Background
The seasonal ice zone (SIZ)
is defined generally as the area between the summer minimum and the
winter maximum extent of the polar ice pack. The SIZ is the crucial
region in which the polar atmosphere, ocean and sea-ice cover interact
with the bordering atmospheres and oceans. The exchange processes taking
place determine the advance and retreat of the sea-ice and profoundly
influence climate on regional and possibly global scales. These processes
also exert a significant effect on marine productivity, commercial
fisheries, off-shore activity and marine operations [e.g. Sandven and
Johannessen, 1990; Johannessen et al., 1992; Sandven and Johannessen,
1993; Sandven et al., 1999; Stenseth et al., 2002; Moritz et al., 2002].
In the Nordic seas – from the East Greenland coast through the
Fram Strait and the Barents Sea – the SIZ comprises a major limb
of the climate system, by connecting the warm, saline waters of Atlantic
origin with the cold, fresh polar waters and their atmospheric counterparts.
The strong heat transfer from ocean to atmosphere and the freshwater
transport from the Arctic to the North Atlantic both impinge on the
deep-water forming areas of the Greenland, Irminger and Labrador seas,
and therefore represent one of the driving mechanisms for the global
oceanic thermohaline circulation [Ganopolski and Rahmstorf, 2001; Alekseev
et al., 2001].
The ice edge can migrate hundreds of kilometres over a seasonal cycle, and thereby
produce large variations in maximum and minimum extent. In the past these variations
have been even more pronounced, e.g., 15.000 years ago most of the Norwegian
Sea was ice covered, indicating the response of the SIZ to the orbital forcing
[Koç et al., 1993, see front page]. Numerous studies have demonstrated
the connection between weather, climate and the state of the polar ice cover.
It is thus of fundamental importance to understand how seasonal, interannual
and decadal variations in ice extent are related to large-scale and mesoscale
circulations in the atmosphere (e.g. North Atlantic Oscillation (NAO), Arctic
Oscillation (AO), ENSO El Niño – Southern Oscillation (ENSO)) and
ocean [Dickson et al., 2000; Häkkinen and Geiger, 2000], including tropical-extratropical
teleconnections [e.g. Deser et al., 2000; Hurrell, 1996; Hoerling et al., 2001;
Peterson et al., 2002].
As an important area for biological production, the marine ecosystem of the SIZ
are also strongly affected by climate fluctuations [Stenseth et al 2002]. The
annual production during a warm year can be much higher than for a cold year
[Slagstad and Wassmann 1996]. This variability is transferred higher up in the
food-web, in which the capelin is a key species and a very important fish resource
for Norway [Gjøsæter, 1998]. Thus, a great challenge is to assess
the consequences for capelin of a future climate change.
A consensus result from 19 different climate system models [CMIP-2, 2002] is
that greenhouse induced global warming should be significantly enhanced in the
high northern latitudes [IPCC, 2001; Räisänen, 2001]. Furthermore,
simulations [Johannessen et al., 2002a] indicate a reduction of the polar ice
pack of 20% during winter and 80% during summer at the end of this century, including
a permanently ice free Barents Sea. Consequently, the ice edge will retreat dramatically
northwards, and the SIZ will be widened, in effect enlarging the Nordic Seas.
A range of potential consequences of an expanded SIZ and a shrinking ice cover
(see front page) can be hypothesized: 1. Reduced albedo and increased open water
would have significant effects on energy balance and atmospheric and oceanic
circulation in the high latitudes [Moritz et al., 2002] 2. Changes in the pathways
and spreading of melt water and stratification of the Nordic seas, and the patterns
and strength of deep-water formation in Greenland, Irminger and Labrador seas
and its impact on the global thermohaline circulation, thereby strongly altering
the climate on northern latitudes [Ganopolski and Rahmstorf, 2001]. 3. Exposure
of vast areas of the Arctic Ocean with cold open water, could become a new important
sink of atmospheric CO2 [Anderson and Kaltin, 2001], 4. Broad changes in the
marine ecosystem, perhaps degrading arctic and sub-arctic biodiversity [Beaugrand
et al., 2002], while at the same time opening new areas for potential fisheries,
as well as improving conditions for shipping and off-shore activities [Ragner
(Ed.), 2000].
From these concerns, we pose
the following three comprehensive questions:
1. To what degree can the changes in the SIZ and the Arctic climate system
be ascribed to natural and anthropogenic forcing?
2. To what degree may anthropogenic
forcing reduce, or even remove, the Arctic sea-ice cover in this century?
3. How will a widened SIZ
and a shrinking ice cover impact on the marine ecosystem?
Based on the above rationaleand
its interdisciplinary nature, nine Norwegian institutions (cf. Annex
2) covering the relevant fields of research have joined efforts and
defined the following objectives:
Objectives
The over-all objective is
to:
Explore, quantify and simulate past, present and future natural and anthropogenic
climate variability and changes, and the response of the marine ecosystem,
in the seasonal ice zone of the Greenland Sea, the Fram Strait, and the Barents
Sea
The five specific objectives
or research questions are:
1. Which marine climate processes
determine the seasonal ice zone variability and changes in the Greenland
Sea, the Fram Strait and the Barents Sea?
2. What is the response of the marine ecosystem to climate variability and
changes in the seasonal ice zone?
3. How well are climate processes that determine the seasonal ice zone variability
and changes represented and simulated in climate and ecosystem models?
4. How did the seasonal ice cover respond to past climate changes and to what
degree are recent climate changes in the seasonal ice zone of natural or anthropogenic
origin?
5. To what degree may the ice cover retreat or even disappear in this century,
and what is the impact on the marine ecosystem of the seasonal ice zone?
State of the art
The northern hemisphere climate has undergone major fluctuations over the last
century, e.g. the substantial warming between 1925 and 1945, the marked
cooling in 1950 – 1970 and the ongoing warming trend which started
around 1980 with a pronounced enhancement in the Arctic region, approximately
in accordance with climate model simulationss [e.g. Delworth and Knutson,
2000; Johannessen et al., 2002a; CMIP-2, 2002; Moritz et al., 2002]. Predictions
for the high Arctic is 3 – 4 °C for the next 50 years, about
twice the global average [IPCC, 2001]. Melting of sea ice [e.g. Johannessen
et al., 1995; Bjørgo et al., 1997; Johannessen et al., 1999; Parkinson
et al., 1999; Vinnikov et al., 1999; Hilmer and Lemke, 2000; Holloway and
Sou, 2002] and changes in the oceanic conditions are also observed [Alekseev
et al., 2001]. However, the dynamic–thermodynamic response of the
ice-cover is neither straightforward nor necessarily linear [Zhang et al.,
2000.], nor is the response of the atmosphere to sea-ice reductions [Deser
et al., 2000].
The recent warming has also coincided with a positive tendency in the NAO-index,
with its well-established link to European climate [e.g Hurrell, 1995; Thompson
and Wallace, 1998; Rodwell et al. 1999; Hilmer and Jung, 2000; Delworth and
Mann., 2000; Hoerling et al. 2001]. However, it remains open to debate whether
it is solely or partly due to intrinsic multi-decadal variability or anthropogenic
forcing.
A major mode of variability of the high northern latitudes is the AO [Thompson
and Wallace, 2000] (highly correlated with NAO [Itoh, 2002]), with considerable
impact on high-latitude precipitation, snow cover, hydrological cycles, sea
ice and upper-layer ocean circulation, although the physical mechanisms behind
it is largely unknown [Cohen et al., 1999; Cohen and Entekhabi, 2001; Cohen
et al , 2001 ; Gillett et al., 2002]. Other studies have linked snow cover
with other modes as the NAO or the Scandinavian pattern [Bojariu and Gimeno,
2002] emphasizing the impact of persisting spring and summer anomalies. However,
current explanations rely on statistical linkages that may be data dependent
(e.g. reanalyses and remotely sensed data), unambiguous mechanistic relations
remains to be explained.
By altering the planetary energy balance and affecting the oceanic overturning
circulation, positive feedback of the sea ice cover is the most likely mechanism
of translating modest changes in the forcing into high amplitude responses
in the past, such as: Initiation of glaciation due to orbital forcing [Imbrie
et al.1992], the post glacial climatic optimum 8000-6000 years BP and subsequent
neoglaciation in mid Holocene time 5-4000 years BP [e.g. Koç et al.,
1993; Koç and Jansen, 1994], and the onset of the Little Ice Age 1400-1500
AD [Lamb, 1977]. A major drawback is our poor capacity to quantify the sea
ice response, i.e. past sea ice variability, the position of the SIZ, and its
seasonal cycle. Improved methodology will provide the climate modeling community
with a unique possibility to study the role and response of the sea-ice feedbacks
under variable forcing conditions. Furthermore, time series of sea ice variations
extending beyond the instrumental observations will provide details on the
magnitude of the natural variability of the sea-ice cover, and on the response
to external forcing, where the forcing was more pronounced than recently.
For all emissions scenarios [IPCC, 2001], climate models predict a gradual
warming world wide, with maximum warming at high northern latitudes in winter.
There is also a spread between the different model realisations, also largest
in the high north, ranging from 1 to 10 ºC at 2xCO2 [http://www-pcmdi.llnl.gov/cmip,
Räisänen, 2001]. A first order cause of these differences can be
traced the way the models describe the SIZ and the arctic pack ice, and the
rapidity with which the ice cover retreats is a measure of the models’ climate
sensitivity. The non-trivial problem of correctly modelling the ice melt depends
on processes such as turbulent surface layer fluxes of heat, moisture and momentum,
radiative heat flux, in particular the radiative properties of clouds, atmospheric
dynamics on all scales, and the state of the sea ice-ocean system [Prisenberg
et al. 1997; Grønås and Skeie 1999; Lopez et al. 2000; Kvamstø et
al 2002]. Many of these processes are active on sub-grid scale, and are usually
described in a very simple manner in climate models. Small but systematic errors
in any of these processes may lead to erroneous response in climate scenario
integrations. Therefore, improved parameterisations of these processes are
a prerequisite for improving the climate models, particularly in the SIZ, and
thereby reducing the regional uncertainty associated with the predictions.
In the SIZ the shallow Barents Sea is an important area for biological production.
This is to a large extent governed by long term periodic, quasi-periodic and
non-periodic variability and trends in the Atlantic inflow [Ottersen et al.
2000], with corresponding impact on the water mass distribution especially
in its eastern part, where the polar front is weak (Johannessen and Foster,
1978; Ingvaldsen et al., in press), and on the winter maximum ice extent (Loeng
1991). This also explains the large variability in ice cover observed during
winter and in the primary production. Simulations show that the annual production
during a warm year can be about 30 % higher than a cold year for the basin
[Slagstad and Wassmann, 1996]
The melting and heating during spring and summer, create a pycnocline which
is hardly broken down even with strong wind events. Along the ice border, upwelling
may take place due to different wind stress on ice and open water [Johannessen
et al., 1983; Fennel and Johannessen, 1998; Lygre et al., 2002]. These different
physical processes create regimes with variability in annual primary production
from 10-20 gC m-2 in the northern Barents sea to more than 140 gC m-2 in the
central Barents Sea [Sakshaug and Slagstad, 1992]. Local variations were between
25 to 250 %. The variability of the primary and secondary production is transferred
higher up in the food-web, in which the capelin is a key species. Its migration
depends on the ice condition and the size of the capelin stock [Skjoldal et
al., 1992] A great challenge is to estimate the consequences for the capelin
stock of a climate change, e.g. whether the stock will move permanently northwards.
In summary, we want to advance our current knowledge and improve the predictability
of the SIZ climate and ecosystem by 1: Updating and improving data sets on
the ice cover, including remote sensing, surface air and sea temperatures (SAT,
SST) and heat budget, 2: Exploring the complex ecosystem interactions by data
and simulations, 3: Simulate and explain climate variability by validating
and applying climate models and identifying their shortcomings, 4. Deriving
data for the past and investigate pronounced events, in order to compare with
present conditions and future simulations, 5. Performing and analysing climate
scenario runs including the impact on the marine primary production and the
consequences for the marine ecosystems in the SIZ.
|
Links