3
Chapter 2
Long-term Change in the Arctic
shifts in temperature, perhaps of only a degree or two,
2.1. The distant past, and recorded history
account for the so-called Medieval Warm Period (1100 -
During the last 400 000 years, the Earth has experienced
1400 AD) and subsequent Little Ice Age (1450 -1850 AD)
four ice ages which have left records in glacial ice accu-
(for the relevance of these terms see Bradley and Jones,
mulating in Antarctica (Petit et al., 1999) and in Green-
1993; Crowley and Lowery, 2000). Both of these minor
land (Dansgaard et al., 1993; Sowers and Bender, 1995).
and sporadic deviations in the temperature record had
The overall surface air-temperature change between gla-
dramatic consequences for humans especially those liv-
cial and interglacial periods is thought to have been
ing on the margins of northern oceans (Alley et al.,
about 12°C, but perhaps more significant than tempera-
2002; McGhee, 1996; Ogilvie and Junsson, 2000). Dur-
ture were the accompanying changes in continental ice
ing the past two centuries, small changes in ice and wa-
masses, sea-ice climate and global ecosystems. In partic-
termass distribution have continued to have an impact
ular, sea-ice cover has proven to be a master variable in
on humans and ecosystems, sometimes leading to migra-
the equation of change. During the last glacial maxi-
tion or abandonment of locations, but certainly requir-
mum, sea ice was locked within the Arctic and seasonal
ing adaptation (Miller et al., 2001; Vibe, 1967).
or perennial sea ice extended well south into the North
For most of the past 10 000 years (the Holocene), cli-
Atlantic Ocean (Darby et al., 1997; de Vernal et al.,
mate change was not accompanied by the added com-
1993). The change from glacial to de-glacial to inter-
plexity of anthropogenic pollutants. However, over the
glacial can be seen widely in Arctic sediments, both in
past two millennia and especially during the past two
terms of sedimentation rate and in the amounts and
centuries, Arctic glacial ice has recorded the transient
sources of organic material derived from marine primary
rise in the levels of virtually every contaminant emitted
production or land vegetation (see for example, Darby
by human activities (Boutron et al., 1995, 1998; Gregor
et al., 2001; Nørgaard-Petersen et al., 1998; Phillips and
et al., 1995; Hong et al., 1994; Masclet and Hoyau,
Grantz, 1997; Stein et al., 1994, 2001).
1994; Rosman et al., 1997). These include the green-
Sea level dropped by about 120 m during the last
house gasses (GHGs) that force atmospheric tempera-
glacial maximum (Fairbanks, 1989). This exposed much
ture change (Petit et al., 1999), and it is the dramatic rise
of the Arctic Ocean's enormous continental shelves, forc-
in the levels of GHGs during the past several decades
ing rivers to cut channels across them to enter the interior
that make future projections based on past climates sub-
sea directly, and severing the connection between the Arc-
ject to such uncertainty.
tic and Pacific Oceans. With sea-level rise, about 15 000
years ago the Bering land bridge was flooded (Hopkins,
2.2. The present and future
1979) and then gradually submerged (Dyke et al., 1996b)
allowing the Pacific Ocean access to the Arctic Ocean.
The twentieth century has been the warmest in the Arc-
This sequence of events together with inundation of the
tic for the past 400 years (Overpeck et al., 1997). The
continental shelves must have had enormous consequen-
Intergovernmental Panel on Climate Change (IPCC)
ces for the oceanography and regional biogeography of the
suggests that over the past century the global mean sur-
western Arctic and the Canadian Arctic Archipelago (Dun-
face temperature has increased by about 0.3 to 0.6°C,
ton, 1992; Dyke et al., 1996a,b; Hequette et al., 1995).
mostly attributable to human activities, and will proba-
Although the climate has been described as `excep-
bly further increase by 1.4 to 5.8°C between 1990 and
tionally stable' during the past 10 000 years (Dansgaard
2100 (Houghton et al., 1995; IPCC, 1995, 2002; Show-
et al., 1993), it has actually continued to undergo sub-
stack, 2001). According to models, warming will be
stantial fluctuations. Indeed, it seems that very small
more pronounced in polar regions (Figure 2·1); perhaps
Temperature change, °C
1.5
1.5
0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
1.5
0.5
0
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
Figure 2·1. Predicted change in surface
0.5
air temperature for 2020-2030 relative
0.5
0.5
to 1990-2000 (from the CGCM2 (Can-
0.5
adian Global Circulation Model 2)),
0.5
0.5
courtesy of the Canadian Centre for Cli-
0.5
0.5
mate Modeling and Analysis). Global
warming is expected to have an uneven
1.5
0.5
distribution with the Arctic experienc-
1.5
ing the highest projected warming.
4
AMAP Assessment 2002: The Influence of Global Change on Contaminant Pathways
Actual sea-ice extent, northern hemisphere,
and mean transit times by perhaps 10% (Holzer and Boer,
million km2
2001). Furthermore, the greatest warming will occur in
16
the autumnwinter period due to delay in the onset of
sea-ice cover (Manabe et al., 1992; Serreze et al., 2000).
Winter (J,F,M)
14
Continental interiors will become dryer and sea level
Spring (A,M,J)
will continue to rise, perhaps by a further 50 cm in ad-
dition to the estimated rise of 10 to 25 cm during the
12
Annual
past century (Proshutinsky et al., 2001; Serreze et al.,
Autumn (O,N,D)
2000).
10
Models predict that after about 80 years of atmos-
pheric CO2 increase at 1% per year, precipitation will in-
Summer (J,A,S)
crease within the Arctic and subpolar regions to 0.5 to 1
8
m/yr, which would more than double the current mois-
a
ture flux convergence north of 70°N estimated at about
6
20 cm/yr (Manabe et al., 1992; Walsh, 2000), making
1900
1950
2000
the Arctic a considerably `wetter' place. Over the past
Projected sea-ice extent, northern hemisphere,
four decades sea-ice extent in the Arctic Ocean has de-
million km2
creased in summer by as much as 25%. By the end of the
12
twenty-first century, GHG forcing might produce an
Arctic Ocean seasonally clear of ice (Figure 2·2, Flato
and Boer, 2001).
10
Simulations based on GHG forcing predict that
mean annual river discharge will increase by about 20%
8
for the Yenisey, Lena and Mackenzie Rivers, but de-
CGCM2
crease by 12% for the Ob (Arora and Boer, 2001; Miller
6
and Russell, 1992). Furthermore, the projection that
high-latitude rivers will undergo marked changes in am-
plitude and seasonality of flow due to decreased snow-
4
fall and earlier spring melt (Arora and Boer, 2001) may
CGCM1
already have some support in observations (Lammers
2
et al., 2001).
b
The coupling of the runoff cycle with northern lake
hydrology is probably one of the points most sensitive
0
1900
1950
2000
2050
2100
to climate change (see for example, Vörösmarty et al.,
2001) but the understanding of the processes involved is
Figure 2·2. Sea-ice extent in the Northern Hemisphere. This figure
not yet sufficient to make confident projections. If Arctic
illustrates a) time series of `actual' annual and seasonal sea-ice ex-
lakes become more `temperate' in character, productivity
tent between 1990 and 2000 derived from long-term observations
and satellite images (adapted from Walsh and Chapman, 2000)
is likely to be enhanced due to less ice cover and more
and b) simulations of annual mean sea-ice extent from the models
mixing, and there will be greater opportunity for runoff
CGCM1 and CGCM2 (Canadian Global Circulation Models 1 and
to mix into the lake during freshet, further supporting a
2 of the Canadian Centre for Climate Modeling and Analysis),
more vigorous aquatic food web.
where the latter differs from the former in mixing parameterization.
With these primary changes, permafrost melting can
(After Flato and Boer, 2001.)
be expected to accelerate, disrupting vegetation and en-
5°C or more near the pole and 2 to 3°C around the mar-
hancing nutrient, organic carbon and sediment loading
gins of the Arctic Ocean, with a decreasing temperature
of rivers and lakes (Vörösmarty et al., 2001). The loss of
contrast between poles and the equator (Manabe et al.,
sea ice in the marginal seas, together with sea-level rise
1992; Mitchell et al., 1995; Zwiers, 2002). For gases
will promote further erosion of poorly bonded, low-gra-
emitted to the atmosphere, climate warming will in-
dient coasts, particularly during the period of autumn
crease inter-hemispheric exchange times, mixing times,
storms.