ice sheet; they contain enough ice to raise the general level of the oceans by 4 feet (1.2 meters) and melt faster than most scientists expect; for the lead author (Eric Rignot), these findings imply an upward revision of the current sea level rise forecasts.
Paleoclimatology seeks to better understand what happened during previous deglaciations, especially after the last glacial maximum (occurred 26,000 to 19,000 years ago). We know that the North Pole has irregularly lost or gained significant amounts of ice, but until around 2010 little information was available for the South Pole. What was known about the response of the Antarctic ice cap to the last postglacial warming was mainly based on chrono-sequences from isotopic analyzes. These analyzes came partly from a few cores of ice and secondly from cores of marine sediments, which are temporally quite inaccurate and geographically limited to a few terrestrial areas, or else marine and shallow.
Recently, the study of marine deposits of debris massively transported by icebergs (so-called “BIRD” for iceberg-rafted debris) has made it possible to reconstruct the melting dynamics of Antarctica in previous millennia and to compare it with similar data already available and used for the North Atlantic. In 2014, eight events were documented increasing the flow of icebergs from various parts of the Antarctic ice cap between 20,000 years ago and 9,000 years ago. This corrects the previous scenarios according to which the main glacial retreat would have been initiated by a continuous melting of the ice until the end of the Holocene.
The flow of large icebergs dropped by the Antarctic ice sheet peaked around 14 600 years ago, the first direct evidence of a contribution from Antarctica to a sudden rise in the ocean level. According to Weber et al (2014), climate simulation models incorporating this type of forcing suggest positive feedbacks, and suggest that small ice cap disturbances could contribute to a possible mechanism for rapid sea level rise.
From 2015 to 2016, several new studies confirm that a region considered stable for at least 10,000 years melts, that an apparent stability of the frozen expanses concealed ice thickness losses; these studies shed new light on the subject and suggest that by better integrating the atmosphere and ocean currents in the models, the stability of the Antarctic ice cap would be less than previously thought.
In Nature in October 2015 Nick Golledge, ice sheet modeler and his colleagues, suggest that a warming of more than 1.5 to 2 °C in 2100 could via the melting of Antarctic ice make the sea rise by 39 cm more before 2100 , and 3 meters in 2300 … with effects that will be measured in centuries or millennia.
Another study (in Nature, December 2015) suggested an equivalent contribution of + 30 cm in 2100 for the average sea level, due to the melting of the sea ice, but the authors acknowledged that they had not taken into account the complex phenomena surface melting and collapse of ice cliffs which – they estimate – could aggravate the loss of ice.
Finally, in March 2016, Pollard & DeConto seek to take these phenomena into account. They conclude that because of them, if the threshold of 2 °C is exceeded, then the situation will be much more serious than previously estimated by the IPCC with an inevitable melting of the Antarctic ice cap, which will raise the sea more than one meter by 2100 and more than 15 meters in 2500; which will literally ‘recartograph’ countries and lands. Their new climate model integrates the ice loss induced by currents warming ocean (which can eat ice from below as the atmosphere refines it). It is more sensitive to the fact that surface meltwater lakes and ponds tend to seep into the ice sheet (by warming it) and by digging channels, widening faults, which can chain reaction resulting in the disintegration of pieces of ice trays and accelerated collapse of ice cliffs. This model shows itself in retrospect capable of better explaining that we have done so far several key geological periods that have long intrigued scientists. Thus, before the beginning of the last glaciation about 130,000 – 115,000 years ago, the sea was 6-9 meters higher than today but with a CO2 level about 30% lower. And 3 million years ago, with a CO2 level comparable to ours, the ocean was 10 to 30 meters higher. This model has been tested and calibrated based on data describing the past to make it more predictive of future sea level rise. Nick Golledge believes that the new evaluation (of 2016) is credible even if the method of calculation is still a little speculative given the complexity of the phenomena involved.