Whale oil

Wrecked by ice off the South Orkneys, 24 Feb. 1928.

Gotland II (1,5001)

Wellington to Yule Bay


Crushed and sunk between pack and fast ice in Yule Bay, Oates Coast, 17 Dec. 1981.

Southern Quest (3281)

Hobart to McMurdo Sound


Caught and sunk by pack ice near Beaufort Island, south-west Ross Sea, 11 Jan. 1986.

Bahia Paraiso (9,6001)

Marguerite Bay to Esperanza

Expedition, 82 tourists

Ran aground and sank off Anvers Island, off Antarctic Peninsula, 31 Jan. 1989

Fig. 4.5. Distribution and mean concentration of icebergs in the Pacific Sector (after Romanov, 1984). Various zones delineate areas where a given number of icebergs per 1,000 km2 of ocean can be expected.

1 Other vessels lost at sea from unexplained causes plying this or similar routes were the Glenmark, 968 tonnes (1872), Loch Dee, 738 tonnes (1883), Kilmery, 813 tonnes (1882), Loch Fyne, 1,290 tonnes (1883), Dunedin, 1,341 tonnes, (1890). Other vessels not involved in the New Zealand trade would have been lost on other shipping routes in the Southern Hemisphere. Ice could have been involved.

2 Numerous ships have been beset by pack ice and some were damaged or severely threatened before escaping.

In the Pacific Sector, the Belgica (trapped for one year in the Bellingshausen Sea from March 1898), Aurora (10 months in the Ross Sea and south-west Pacific Ocean from May 1915), Ob (3 months off Oates Land from April 1973), and Mikhail Somov (8 weeks off Oates Land in 1977, and 19 weeks east of the Ross Sea from March 1985) gave useful information on ice movement and currents as they drifted with the pack. Other vessels have been stuck for shorter periods where pack ice concentration is high, for instance early in the navigation season (e.g., Gotland II, for several days from 23 November 1981 north of the Ross Sea; John Biscoe for a few days from 13 November 1985 off the western coast of the Antarctic Peninsula) or in areas of persistent pack ice (e.g., Polar Sea once or twice in late February 1984 in the Amundsen Sea).

noted trapped in fast ice within a few kilometres of the shore of South Victoria Land but these are not representative of average concentrations offshore there.

Concentrations of icebergs decrease away from the coast although this decrease is not the same everywhere. Zones of differing concentration, roughly parallel to the coast (Fig. 4.5) are mostly parallel to regional current directions. However, a zone shown trending north-east of Cape Adare in Fig. 4.5 and also shown by Keys (1983) is not consistent with known currents. This iceberg pattern may be due to an accumulation of icebergs drifting in converging currents (Fig. 4.6). Alternatively, the zone could be an artifact of a divergence in the coastal current north-east of Cape Adare as the main portion of the coastal current swings to the west around the Cape. A zone north to north-west of Cape Colbeck, not shown on Fig. 4.5, is illustrated by O. Orheim (pers. comm.) and is due to the subdivision of the East Wind Drift there (Fig. 4.6).

Icebergs move around the continent initially in coastal currents, particularly the East Wind Drift (Fig. 4.6). They drift in an erratic fashion because of current eddies, tides, wind, waves, and the Coriolis force due to the earth's rotation. However, the nett drifts of icebergs follow the west-setting currents at rates thought to be between 50% and 75% of the speed of the surface currents (Tchernia, 1977; Swithinbank et al., 1977; Tchernia and Jeannin, 1983). Offshore in the Pacific Sector, measured speeds average 3-12 km day1 (0.03-0.14 m sec."1). Various studies have shown that ocean currents are more important than winds (even strong ones) in determining directions of movement (e.g., Tchernia and Jeannin, 1983) but wind can be significant (Veskin, 1982), especially when other currents are slow or absent.

Near shore, average speeds are much slower and the bergs frequently run aground. The giant berg B-9 averaged a little over 1 km day1 during the first three months of its drift in which time it collided with the ice shelf (Keys, 1988, 1989) (Fig. 4.4). The furthest distance a positively-identified berg has travelled in one open-water season northwards up the coast of South Victoria Land is only about 100 km, which gives a minimum estimate of iceberg drift rate, including any time spent aground, of 1-2 km day1 (Keys and Fowler, 1989). This is about half the mean current speed inferred by Lewis and Perkins (1985) for the north-setting current there. These speeds are much slower than the drift of some icebergs that kept apace with the Aurora while it was beset in pack ice in western Ross Sea in 1915 (8 km day1, from Wordie, 1921). Wright and Priestley (1922) reported that some icebergs moved as fast as 2 m sec."1 for short periods in coastal areas.

There are a number of places around the continent where drifting icebergs turn north from the coastal zone (Tchernia and Jeannin, 1983). In the Pacific Sector, the best defined of these appears to be north of Cape Colbeck east of Ross Sea (Fig. 4.6) where the continental shelf break starts trending north-west and so subdivides the East Wind Drift. Consequently, relatively few bergs seem to enter Ross Sea from the east, most instead being carried northwards. Thereafter, they are carried into the great Antarctic Circumpolar Current (West Wind Drift) with sinuous but generally eastward drifts. A similar north-turning zone is located north of Cape Hudson (Fig. 4.6) west of Cape Adare on the Oates Coast (Tchernia and Jeannin,

Fig. 4.6. Patterns of ice movement and ship drifts in the Pacific Sector, adapted from Elliot (1977), Swithinbank et al. (1977), Tchernia and Jeannin (1983), Romanov (1984), Sturman and Anderson (1986), Keys and Fowler (1989), R. Moritz, D. Harrowfield and B. Mcintosh (pers. comm.) and miscellaneous ice charts from the U.S. Joint Ice Center.

Fig. 4.6. Patterns of ice movement and ship drifts in the Pacific Sector, adapted from Elliot (1977), Swithinbank et al. (1977), Tchernia and Jeannin (1983), Romanov (1984), Sturman and Anderson (1986), Keys and Fowler (1989), R. Moritz, D. Harrowfield and B. Mcintosh (pers. comm.) and miscellaneous ice charts from the U.S. Joint Ice Center.

1983). Icebergs may also turn north from the coast in the Bellingshausen Sea in a possible current gyre revealed by the year-long drift of the barque Belgica beset in the ice pack from March 1898. Cape Colbeck, Cape Hudson and Thurston Island may all mark places where iceberg populations are subdivided by regional current patterns and only those icebergs driven by local and tidal currents mix across them.

The extreme northern limit reached by icebergs in European history (c. 200 years) is approximately 40°S in the Pacific Ocean near the Subtropical Convergence (Brodie and Dawson, 1971; Burrows, 1976). Furrows gouged on the Chatham Rise by grounding icebergs imply that bergs were common in New Zealand waters during the Pleistocene Glaciation (Kudrass and von Rad, 1984).

However, bergs cannot survive long in the rough temperate ocean north of the present-day Antarctic Convergence or even the pack ice zone. Cessation of signals from icebergs carrying radio transponders and decrease in observed sizes of icebergs in open water (I. Allison, pers. comm.) suggest that most bergs break up soon after they reach open water; this creates a serious, possibly fatal, obstacle to the towing of icebergs for their water. Few modern bergs have been sighted north of about 55°S in the Pacific (Fig. 4.5).

Icebergs decay mainly by splitting, wave action at the waterline, melting below the waterline and calving off the sides above the waterline (Budd et al., 1980; Orheim, 1980; Keys and Williams, 1984; Hamley and Budd, 1986), probably in that order. Observations near the coast of south-west Ross Sea suggest that splitting is caused mainly by stresses developed during grounding on the sea-floor, collisions by other bergs and upwards buoyancy on underwater portions (J.R. Keys, unpubl.). Offshore, stresses induced by ocean waves (Wadhams et al., 1983) are probably more important. Calving occurs when waves erode notches at the waterline, undercutting the ice at about 100 m yr1 (Keys and Williams, 1984). This often leads to caves, arches and wave-cut platforms at the waterline. Melting below the waterline is enhanced by roll-over and wallowing of bergs, strong currents and higher sea temperatures. Iceberg melt rates, which have been estimated from measurements of side melting to be about 20 m yr1 at -1°C (Keys and Williams, 1984), are therefore much faster than basal melt rates of ice shelves in water of the same temperature.

The life expectancy and residence time of icebergs depend on the mechanisms and rates of decay, drift speeds and on their initial size. Various tabular bergs drifing in the Pacific Sector were tracked by Tchernia and Jeannin (1983) and the U.S. Joint Ice Center (R. Godin, pers. comm.) for up to two years but their average lifetime is thought to be longer than this (Hult and Ostrander, 1973) (some icebergs elsewhere in the Antarctic have been tracked for several years, Swithinbank et al., 1977). The sparse data given above suggest that it might take one to two years for bergs to drift up the 800 km long Victoria Land coast of Ross Sea. In general, the half-life of bergs near the coast is possibly two to five years but only a year for bergs that have drifted north from the coast (Orheim, 1985). For example, the average half-life before breakage or roll-over of bergs less than 1,000 m wide in water of about + 1°C (north of the coastal zone) is estimated to be about 0.2 years (Hamley and Budd, 1986). However, bergs grounded near the coast have been known to survive for many years, perhaps decades, in the cold water there, especially when they are locked in fast sea ice for most of the year and protected from the effects of waves.

In coastal waters in the Ross Sea Sector, icebergs commonly ground in water depths less than 200 m, although a few ground in up to 300 m. About 75% of 285 bergs trapped in fast ice near the coast in the south-west Ross Sea were aground (Keys, 1985). Soundings, using a weighted wire line deployed through holes drilled in fast ice beside grounded bergs there, revealed water depths from 10 to 273 m. Icebergs in seaward extensions of fast ice off Cape Colbeck and Grant Island further east (Hult and Ostrander, 1973) were in about 200 m of water and were also probably aground. Stationary icebergs observed in open water off Cape

Adare Peninsula and Franklin Island were also in less than 200 m (Keys, 1983). No bergs have been observed grounded in water deeper than 300 m in this area. The draft of most modern bergs in Ross Sea is therefore probably less than 300 m.

Icebergs having drafts of more than 300 m may occur in the Ross Sea, for example some derived from the thick Mackay Glacier Tongue (Keys, 1983). Measured thicknesses (Bentley et al., 1979; Jacobs et al., 1986) and post-measurement deformation of the source area in Ross Ice Shelf of the giant berg B-9 suggest that its maximum draft may be about 300 m. Fragmentation and tilting of B-9 could produce bergs with drafts deeper than this. Its slow and seemingly erratic behaviour while over a deep submarine bank at 77°40'S, 166°50'W suggests that it was temporarily grounded there (Navy Polar Oceanographic Center, written communication, February 1988). Water depths between 400 and 500 m have been measured in the area (Davey and Cooper, 1987) so an alternative reason for its erratic behaviour may be a lack of strong currents over the bank.

About one-third of the bergs near the coast of south-west Ross Sea are tabular (Keys, 1985). Irregular shaped and rounded bergs are present in similar proportions. Close inspection indicates that about one-third of all bergs contain some dirt in the full range of environments for glacial sediments. About half of the tabular bergs are quite featureless in that they have no distinguishing upper surface that could be used to trace their source.

About 40% of the tabular bergs are less than 100 m thick and probably come from Ross Ice Shelf. The water depth beside grounded icebergs, together with their height (freeboard) give an indication of their thickness. Table 4.2 lists 10 tabular bergs representative of those in south-west Ross Sea which have a stratified snow pack and a planar upper (top) surface. The composition and density of these bergs are similar to that of Ross Ice Shelf (Keys, 1985) whose thickness at the ice front is thinner than 100 m west of 173°E (e.g., Fig. 4.3) and in Bay of Whales (see above). Calving of the ice shelf in recent years has probably been most prevalent west of 178°E (Jacobs et al., 1986 — see above). Most bergs calved from the Ross Ice Shelf probably drift westwards, parallel to and within about 100 km of the ice front (Keys and Fowler, in press). It is therefore highly likely that most of the bergs listed in Table 4.2 came from Ross Ice Shelf west of 178°E.

It is unlikely many tabular bergs of similar thickness come from sources east of Cape Colbeck as the subdivision of the East Wind Drift north of this cape (Fig. 4.6) would tend to prevent this. However, the Sulzberger Ice Shelf immediately to the east of Ross Sea is probably up to about 40 to 70 m thick near the ice front (Anderson, this volume) and may have produced some of the bergs present.

These "Ross Ice Shelf" tabular bergs are in a minority near the South Victoria Land coast. Only about 10% of the bergs trapped in fast sea ice near this coast in November 1984 were tabular bergs which appeared to have been little decayed and therefore must have been derived relatively recently from the Ross Ice Shelf west of 178°E. The composition and shapes of a further 30% of bergs suggest that they too had been derived from the ice shelf. Their combined volume was estimated to be in the order of 0.1 km3, only 1% of the 10 km3 that may be calved annually on average from this part of the ice shelf (see Ross Ice Shelf, above). It is

Was this article helpful?

0 0

Post a comment