Known as the "Scan Eagle," the unmanned aircraft was launched in May and June of 2009 from the NOAA vessel McArthur II over the Bering Sea west of Alaska. The drone has a 10-foot wingspan and is owned and operated by the University of Alaska.
The image recognition software was developed by Boulder Labs Inc. in Boulder, Colo., and used to automate the identification of seals in 27,000 images that were collected during the flights. "The results show that the seals have distinct preferences for specific types of ice, demonstrating that ice extent is not the only factor affecting seal populations," said Weatherhead.
The Scan Eagle flights lasted from two to eight hours and flew at altitudes ranging from 300 to 1,000 feet. While the amount of ocean and ice scanned by the unmanned aircraft was small -- it flew 3- to 5-mile-long transects over the Bering Sea -- the researchers were eager to see whether the image recognition system would work for characterizing both the ice and the seals. "The answer was a resounding yes," Weatherhead said.
The analysis of sea ice by the team included edge-to-area calculations of ice floes as well as ice floe size and distribution. "There is an incredible variety of ice and we are trying to come up with mathematical ways to describe it," she said. "One thing that really interests us is how broken up the ice is in particular areas."
According to CU-Boulder's National Snow and Ice Data Center, the total loss of Arctic sea ice extent from 1979 to 2009 was an area larger than the state of Alaska. Scientists there believe the Arctic may become ice-free during the summers within the next several decades.
In December 2010, NOAA's Fisheries Service proposed to list the Arctic ringed seal as threatened under the Endangered Species Act because of diminishing sea ice and snow cover. Arctic ringed seals do not com
|Contact: Elizabeth Weatherhead|
University of Colorado at Boulder