A NASA study released in early April has found that very little of the thick sea ice in the Arctic has been replaced during the perennial melt/freeze cycle the Arctic undergoes. The replenishment of the sea ice is critical to maintaining a stable covering of ice during the Arctic summer season. These findings reinforce earlier data from NASA that indicated a 14 percent drop in perennial ice between 2004 and 2005.

The perennial ice coverage in the Arctic falls due to summer melting and transport of ice out of the Arctic. Perennial ice is typically ten or more feet thick while seasonal ice is much thinner. The seasonal ice replaces the lost perennial ice, but must survive an entire summer season to become part of the perennial ice cover. If none of the seasonal ice survives then previously ice-covered areas may become open water.

"Recent studies indicate Arctic perennial ice is declining seven to 10 percent each decade," explained Ron Kwok of NASA's Jet Propulsion Laboratory in Pasadena, California. "Our study gives the first reliable estimates of how perennial ice replenishment varies each year at the end of summer. The amount of first-year ice that survives the summer directly influences how thick the ice cover will be at the start of the next melt season."

Using data from NASA's QuikScat satellite and other sources, data on six annual cycles of perennial Arctic ice were studied. The QuikScat satellite uses radar pulses reflected off Arctic ice to inventory both perennial and first-year ices. After the 2005 summer melt only about four percent of the 965,000 square miles of seasonal ice that formed the previous winter survived the summer and replenished the perennial ice cover. Additionally, weather conditions in the Arctic created winds that pushed ice out of the Arctic at an increased rate. Perennial ice that moves out of the Arctic in the summer time leaves open-water areas behind that absorb more heat than ice cover would and do not refreeze until summer's end. The increased heat absorption leads to additional thinning of the ice cover.

These findings suggest that the amount of first-year ice that survives is related to the number of freezing temperature days during the prior season. More freezing temperature days increases the ice thickness, making it more likely to survive the summer heating. "The winters and summers before fall 2005 were unusually warm," Kwok said. "The low replenishment seen in 2005 is potentially a cumulative effect of these trends."

When the six years of data were examined within the larger context of longer-term temperature records dating back to 1958, Kwok found a gradual warming trend in the first 30 years, which accelerated after the mid-1980s. Additionally, there was no data to indicate any reversal of the warming trend. "The record doesn't show any hint of recovery from these trends," he stated. "If the correlations between replenishment area and numbers of freezing and melting temperature days hold long-term, it's expected the perennial ice coverage will continue to decline."

This study suggests that on average the area of seasonal ice may no longer be large enough to sustain a stable perennial ice cover. Data from the 2005-2006 season have not yet been analyzed. The study appeared March 2 in Geophysical Research Letters.