Sea

Sea ice is frozen sea water and in the Antarctic is mostly mobile pack ice less than one year old. When the sea surface begins to freeze, from about mid February on, small needles or platelets of ice appear, growing and agglomerating to form a slush known as grease ice. As it thickens, this slush coagulates and breaks up under the influence of waves and wind into intermediate soft and hard plates which collide with each other and may form pancake ice. Still thickening, these alternately freeze together and break up forming even larger, still thickening slabs of ice called floes. Colliding floes may raft over one another when still thin or form pressure ridges of ice rubble along the colliding margin, particularly in shear zones near the coast. The whole accumulation of broken sea ice is termed pack ice and collectively referred to as the ice pack (or simply pack).

In most places, Antarctic pack is not constrained by land and so divergent motion is common. Floes move under the influence of wind, storms and currents. Cracks and linear openings (leads) formed between (or through) floes, and nonlinear openings (polynyas) are sites of rapid formation of new ice during autumn, winter and spring. The areal ratio of ice to water is greatest during winter (when ice concentration is mostly nine-tenths or more) and least in late summer. Other characteristics of Antarctic pack ice include an extensive snow cover, a high proportion of fine-grained spicule and plate crystals, a frequent underlayer of unconsolidated ice platelets, algae growth within the ice and a general lack of surface melt features (Lewis and Weeks, 1970; Gow et al., 1982; Jacka et al., 1987). These characteristics are in large part due to the dynamic and climatic regimes producing divergent ice motion, lead formation, ice formation and snowfall.

In the Pacific Sector (160°E to 60°W), the pack ice reaches a maximum extent averaging 6.3 million km2 by late September (U.S. Navy, 1985), one-third of the total Antarctic ice pack extent. By the end of the growth season, undeformed sea ice has thickened to about 3 m at 77°S (Budd, 1981). Hoe size and thickness probably vary greatly at all locations but generally increase from north to south (Gilbert and Erickson, 1977; Jacka et al., 1987). Thickness averaged over time and space perhaps 0.5 to 1.5 m (Budd, 1986b) because of continual formation of leads, new ice and northwards advection of ice. From October to February, the pack decays by breakage, dispersal and melting to a minimum extent of about 1.5 million km2 in the Pacific Sector. Most of the ice is therefore less than one year old (first-year ice).

Other forms of sea ice also occur in the Pacific Sector. Fast ice (ice in a sheet attached to the shore) develops along most of the coastline during winter, especially in coastal indentations sheltered from the wind and waves. The fast ice belt extends up to 150 km offshore near Russkaya Station (74°50'S, 137°E) (Romanov, 1984). Most of this fast ice has broken up and dispersed by February, except in some coastal indentations some years. Second-year (or multi-year) pack ice is found in areas of pack which survive the summer melting (Fig. 4.7). It therefore tends to be thicker than first-year ice and takes longer to melt. An ice-foot consisting of frozen sea water with blocks of sea ice and glacial ice, or ice-push ridges of heaped up sea ice, develop along rocky shores throughout the year but most invariably melt in summer.

Antarctic sea ice has a significant effect on the climate, weather, seas, and ecosystems of the region and human activities there. The seasonal sea ice cover varying so much between light coloured ice and dark ocean or thin ice modifies the surface albedo and energy exchange affecting the earth's heat balance in the Antarctic region. Sea ice acts as an insulator between the cold polar atmosphere and the relatively warm ocean, so moderating the exchange of heat and moisture between them, and hence affecting their characteristics, particularly of the upper layer of the ocean and deep oceanic circulation. Snow-covered pack ice reflects most of the incoming solar radiation, amplifying the cooling effect of the region as well as delaying and limiting the summer warming of the continent and its surrounding seas. The generation and paths of storms in the Southern Ocean are related to the ice pack extent (see Mullan and Hickman, this volume). Sea ice formation increases the salinity of the sea water beneath causing vertical mixing in the water column. In particular, polynyas and leads in the pack are thought to contribute to the high salinity Antarctic Bottom Water (Zwally et al., 1985) found throughout the world's oceans. Melting in summer dilutes and can increase the stability of the upper part of the water column increasing phytoplankton activity there (Ainley and Jacobs, 1981; Smith and Nelson, 1985; Smith, 1987). The ice algae that grow in sea ice are believed to be an important source of food for secondary consumers as indicated by increased productivity at the pack ice edge and the ecosystem characterstics (Ainley et al., 1986). Polynyas in the pack are thought to be important biologically as they may contribute to ecosystem productivity and provide winter refuges. Pack ice effectively limits most Antarctic shipping to less than four months of the year from mid November and has recently caused two ships to sink in the Pacific Sector (Table 4.1).

This section outlines the distribution and movement of sea ice in the Pacific Sector. Further information and reviews of Antarctic sea ice are given by Lewis and Weeks (1970), Ackley (1981), Zwally et al. (1983a, b), Romanov (1984), Keys (1984), U.S. Navy (1985), Sturman and Anderson (1986) and Lewis (1987). Most of our knowledge of Antarctic ice pack has been derived using remote sensing techniques especially satellite-borne sensors. These may underestimate the amounts of new or young ice present (Jacka et al., 1987).

Different methods of interpretation and analysis have produced different results. Our knowledge is therefore still incomplete especially since satellite observations of the whole pack have been made only over the last 15 or so years and not all sea ice parameters can be measured in this way. In addition, the distribution of pack ice is complex and variable due to a variety of dynamic, thermal, climatic and oceanographic factors.

The main feature of pack ice distribution in the Pacific Sector, like elsewhere in the Southern Ocean, is the large seasonal change in ice extent (area covered by pack ice) between summer and winter. This change is particularly evident in the Ross Sea. Fig. 4.7 shows the mean extent of the pack in late September and in mid Feburary, the latter time being close to the end of the summer melt season. Four areas of pack normally survive the summer. Although the two relatively small areas in the western Ross Sea have completely dissipated some summers (Fig. 4.7), persistent cores of pack ice survive mainly in the Amundsen and Bellingshausen Seas but also off the Oates Coast (Zwally et al., 1983a; U.S. Navy, 1985). Extensive freezing occurs from the south from late Feburary-March due to rapidly cooling air and sea temperatures. Together with ice advection driven by wind and currents (see below), this freezing leads to rapid areal increase in sea ice cover from early April to late July (Ackley, 1981; Zwally et al., 1983a). Thereafter, the pack continues to expand more slowly until September-October (Figs 4.7; 4.8). The rapid decrease in ice cover from mid November to January is thought to be due to upwelling of relatively warm deep water, for example, along the Antarctic Divergence, and to exchange of heat between sea and air (Gordon, 1981).

Two other important and related features of pack ice distribution are the

Fig. 4.7. Sea ice in the Pacific Sector adapted from U.S. Navy (1985) showing ice extents in mid September (extreme maximum and mean maximum), and mid February (mean minimum and extreme minimum) averaged over the years 1973 to 1982. Recurring or intermittent polynyas (areas of relatively thin pack ice with a higher incidence of open water) within the ice pack and along the coast are also shown, after Romanov (1984) and Zwally et al. (1985). Other small polynyas probably occur at the edge of the fast ice similar to the polynya in McMurdo Sound.

Fig. 4.7. Sea ice in the Pacific Sector adapted from U.S. Navy (1985) showing ice extents in mid September (extreme maximum and mean maximum), and mid February (mean minimum and extreme minimum) averaged over the years 1973 to 1982. Recurring or intermittent polynyas (areas of relatively thin pack ice with a higher incidence of open water) within the ice pack and along the coast are also shown, after Romanov (1984) and Zwally et al. (1985). Other small polynyas probably occur at the edge of the fast ice similar to the polynya in McMurdo Sound.

differences in area from year to year of pack ice extent, and polynyas (areas of open water or reduced ice concentrations within the ice pack). Variability is most pronounced in the summer months (Fig. 4.8) when the area of the ice pack at any given time may vary by up to 50% between years (Zwally et al., 1983a; U.S. Navy, 1985). The Ross Sea pack is especially variable. Some summers are heavy ice seasons when the pack is very slow to dissipate (e.g., 1975-76,1976-77, 1977-78) and others have extensive and long (e.g., three months) open-water seasons (e.g., 1978-79,1979-80) (Zwally et al., 1983b; Keys, 1984). Jacka (1983), Romanov (1984) and U.S. Navy (1985) also portray the pack ice distribution and the variability of the present-day ice edge.

The areal extent of sea ice also varies over longer periods but the most reliable records date from 1973. Year-to-year variability in the Ross Sea sector is too great to show any significant trends over this period but both maximum winter and minimum summer yearly sea ice extents in the eastern Pacific Sector (130°W to 60°W) have increased and then decreased slightly since 1973 (Fig. 4.8; U.S. Navy,

1985). However, total Antarctic pack ice extent at the winter maximum decreased and then increased over the same period (Zwally et al., 1983b; Jacka, 1983; U.S. Navy, 1985). There is therefore no evidence for any recent decrease in sea ice extent due to elevated carbon dioxide levels in the atmosphere or higher sea temperatures. Antarctic pack ice may be quite sensitive to such changes (Parkinson and Bindschadler, 1984) but geologic evidence relating to ice-rafted debris suggests that current models which predict an early removal of Antarctic pack ice due to greenhouse warming should be viewed with caution (Robin,

1986). During the 1930s, however, summer ice conditions were heavier than during present summers (Kukla and Gavin, 1981) and pack ice extended further north (Doake, 1985). In addition, sea ice extent in winter-spring 18,000 years ago was probably double the present day winter maximum extent (Burckle et al., 1982). Any local influence that human activity might have on ice extent, duration

Fig. 4.8. The extent of sea ice, mainly pack ice, in the Pacific Sector as interpreted from visual, thermal, infrared and passive microwave satellite imagery (after U.S. Navy, 1985). These weekly observations show the marked seasonal change in ice extent between late summer and winter and indicate the interannual variability between January 1973 and December 1982.

Fig. 4.8. The extent of sea ice, mainly pack ice, in the Pacific Sector as interpreted from visual, thermal, infrared and passive microwave satellite imagery (after U.S. Navy, 1985). These weekly observations show the marked seasonal change in ice extent between late summer and winter and indicate the interannual variability between January 1973 and December 1982.

or albedo is likely to be outweighed by such seasonal and interannual variability of the highly mobile pack.

The development of polynyas strongly influences pack ice distribution and dissipation including the length of the open-water season. At least nine intermittent or recurring polynyas have been recognized in the Pacific Sector (Fig. 4.7). These areas of reduced ice concentration (and thickness) or open water are thought to be caused by a combination of synoptic (as produced in a regional weather event or events) and katabatic winds acting in concert with suitable topography of land, coast and seafloor (Kurtz and Bromwich, 1985; Zwally et al., 1985). Probably some vertical component of water motion is also involved.

The stable, recurring polynya in Terra Nova Bay (Fig. 4.7) is most convenient for study and its existence has been the most fully explained. Strong, persistent westerly katabatic winds blowing offshore and the Drygalski Ice Tongue which blocks northward advecting sea ice from moving into the Bay are the primary cause of the polynya (Kurtz and Bromwich, 1985). Synoptic winds, deep water and water circulation patterns are secondary influences. Sea ice is continually forming in this 1,000 km2 polynya for probably 9-10 months of the year to be advected eastwards contributing to the persistent pack ice zone in the north-west Ross Sea (Fig. 4.7). The same polynya, however, also helps disperse the pack there in late summer due probably to the increased ice divergence and upwelling of deeper water.

The large Ross Sea polynya appears to be driven by southerly through easterly winds and probably by some upwelling of warmer or more saline water as well (Jacobs etal., 1970; Keys, 1984; Pillsbury and Jacobs, 1985; Zwally etal., 1985). Ice advection into the southern and central Ross Sea is limited because of the current and wind directions (Fig. 4.6), whereas the winds and upwelling act to drive and melt ice out of this area. The polynya therefore tends to open the Ross Sea from the south due to enhanced advection and melting (especially in summer). The strong but variable synoptic forcing of the polynya is probably the main reason why the development of open water in the area is not the same from year to year (Zwally et al.,1985).

The large seasonal changes in pack ice extent mean that the region, especially the Ross Sea, has an extensive marginal ice zone. This is the zone across which the advancing or retreating edges of pack ice sweep during the year. Studies made of this zone in the Arctic have considerable relevance to the Antarctic. The complex vertical and horizontal air-sea-ice interactions of the zone are, in general terms, likely to be the same in both polar regions. For example, direct observations of Arctic ice edge compaction during on-ice winds and divergence during off-ice winds also apply to the Antarctic. Divergence is also due in part to ice advecting over warmer water and thinning, and to warm water incursions into the pack in eddies. Such processes together with synoptic weather events and local ice and ocean dynamics help explain the pronounced variability in the position of the ice edge (e.g., Jacka, 1983).

The patterns of sea ice movement in the Pacific Sector are known only in general terms. Average directions and speeds of regional ice movement are portrayed by Romanov (1984) and Sturman and Anderson (1986) and shown here in Fig. 4.6. Near the coast, from about Thurston Island west to Cape Colbeck, ice drift is generally westwards apart from localized areas where katabatic winds force the ice offshore. In most of the Ross Sea, drift appears to be more towards the north-west and north especially in the north. However, in the southern Ross Sea, westward drift may also occur turning to the north near Ross Island then north-west again around Cape Adare. North of the Ross Sea, northward movement appears to be most common until the latitudes of the West Wind Drift are reached and where eastward drift predominates (Fig. 4.6). These sea ice drift directions are not always consistent with iceberg trajectories. Elsewhere in the Pacific Sector, sea ice movement tends to be more variable and information less reliable but there is evidence for circular (gyral) motion in the vicinity of the Antarctic Divergence (Romanov, 1984). Southwards drift may be most common at longitudes 130° to 140°W (Fig. 4.6).

The pattern and direction of predominant winds (e.g., Weyant, 1967), surface currents (e.g, Elliot, 1977) and the drifts of ships beset in the pack (Tchernia and Jeannin, 1983) are similar to that of the ice (Fig. 4.6). This implies that ice advection can be attributed qualitatively to both wind and current. However, the direction of ice drift can be highly variable over short periods (De Rycke, 1973). Drift rates are highly variable also with mean speeds normally less than 17 km day1 (0.2 m sec.1; Romanov, 1984; R. Moritz, pers. comm.) but maximum speeds probably exceeding 1.2 m sec."1 when driven by strong winds or tides (Wordie, 1921; De Rycke, 1973; Keys, 1984). Probably sea ice drifts faster than icebergs on average due to the greater influence of wind on the former but it is not uncommon to see icebergs caught in strong currents ploughing through slower moving pack ice.

There is as yet no widely used model to describe the persistent areas of pack ice (Fig. 4.7) in the Pacific Sector. Sturman and Anderson (1986) concluded that higher concentrations in the eastern Ross Sea and north of the Oates Coast were probably due to slower ice movement, convergence and accumulation of ice in these areas throughout the year. Persistent pack in the western Ross Sea was thought to be due to convergence caused by both offshore (katabatic) winds off Victoria Land and southerly (and south-easterly) winds further out to sea and in the south. While this model is very plausible, advection of ice and cold water in regional currents may have been underestimated. Advection appears to be significant off the Oates Coast and the eastern Ross Sea where regional currents have been best defined by iceberg drift (Fig. 4.6).

Nevertheless, convergence remains an underlying factor. It seems a fundamental cause of the two other major areas of persistent pack ice in the Pacific Sector, in the Amundsen and Bellingshausen Seas (Figs 4.6; 4.7). Surface air temperatures in these areas are about -2° to -6°C in January, colder than elsewhere along the coast of the Pacific Sector at this time (Schwerdtfeger, 1970). Sea ice, therefore, probably forms in the vicinity throughout the year and is made to converge in these areas by the pattern of winds and currents. Katabatic winds off the continent could contribute to this convergence for instance north of the coastal polynyas at 110° to 150°W (Fig. 4.7). Elsewhere such coastal polynyas tend to develop sufficiently to break through the most persistent band of pack ice further offshore (Fig. 4.7). Convergence may also be increased locally by coastal topography for example in Lady Newnes Bay where the coastline trends northeast approximately normal to south-easterly winds and the regional current in the western Ross Sea.

A further example of local convergence may be caused by the Drygalski Ice Tongue which is advancing out into south-west Ross Sea. This could tend to increase sea ice convergence south of it (up current and up wind) and would have implications for logistic, scientific and biological activities in McMurdo Sound.

0 0

Post a comment