One-third of Arctic sea-ice retreat could be due to natural variability

Most of the Arctic Ocean is covered by a floating platform of sea ice, which grows and shrinks with the seasons each year. At its maximum, usually in March, it covers around 15.86 million square kilometres (6.12 million square miles), while its minimum, usually in September, averages 6.71 million square kilometres (2.59 million square miles). However, these maxima and minima (based on a 1979 to 2000 average) have become smaller and smaller in recent years, with the September 2011 minimum, for example, at just 4.33 million square kilometres.

The shrinking of Arctic sea ice could be a boon for shipping and resource extraction, but for Arctic wildlife these sudden changes are proving hard to adapt to. Shrinking sea ice also has global effects, influencing the climate far beyond the Arctic Circle.

Climate models have failed to capture this diminishing sea-ice trend, suggesting that the models are missing significant climate processes. One such component could be an element of natural climate variability that until now has been overlooked.

To investigate this possibility Jonny Day from the University of Reading, UK, and his colleagues at the Japan Agency for Marine Earth-Science and Technology (JAMSTEC), decided to study the role that the Arctic oscillation (AO), Atlantic multi-decadal oscillation (AMO) and the Atlantic meridional overturning circulation (AMOC) play in decadal sea-ice variability, using simulations of pre-industrial climate.

These three natural cycles play a vital role in moving heat around the northern hemisphere. The AMOC is the Atlantic part of the thermohaline circulation. "It pulls warm water along the surface to the Arctic where it cools and sinks, returning south along the bottom of the ocean," explained Day. The AMO, meanwhile, is a sea-surface temperature pattern in the North Atlantic, most likely driven by the AMOC, which appears to oscillate on a 65 to 80-year time period. The AO describes wind-driven cycles that move pockets of air around in the northern hemisphere. In negative years, warm air extends into the Arctic, while in positive years the Arctic is more isolated.

Day and his colleagues ran multi-centennial pre-industrial model simulations, comparing the results with modern-day satellite data and observational records going back to the 1950s. They found that the AO had very little influence on sea-ice extent, but that the AMO and AMOC played a significant role.

"We show that when the AMOC is high there is more heat transport to the Arctic, limiting ice formation and increasing ice melt, depending on the season," Day told environmentalresearchweb. Right now the AMO is in a positive phase, bringing excess warmth to the Arctic, and possibly contributing to as much as 30% of the currently observed decline.

However, as Day points out, that still leaves a large chunk of ice melt which cannot be explained by natural variability, and is most likely due to man-made greenhouse-gas emissions.

The AMOC and AMO are still poorly understood and it is not yet clear what kind of influence they will have on Arctic sea ice over the coming decades. Day and his colleagues will continue to work on this issue. They hope that eventually their work will help provide more accurate predictions of sea-ice extent, ultimately feeding into more reliable climate models and enabling us to better prepare and plan for the future.

The results are published in Environmental Research Letters (ERL).