EDIT
Geophysical Institute researchers Martin Jeffries and Kim Morris began investigating the Tanana River ice break up record along with temperature data in 2004. They found links between the Nenana Ice Classic data and changes in the local and regional climate. "The correlations we are seeing between break up on the Tanana River and Arctic climate adds to our confidence that river break up is a good indicator of climate variation," Jeffries said. Jeffries recently presented his findings at the Arctic Science Conference.
The Nenana Ice Classic began in the spring 1917 when Alaska Railroad workers surveying to build a waterfront dock bet on the exact date and time the ice on the Tanana River would break up. Over the past 89 years, the contest has grown into an annual gaming event, the proceeds from which support local charities. Each spring a 26-foot, five-legged tripod is built on the frozen Tanana River. A cable connects the tripod to a clock tower and when the ice moves for spring break up, the cable trips the clock, indicating the exact time of spring break up.
Jeffries, a research professor at the Geophysical Institute, who has never participated in the contest, began investigating the Nenana Ice Classic as a potential indicator of Arctic climate change after reading an article in the journal Science, which highlighted the ice classic contest and a trend toward early break up dates. In the article, two California scientists discussed the possibility of a climate signal in the ice break up data. As an Alaska resident for 21 years, Jeffries wanted to take a closer look at the ice break up data, "to see if we can tease more climate information out of the Ice Classic."
Their research showed the Tanana River break up data matched with periods of prolonged warm or cool air. The findings prove the Nenana Ice Classic is a good proxy for climate in interior Alaska, according to the study. The most significant determinant of variation in the ice break up data was the Fairbanks average air temperature. Overall, the ice classic data are very sensitive to large scale temperature change, one factor that makes the Nenana results a good indication of climate variation and change in Alaska.
EDIT
http://www.juneauempire.com/stories/102306/sta_20061023023.shtml