Jökull - 01.12.2007, Blaðsíða 68

Jónsdóttir and Sveinbjörnsson
monitor the migration of a large ice band across the
Denmark Strait and into Dýrafjörður, NW-Iceland. A
MODIS image from January 25th and an ENVISAT
image on the 26th (Figure 6) record the rapid migra-
tion of the sea ice under strong winds. During an ice
reconnaissance flight on January 28th it became clear
that the sea-ice band had been almost entirely pac-
ked into Dýrafjörður. It appears that the amount of
ice, packed into such a limited space was sufficient
to resist melting and create a microclimate within the
fjord, as sea ice was forming there while it melted fast
in nearby fjords. The next day, winds and tides bro-
ught the ice out of the fjord again where it soon melted
within the more spacious Denmark Strait.
November
Through the spring and summer, a long peninsula
of spread sea ice remained fairly close to Iceland (Fi-
gure 7). It was not until September 21st that no sea ice
was observed along our survey line (Figure 3). ENVI-
SAT images show that the sea ice lingered close to
Scoresbysund throughout the year, growing rapidly in
the early autumn (Polar View, 2007). A comparison
between the last six years at a similar date at the be-
ginning of winter shows that the ice extent was consi-
derably higher in 2007 than during the previous years
(Figure 8). The sea-ice extent was very low in 2002,
due to the mild weather conditions in the Greenland
Sea earlier that autumn.
On November 18th a blocking high pressure sys-
tem started developing SW of Iceland. After a short
spell of strong northeasterly direction, the wind tur-
ned to SW. The situation from January was repeated,
a low pressure system formed off East Greenland, de-
epening and increasing westerly windflow in Den-
mark Strait. Negative pressure difference between
Scoresbysund and Bolungarvík remained until No-
vember 23rd, but not as strong as in January (Figure
9). Subsequently, the wind turned northeasterly again
and the high pressure system disappeared. Earlier in
November, the pressure difference had also been ne-
gative, which no doubt explains why the ice edge was
so close to Iceland most of the month. During this
time, the sea ice was again driven over to Iceland and
came within 18 nm of the coast (Figure 10) before the
easterly winds drove the ice away.
CONCLUSIONS
Sea-ice conditions during 2007 have revealed that de-
creased amount of sea ice in the Northern Hemisphere
is no guarantee for ice-free waters around Iceland.
Blocking high pressure systems have brought the ice
to the coasts even if the amount of ice in the region
was relatively small. The most extreme cases, in Ja-
nuary and November 2007, were generated by unusu-
ally high pressure difference between Scoresbysund
and NW-Iceland with prevailing westerly winds un-
der which the sea ice reached the Western Fjords in a
few days.
As winter started, ice formation and drift south
with the East Greenland current began rapidly. ENVI-
SAT radar images show areas of sea-ice formation
very clearly as very dark regions (little backscatter),
and large patches have been observed on those ima-
ges in the Autumn. This interpretation has been con-
firmed during an ice reconnaissance flight.
Initial fluxes of sea ice into Icelandic waters usu-
ally consist of spread ice or long and narrow bands.
Open access to detailed satellite images, such as radar
images or medium resolution optical images, is there-
fore essential in order to detect and monitor such sea
ice in near real time as it can, even in small concentra-
tions, pose severe threats to passing ships as well as
near-shore fish farming.
For sea-ice historians, observations of sea-ice be-
haviour in 2007 are quite important, as they revealed
scenarios which would have been hard to understand
had they existed in earlier accounts, with a single fjord
being packed with ice, whilst the neighbouring fjords
are more or less ice free.
The satellite images collected in 2007 provide a
valuable dataset for studying ice drift and ice thermo-
dynamics in more detail, thus hopefully improving fu-
ture ice models.
Acknowledgements
The authors would like to thank the Icelandic Coast
Guard for co-operation on comparing satellite ima-
ges with ice observations on ice reconnaissance fli-
ghts, Dr. Þór Jakobsson, Eiríkur Sigurðsson, and Sig-
þrúður Ármannsdóttir at the Icelandic Meteorological
Office for supplying ice reports and information from
66 JÖKULL No. 57