One of the key issues facing climate modelers is the feedback between shrinking sea ice and the Arctic region's declining albedo, the fraction of incoming sunlight that is reflected back into space. Lower albedo means that ice and seawater absorb more solar warmth, causing more ice to melt, which further decreases albedo.

Donald Perovich of the US Army's Cold Regions Research and Engineering Laboratory in New Hampshire told attendees that modeling would be easier if the ice stayed the same all summer long. But the Arctic Ocean presents an ever-changing mosaic of ice, ponds and open water, he said.

Ponds, areas of meltwater on the ice surface, vary from day to day and they are difficult to track by satellite, especially through the clouds typical of Arctic summer. Ponds can cover as much as 80% of undeformed ice one day, then drop to just 10% a few days later, as they drain.

As sea ice breaks, areas of open ocean water, known as leads, appear. These have lower albedo than both ice and ponds, Petrovich said. The key issue regarding the changing albedo is timing: when does melt begin in the spring, when do ponds first form, when do they develop and when does fall freeze-up begin? These dates are not uniform across the Arctic in any given year, he added, and they must be determined annually for the entire region. Further complicating the equations is the increasing ratio of first-year to multiyear ice. First-year ice is thinner and melts sooner, while ponding on its surface occurs earlier.

Albedo also varies depending on whether the ice is snow-covered. With 10 centimetres or more of snow on the surface, the albedo is the same anywhere, Petrovich said, regardless of whether an ice floe or a parking lot lies beneath it.

The ice-albedo feedback cycle begs the question of why there is any ice left in the Arctic at all. There are restraints on the process, Petrovich said. One is the surface-scattering layer – one to three centimetres thick – that forms on the ice surface, looking like snow, which is actually a slush of refrozen meltwater. This raises albedo. Also clouds, prevalent in summer, cut incident sunlight in half – another brake on feedback.

Petrovich concluded by affirming that Arctic solar-heat input is greatly increasing, especially over ocean water, whose extent is rising. Especially worrisome is the effect of warming on first-year ice, which is becoming relatively extensive. Scientists must learn how ponds evolve over the summer and what happens as ice thins, he said.