Arctic sea ice variability in the context of recent atmospheric circulation trends

Forty years (1958-97) of reanalysis products and corresponding sea ice conc entration data are used to document Arctic sea ice variability and its asso ciation with surface air temperature (SAT) and sea level pressure (SLP) thr oughout the Northern Hemisphere extratropics. The dominant mode of winter ( January-March) sea ice variability exhibits out-of-phase fluctuations betwe en the western and eastern North Atlantic, together with a weaker dipole in the North Pacific. The time series of this mode has a high winter-to-winte r autocorrelation (0.69) and is dominated by decadal-scale variations and a longer-term trend of diminishing ice cover east of Greenland and increasin g ice cover west of Greenland. Associated with the dominant pattern of winter sea ice variability are larg e-scale changes in SAT and SLP that closely resemble the North Atlantic osc illation. The associated SAT and surface sensible and latent heat flux anom alies are largest over the portions of the marginal sea ice zone in which t he trends of ice coverage have been greatest, although the well-documented warming of the northern continental regions is also apparent. The temporal and spatial relationships between the SLP and ice anomaly fields are consis tent with the notion that atmospheric circulation anomalies force the sea i ce variations. However, there appears to be a local response of the atmosph eric circulation to the changing sea ice cover east of Greenland. Specifica lly, cyclone frequencies have increased and mean SLPs have decreased over t he retracted ice margin in the Greenland Sea, and these changes differ from those associated directly with the North Atlantic oscillation. The dominant mode of sea ice variability in summer (July-September) is more spatially uniform than that in winter. Summer ice extent for the Arctic as a whole has exhibited a nearly monotonic decline (-4% decade(-1)) during t he past 40 yr. Summer sea ice variations appear to be initiated by atmosphe ric circulation anomalies over the high Arctic in late spring. Positive ice -albedo feedback may account for the relatively long delay (2-3 months) bet ween the time of atmospheric forcing and the maximum ice response, and it m ay have served to amplify the summer ice retreat.