Trends in Arctic wind speed and sea ice speed during 1992-2010: reanalysis, model, and observations

During the last 20 years substantial changes of the Arctic sea ice cover have been observed. Although the most pronounced changes are the reduction in summer sea ice extent and thickness, an increase in sea ice speed and deformation has also been inferred from surface buoys and drift stations. Sea ice movement depends on surface wind, ocean currents, and on the structure of the sea ice, e.g., strength, compactness, and drag. Here we study the trends in observed sea ice velocities obtained from satellite measurements and from a coupled ocean and sea ice model, in combination with trends in sea surface wind speed obtained from two atmospheric reanalyses.

The trend in 10-m surface wind speed between the years 1992 to 2009 of the Japanese ReAnalysis (JRA) is spatially heterogeneous. Positive trends are seen in the Western Beaufort Sea, the Chukchi Sea, the Eastern Barents Sea, and the Central Arctic. In other parts of the Arctic, however, negative trends can be observed, e.g., in the Greenland Sea, the Lincoln Sea, the Eastern Beaufort Sea, parts of the Laptev Sea, and the Western Barents Sea. This wind trend pattern is reflected in the spatial trend pattern of sea ice speed obtained from a coupled ocean and sea ice simulation driven by JRA atmospheric forcing. In the Central Arctic, including marginal seas on the Eurasian side of the Arctic, the spatial correlation coefficient between JRA wind speed trend and simulated sea ice speed trend is 0.54, compared to a correlation coefficient of 0.27 between surface temperature trend and sea ice speed trend. Although the pattern of sea ice speed trend correlates with the pattern of wind speed trend, the simulated sea ice speed trend is almost everywhere positive, including in regions where JRA wind speed has decreased. For example, simulated sea ice speed trend in the Greenland Sea, Baffin Bay, and parts of the Barents Sea has opposite sign from JRA wind speed. In these regions, other mechanisms, e.g., ocean forcing or changes in the ice structure, must play the dominant role in producing the simulated ice speed trends. Implications from observed and simulated trends in wind speed, sea ice drift, and sea ice thickness and concentration will be presented.