16.1.3.1. The Arctic

Figure 16-2: Location map for the Antarctic. Elevations on Antarctic continent are indicated by dashed contour lines (500-m interval). Topography and ice shelf outlines are from U.S. Geological Survey and British Antarctic Survey data. Sea-ice extent data are from Couper (1989).

Figure 16-3: Time series of April sea-ice extent in Nordic Sea (1864-1998) given by 2-year running mean and second-order polynomial curves. Top: Nordic Sea; middle: eastern area; bottom: western area (after Vinje, 2000).

Instrumental observations of climatic parameters over the 20th century are available from standard climate stations on land and measurements taken on drifting ice floes in the Arctic Ocean. The land stations show that warming in the Arctic, as indicated by daily maximum and minimum temperatures, has been as great as in any other part of the world. Although not geographically uniform, the magnitude of the warming is about 5°C per century, with areas of cooling in eastern Canada, the north Atlantic, and Greenland (Koerner and Lundgaard, 1996; Borzenkova, 1999a,b; Jones et al., 1999; Serreze et al., 2000). Data from ice floe measurements show a slight warming on an annual basis, with statistically significant warming in May and June over the 1961-1990 period. Air temperature anomalies in the Arctic basin have been strongly positive since 1993. In the period 1987-1997, air temperature in the Arctic increased by 0.9°C (Alexandrov and Maistrova, 1998).

Significant warming from the beginning of the 20th century has been confirmed by many different proxy measurements (Maslanik et al., 1996; Bjorgo et al., 1997; Smith, 1998). Magnuson et al. (2000) found consistent evidence of later freeze-up (5.8 days per 100 years) and earlier breakup (6.5 days per 100 years) of ice on lakes and rivers around the northern hemisphere from 1846 to 1995 and an increase since 1950 in interannual variability of both dates. Glaciers and ice caps in the Arctic also have shown retreat in low-lying areas since about 1920. Numerous small, low-altitude glaciers and perennial snow patches have disappeared. However, no increasing melting trend has been observed during the past 40 years (Jania and Hagen, 1996; Koerner and Lundegaard, 1996; Dowdeswell et al., 1997). Glaciers in Alaska have receded, with typical ice-thickness decreases of 10 m over the past 40 years, but some glaciers have thickened in their upper regions (BESIS, 1997). Greenland's ice sheet has thinned dramatically around its southern and eastern margins. Above 2,000-m elevation, the ice sheet is in balance, on average. The net effect is a loss of about 51 km³ yr^-1. (Krabill et al., 1999, 2000).

Snow-cover extent in the northern hemisphere has been reduced since 1972 by about 10%, largely as a result of spring and summer deficits since the mid-1980s (Brown, 2000; Serreze et al., 2000). Most Arctic regions have experienced increases in precipitation since at least the 1950s (Groisman et al., 1991; Groisman and Easterling, 1994; Georgiyevskii, 1998). Measurements from Spitsbergen show a statistically significant increase in precipitation during all seasons, except winter (Hanssen-Bauer and Forland, 1998).

Groisman et al. (1994) analyzed seasonal snow extent in the northern hemisphere and demonstrated an inverse relationship with near-surface air temperature. Recent findings have provided evidence of a significant decrease in spring snow extent since 1915 over Eurasia (Brown, 1997) and southern Canada (Brown and Goodison, 1996). Such trends may be related to low-frequency fluctuations in hemispheric atmospheric circulation patterns (Serreze et al., 2000).

Arctic sea-ice extent decreased by approximately 3% per decade between 1978 and 1996 (Cavalieri et al., 1997; Parkinson et al., 1999; Johannessen et al., 1999; Serreze et al., 2000). The most significant contractions were detected in 1990, 1993, and 1995 (Maslanik et al., 1996). Ice composition also has changed, with a reduction in the area of multi-year ice in winter. Summer sea-ice extent has shrunk by 20% (880,000 km²) over the past 30 years in the Atlantic part of the Arctic Ocean (Johannessen et al., 1999), but the shrinkage has only been 5% in the Canadian Arctic Sea. Sea-ice extent in the Bering Sea experienced a dramatic reduction when a regime shift occurred in 1976 and has continued to decrease (BESIS, 1997). New compilations of Arctic sea-ice extent, using historical data from the past 135 years (Vinje, 2001), show that overall sea-ice extent in April has been reduced in the Nordic seas by 0.79 x 106 km² (33%) (see Figure 16-3). Nearly half of this reduction took place between 1860 and 1900. Although there are large interannual and seasonal variations in sea-ice extent, the reduction is greatest in spring. This is consistent with the 1912-1996 temperature record from Svalbard, which shows significant warming (3°C) in spring (Hanssen-Bauer and Forland, 1998). There is an approximate 10-year climate signal in the Arctic and subarctic, with a clockwise propagating signal in sea-ice concentration anomalies and a standing oscillation in sea-level pressure anomalies—the latter linked to the two phases of the North Atlantic Oscillation (NAO) (Mysak and Venegas, 1998). Comparison of these trends with outputs from climate models (forced by observed GHGs and tropospheric sulfate aerosols) reveals that the observed decrease in northern hemisphere sea-ice extent agrees with transient simulations (see Figure 16-6) and that both trends are much larger than would be expected from natural climate variations (Vinnikov et al., 1999).

The most effective method for determining the thickness of sea ice is to use sonar, directed upward at the floating ice from submarines or moorings (Melling et al., 1995). The results indicate great variability. Average thinning of sea ice has been observed since 1976 on some transects of the Atlantic and Arctic Oceans (Wadhams, 1997). A thinning by about 0.13 m over the period 1970-1992,with the maximum decrease detected in the eastern Siberian Sea, was found by Russian scientists (Nagurduy, 1995). Rothrock et al. (1999) found that ice draft at the end of the melt season has decreased by about 1.3 m in most of the deepwater portion of the Arctic Ocean, from 3.1 m in 1958-1976 to 1.8 m in the 1990s (~15% per decade). The decrease is greater in the central and eastern Arctic than in the Beaufort and Chukchi seas (Rothrock et al., 1999). Vinje et al. (1998), however, claim that no significant change in ice thickness can be observed in Fram Strait. Their conclusion is in agreement with submarine observations from 1960-1982. Although there is large variability, ice cover has continued to become thinner and has decreased by 40% over the past 3 decades (Rothrock et al., 1999). Analysis of the duration of summer melt over a large fraction of the perennial Arctic sea ice from 1979 to 1996 reveals an increase of 5.3 days (8%) per decade in the number of melt days each summer (Smith, 1998).

Continues on next page

IPCC Homepage