The polar systems are extremely sensitive to the variability of temperature, and several aspects of these systems will be affected by any further climate change. The primary impacts will be on the physical environment, including ice, permafrost, and hydrology; on biota and ecosystems, including fisheries and terrestrial systems; and on human activities, including social and economic impacts on settlements, on resource extraction and transportation, and on existing infrastructure.
Sea ice covers about 11% of the world's ocean, depending on the season. It affects albedo, salinity, and ocean-atmosphere thermal exchange (IPCC 1996, WG II, Section 18.104.22.168). Studies on regional changes in the Arctic and Antarctic indicate trends of decadal length, often with plausible mechanisms proposed for periodicities of a decade or more (IPCC 1996, WG II, Section 22.214.171.124). At present, however, there is considerable year-to-year variation and no convincing evidence of long-term changes in the extent of global sea ice.
In Antarctica, temperatures are so low that comparatively little surface melting occurs on the continental ice sheet; ice loss is mainly by iceberg calving, the rates of which are determined by dynamic processes involving long response times (thousands of years). Even if Antarctica were to warm in the future, its mass balance is expected to become more positive: The rise in temperature would be insufficient to initiate melt but would increase snowfall (IPCC 1996, WG II, Section 7.4). Little change in Antarctic ice sheets is expected over the next 50 years, although longer-term behavior-including that of West Antarctic ice-remains uncertain, and some instability is possible. Some areas of Antarctica may show a pronounced change and dynamic response. The Antarctic Peninsula, for instance, receives 28% of the continent's snowfall and experiences warmer temperatures and summer melting at sea level. A rise in temperature would be expected to cause continued wasting of marginal ice shelves in the Antarctic Peninsula, but this melting has no direct effects on sea level, nor is it indicative of changes in the Antarctic ice sheets.
In the Antarctic, where the sea-ice cover is divergent and land boundaries are less important, it is more reasonable to suppose that the main effect of global warming will be a simple retreat of the ice edge southward. Even here, however, a complex set of feedback mechanisms comes into play when the air temperature changes. The balance of lead concentration, upper ocean structure, and pycnocline depth adjusts itself to minimize the impact of changes, tending to preserve an ice cover even though it may be thinner and more diffuse (IPCC 1996, WG II, Section 7.4.5). The modeling of Gordon and O'Farrell (1997) demonstrates an ultimate reduction in Antarctic sea ice of about 25% for a doubling of CO2, with the ice retreating fairly evenly around the continent.
A recent study by Nicholls (1997) discounts the significance of air temperature-induced melting of the more massive ice shelves south of the Antarctic Peninsula in a climate-warming scenario. Studies of the seasonal water temperature changes of the sub-ice-shelf cavity of the Antarctic Filchner-Ronne Ice Shelf indicate that the flux of high-salinity shelf water (HSSW) is responsible for the net melting at the ice shelf's base. As rates of sea-ice formation decrease during warmer winters, the flux of HSSW beneath the ice shelf is reduced. Subjected to less warm water flux, the sub-ice cavity will cool. Nicholls concludes that a moderate warming of the climate could, in fact, lead to a basal thickening of the Filchner-Ronne Ice Shelf, perhaps increasing its longevity.
With climate warming, ice cover in lakes and rivers is expected to decrease, with large changes in lake water levels. In the largest Antarctic desert, the McMurdo Dry Valleys, closed-basin lakes fed by glacier meltwater streams rose as much as 10 m from 1970 to 1990, or almost 480 mm per year. In the Antarctic Dry Valley, ice cover has thinned for some permanently ice-covered lakes. Lake Hoare thinned by 20 cm/yr over a 10-year period beginning in 1977. Because light attenuation by the ice is a major limiting factor, these climate-related changes are expected to cause shifts in the biota of such lakes (IPCC 1996, WG II, Section 10.6.1.4).
The Greenland ice sheet, which has no floating ice shelves of consequence, is different from the Antarctic ice sheets. Ice loss from surface melting and runoff is of the same order of magnitude as loss from iceberg calving. Thus, climate change in Greenland could have immediate effects on the surface mass balance of the ice sheet through melting and runoff as well as through accumulation. If there is warming, both the melt rates at the margins and the accumulation rates in the interior should increase. The former rate is expected to dominate. Thus, the mass balance could become negative (IPCC 1996, WG II, Section 7.4). Nevertheless, changes in the general form of the total ice sheet over the next century are expected to be small.
Increased temperatures in the Arctic are likely to shorten the duration of ice cover on Arctic lakes. A longer open-water period, together with warmer summer conditions, will increase evaporative loss. Some patchy wetlands and shallow lakes owe their existence to a positive water balance and the presence of an impermeable permafrost substrate that inhibits deep percolation; enhanced evaporation and ground thaw will cause some to disappear (IPCC 1996, WG II, Section 7.4.4). If there is decreased precipitation and lower flood frequency, the Mackenzie Delta in the Canadian arctic could shrink in several decades (IPCC 1996, WG II, Section 10.5.2), although the pre-Pleistocene depositions of gravel in the delta are sufficiently large that a more likely scenario is a profound change in the delta's form and extent.
A major source of uncertainty about sea-level change relates to the future behavior of the polar ice sheets, which hold most of the nonoceanic water on Earth's surface. Most of their volumes lie on land above sea level. However, much of the West Antarctic ice sheet is below sea level. The discovery of major recent changes in certain Antarctic ice streams has focused public attention on the possibility of "collapse" of this ice reservoir within the next century, with potential impacts on sea level (IPCC 1996, WG I, Section 126.96.36.199). the collapse of the West Antarctic ice sheet by the year 2100-with consequential major sea-level rise-is not impossible, but its likelihood is considered to be very low (IPCC 1996, WG I, Section 7.5.5).
Observed variations in sea-ice thickness are in accord with the predictions of numerical models that take account of ice dynamics and deformation as well as ice thermodynamics. The limited information available does not provide evidence of detectable change in the thickness of Antarctic sea ice (IPCC 1996, WG II, Section 7.2.5). Nevertheless, researchers have identified significant reductions in the summer sea-ice cover in the Bellingshausen and Admundsen Seas in the late 1980s and early 1990s that are consistent with a warming climate west of the Antarctic Peninsula (Jacobs and Comiso, 1993). Measurements from a series of submarine transects near the North Pole show large interannual variability in ice draft over the period 1979-1990. There is some evidence of a decline in mean thickness in the late 1980s relative to the late 1970s (IPCC 1996, WG II, Section 7.2.5). Between 1978 and 1994, the Arctic sea ice appears to have decreased by 5.5%.
Based on some global model scenarios for a doubling of CO2, a large change in the extent and thickness of sea ice is possible, not only from warming but also from changes in circulation patterns of both atmosphere and oceans. There is likely to be substantially less sea ice in the Arctic Ocean (IPCC 1996, WG II, Chapter 7 Executive Summary). Major areas that are now ice-bound throughout the year are likely to have long periods during which waters are open and navigable. Some models even predict an ice-free Arctic, although most scenarios maintain significant summer ice centered on the North Pole. Melting of snow and glaciers will lead to increased freshwater influx, changing the chemistry and salinity of oceanic areas affected by the runoff.
However, the ability of existing GCMs to predict the extent of Arctic ice change is limited by the inadequacy of regional polar models to simulate the multiscale dynamics of sea ice. GCM experiments with simplified treatments of sea-ice processes produce widely varying results and do not portray the extent of and seasonal changes in sea ice for the current climate very well. Given these limitations in existing ice modeling, some authors estimate that with a doubling of greenhouse gases, sea ice would cover only about 50% of its present area; others project a sea-ice reduction of 43% for the Southern Hemisphere and 33% for the Northern Hemisphere. The global area of sea ice is projected to shrink by up to 17 x 106 km2 (IPCC 1996, WG II, Section 7.4.5).
(continues on next page...)
Other reports in this collection