A statistical linear regression model is used to forecast
end of summer ice
conditions in the Beaufort Sea with several months lead
time. The model retains six sea ice and atmospheric parameters, where
relatively more (less) multiyear ice
concentration, fewer (more) heating degree days, lower
(higher) North Atlantic
Oscillation index values, higher (lower) Tropical-Northern
Hemisphere index values, higher (lower) January Pacific-Decadal Oscillation
index values, and lower (higher) June Arctic Oscillation index values are
associated with lighter (heavier) sea ice conditions. Cross validation
diagnostics indicate that variations in these parameters are related to
91% of the variation in sea ice conditions, suggesting a relatively simple
ice-atmosphere statistical model can be used to forecast end of summer
ice conditions in the Beaufort Sea.
Submitted to Geophysical Research Letters
Dr.
Sheldon D. Drobot and Dr.
James A. Maslanik