The Arctic environment is in the midst of significant change (Serreze et al., 2000). The past several decades have seen pronounced warming in northern high latitudes, especially for winter and spring. Over land, there are indications of a shift from tundra to shrub vegetation, as well as regional warming of permafrost. The mass balance of small glaciers in the Arctic has been generally negative (Dyurgerov & Meier, 1997), paralleling a global tendency. Based on data from 1979-1999, Abdalati & Steffen (2001) show that the area of the Greenland ice sheet exhibiting summer melt has increased. It also seems that the lower coastal areas of the Greenland ice sheet thinned in the 1990s, with the coastal ice losses contributing to sea-level rise (Thomas, 2001). Of all the changes that have recently been observed in the Arctic, the decline in sea-ice cover stands out prominently. To complement the other chapters in this book it is useful to review the changes in Arctic sea ice.
Satellite passive microwave imagery has allowed for detailed assessments of Arctic sea-ice extent and concentration. Coverage from October 1978 through to 1987 is provided by the Nimbus-7 Scanning Multichannel Microwave Radiometer (SMMR) and since 1987 from the Defense Meteorological Satellite Program (DMSP) Special Sensor Microwave/Imager (SSMI). Figure 23.1 gives the time series of Arctic sea-ice extent through 2002 from the combined satellite record as normalized monthly anomalies. Sea-ice extent is defined as the region with an ice concentration of at least 15%. There is considerable month-to-month variability in normalized departures as well as evidence of a multiyear
1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004
Figure 23.1 Time series of Arctic monthly sea-ice extent anomalies, 12-month running anomalies and least-squares linear fit for the Arctic basin, based on data from 1979 to 2002. (Courtesy of National Snow and Ice Data Center, Boulder, CO.)
cycle. These features are in turn superimposed upon a statistically significant downward trend. Using data though 1997, Cavalieri et al. (1997) calculated a trend of 2.9 ± 0.4% per decade. Closer inspection of the record indicates that the downward tendency is driven primarily by sea-ice reductions in late summer and early autumn. Extreme minima were observed in September of 1990, 1993, 1995, 1998, 2002 and 2003. September of 2002 set a new record low for the passive microwave era (Serreze et al., 2003). Comparisons with earlier, albeit less reliable data, based on visible band satellite imagery, aircraft and ship reports, suggest that ice extent in September 2002 was the lowest in at least the past 50 years. Conditions were nearly as extreme in September 2003 (Plate 23.1). From satellite time series updated at the National Snow and Ice Data Center (NSIDC) in Boulder, Colorado (http://nsidc.org/data/seaice_index/), 2004 was another low sea-ice year. Records from ship reports and other sources provide some evidence of century-scale reductions over the North Atlantic (e.g. Walsh et al., 1999), but the quality of these data is open to question.
There is additional evidence of attendant thinning of the perennial ice pack in the central Arctic Ocean (Rothrock et al., 1999). This is the part of the ice cover that is present year-round, comprised of ice floes typically 3-5 m in thickness that have survived at least one melt season (known as multiyear ice). This contrasts with first-year ice, which forms in areas of open water during the autumn freeze up and winter. Through the use of upward-looking sonar, submarines can provide information on sea-ice draft—the fraction of the total ice thickness (about 90%) that projects below the water surface. Comparisons between sea-ice-draft data acquired during submarine cruises between 1993
and 1997 with earlier records (1958-1976) indicate that the mean ice draft at the end of the summer melt period decreased by 1.3 m in most of the deep-water portions of the Arctic Ocean. Questions have arisen regarding how much of this thinning represents the effects of melt/reduced ice growth, a wind-driven redistribution of thicker ice towards the coast of the Canadian Arctic Archipelago (Holloway & Sou, 2002), or a flux of ice out of the Arctic basin via Fram Strait (between Svalbard (Spitzbergen) and northern Greenland) (Rigor & Wallace, 2004). Although all of these processes may be at work, the latter seems quite important (see later discussion). An updated analysis by Rothrock et al. (2003), based on both submarine observations and models, gives further supporting evidence for thinning in the late 1980s through to 1997, with some indication of recent recovery.
How can we explain these ice losses? A number of studies (e.g. Maslanik et al., 1998; Serreze et al., 2003) have addressed some of the recent large anomalies (e.g. 1998, 2002) as case studies. These studies document the importance of anomalies in the atmospheric circulation that alter the distributions of surface air temperature (hence ice melt/growth) and the wind-driven circulation of the sea ice cover (e.g. promoting ice drift away from the coastlines). Thermodynamic and dynamic forcing tend to be closely intertwined. Although such studies have certainly been useful, a more basic framework is required.
One possible unifying framework is greenhouse-gas warming. A common feature of climate model projections is that, largely due to albedo feed-backs involving snow and sea ice, the effects of greenhouse gas loading will be observed first and will be most pronounced in the Arctic region (Holland & Bitz, 2003). Sea-ice loss is a near universal feature of these projections. Although the rate of decline varies widely between models, some predict a complete loss of summer ice cover by the year 2070. Given that the Arctic has warmed in recent decades, it is certainly tempting to view the observed ice losses as an emerging greenhouse signal. As reviewed below, however, the more obvious explanations involve changes in the atmospheric circulation. These changes account not only for much of the warming, but have led to important impacts on the sea-ice circulation. Might the circulation changes themselves be at least partly a response to changes in atmospheric trace gas composition? Although there is no firm consensus on this issue, there is growing evidence that this may be the case.
Was this article helpful?