Recent Trends in Arctic Sea Ice and the Evolving Role of Atmospheric Circulation Forcing, 1979–2007 - Arctic Sea Ice Decline: Observations, Projections, Mechanisms, and Implications - page 15

the second level of red or blue shading in Plate 3) are statisti-
cally significant at the 95% level.
The pattern of winter SIC trends in the first half of the
record (199-1993) exhibits positive values in the Labrador
and Bering seas and negative values in the Greenland and
Barents seas and the Sea of Okhotsk. In contrast, winter SIC
trends in the second half of the record (1993-200) are nega-
tive throughout the marginal seas, with the largest declines
in the Atlantic sector. The winter SIC trends over the full
period of record (199-200) are also negative throughout
the marginal seas, except in the Bering Sea where the trends
are near zero. The change in pattern of winter SIC trends be-
tween the first and second halves of the record is notable and
will be discussed further below in the context of evolving
atmospheric circulation trends. The patterns of SIC trends
in the autumn season are very similar to those in winter,
with somewhat reduced magnitudes commensurate with the
lower long-term mean SIC amounts.
In summer, SIC trends in the first half of the record are
negative in the East Siberian Sea and positive in the Barents,
Kara, and eastern Beaufort seas, with the area of reduced
SIC outweighing that of increased SIC. In the second half of
the record (1993-2006), the area of negative SIC trends has
expanded to cover almost all longitudes. The summer SIC
trend over the full period of record (199-2006) is similar
to that for the second half of the record, with the largest de-
clines extending from the Laptev Sea eastward to the Beau-

fort Sea. The SIC trends in spring are a mixture of those in
winter and summer. In particular, the trends in the Atlantic
sector follow those in winter, while the trends in the central
Arctic Ocean resemble those in summer but with weaker
magnitudes. The hybrid nature of the spring SIC trends is
consistent with that of the long-term mean SIC distribution
discussed earlier.
It is instructive to relate the results shown in Plate 3 back
to the behavior of sea ice extent for the Arctic as a whole.
Figure 6 shows the linear trends in Arctic sea ice extent as
a function of time of year based on 5-day running means for
the two halves of the record separately (excluding data since
June 2007). Trend values significant at the 95% level are
shown with a bold line segment. Note that the magnitudes of
the trends in each period may be directly compared as they
are expressed in square kilometers per decade and percent of
the period mean per decade. The negative trends in Arctic
sea ice extent increased dramatically in magnitude between
the first and second halves of the record during October-
March, from values of 0 to −0.3 × 10 6 km 2 (0 to −2%) per
decade to values of −0.8 to −1.1 × 10 6 km 2 (−6 to −8%) per
decade. The weak and statistically insignificant trends in the
first half of the record are due to the large degree of cancel-
lation between negative and positive SIC trends in different
regions not to a lack of SIC trends (recall Plate 3). In terms
of actual magnitude (Figure 6, left), the seasonal dependence
of the trend amplitudes is nearly opposite between the two

Figure 6. As in Figure 4 but for the first (1979-1993, solid curve) and second (1993-2007, dashed curve) halves of the
record. Trend values significant at the 95% level are shown with a bold line segment.