Barents Sea drift ice characteristics * TORGNY VINJE and A N U N D S . KVAMBEKK Vinje T. & Kvambekk AS. 1991: Barents Sea drift ice characteristics. Pp. S9-68 in Sakshaug. E . . Hopkins. C. C. E. & Britsland. N . A . (eds.): Proceedings of the Pro Mare Symposium on Polar Marine Ecology, Trondheim. 12-16 May 1990. Polar Research l o ( / ) . A short review of the ice field characteristics in the Barents Sea is given. A digitised ice distribution data set covering the period 1 9 6 6 8 8 has been used to calculate the probability distribution of sea ice and temporal trends. The ice-covered area between 68 and 83"N and 10 and W E at the end of the melting season is found to have been reduced by 40% during the last 23 years. The maximum extension, however. shows no significant decrease over the same period, and this indicates that there might have been a reduction/increment in the ice thickness/melting effects in the Barents Sea during the last 23 years. Torgny Vinje and Anund S . K u a m b e k k . Norsk Polariistitutt, P.O. Box 158. N-1330 Oslo LufthaLm. N o r w a y . Introduction The sea ice conditions in the different marginal seas bordering the Arctic Ocean are determined by dynamic and thermodynamic processes charac- teristic for each area. In addition, the exchange with the Arctic Ocean may influence, and in some cases dominate, the age-composition of ice in the various seas. The physical boundaries of the different marginal seas are in this connection of major importance. The western and southern parts of the ice fields of the Barents Sea are bounded by the warmer currents spreading north and eastwards from the Norwegian Sea. Towards the east there is a relatively large opening to the Kara Sea between Frans Josef Land and Novaja Zemlja; this facilitates the exchange of ice masses with the neighbouring sea in this area. The bound- ary towards the north is linked by a number of islands which to some extent govern the exchange of ice with the Arctic Ocean. The ice drift in the Barents Sea is mainly wind- driven. The exchange of ice with the Kara Sea and the Arctic Ocean may thus vary considerably from year to year in accordance with changes i n the general atmospheric circulation. This exchange in turn determines to a large extent the amount of ice that remains in the Barents Sea through the subsequent summer. The com- position of the ice fields may therefore at times be very complex, consisting of local and imported ice of various ages. Glaciers in Frans Josef Land and Svalbard feed * Norsk Polarinstitutt Contribution No. 275. icebergs to the area. This occurs at a highly vari- able rate, mainly determined by the surging activity. The subsequent distribution of icebergs is determined by the prevailing wind field. An extreme event occurred in the spring of 1929 when about 20 icebergs were observed along the coast of Finnmark (Hoe1 1962). Distribution The denser, warmer modified Atlantic Water fills up the deeper part of the Barents Sea. The south- ward extension of the sea ice is therefore during the cold season indirectly, topographically con- trolled i n the western part, where the bottom topography shows marked shelf breaks. Similarly well-defined shelf breaks do not exist in the east- ern part. Contingent upon the intensity of the governing dynamic and thermodynamic pro- cesses, the increasingly cooled water flowing east- ward from the Norwegian Sea may here dive under the polar water at quite different latitudes. These features are reflected in the increasing variability of the ice edge position in the eastern part of the Barents Sea (Fig. 1). The degree of cooling of the warmer water masses entering the southern area from the west as well as the ice formation during the cold season are determined by the atmospheric and oceanic circulation. The position and the intensity of the Barents Sea Low are in this regard of particular importance. Prevailing westerly/easterly winds will for example cause less/more ice formation in 60 Torgny Vinje & A n w i d S . Kvambekk FIR 1 . The probability of encountering sea ice at the end of April in the Barents Sca. We note the eastward increasing variability in the ice edge position. From satellite-based Ice maps during 19661988. mainly from the Norwegian Meteorological Institute ( D N M I ) . the N A V Y - N O A A Joint Ice Center. I: S . A . . and observations from ships and aircraft collected by Norsk Polarinstitutt. this area and accordingly determine the position of the ice edge (Vinje 1976-1982). T h e effect of heat transfer from the warmer water masses flowing in from t h e west is illustrated by a cor- respondence between temperature anomalies at the Kola meridian oceanographic section and anomalies in the ice extension in the Barents Sea (Saetersdal & Loeng 1984). Loeng et al. (1983) also found that oceanic temperature changes in the eastern part will occur o n e year later than in the western part. Helland-Hansen & Nansen (1909) found similar indications, mainly that cli- matic variations in the Barents Sea lagged o n e year with respect to the Lofoten section. Oceano- graphic features will accordingly play an impor- tant role in long-term ice forecasting. If the Barents Sea Low center is located in the central part of the Barents Sea, and if the Barents Sea drift ice characteristics 61 circulation is more intense than normal, we will have an increased surface speed of warmer water as well as relatively high atmospheric tem- peratures and northward drift in the eastern part of the Barents Sea. This was the case, for example, in March 1985 when there occurred a broad opening, extending up to the northernmost tip of Novaja Zemlja (DNMI Ice Maps). If on the other hand the air pressure is high during the freezing season, there will be a retarding effect on the eastward flowing warmer water, and cold, continental winds may favour the cooling and freezing over of large areas. This happened in 1979 when an extreme western position of the ice edge was observed towards the end of the freezing season (Vinje 1980). A deviation from the mean ice limit has a traceable persistency in the Barents Sea over a period of 2-3 months (Lemke 1980). This impor- tant prognostic feature, here given a statistical number, is so predominant that it has for a long time (on an empirical basis) been used for the planning of expeditions t o the area. The seasonal variation of the ice-covered area of the Barents Sea and the adjacent part of the Arctic Ocean shows great interannual variations (Fig. 2 , Area: 6goN-83"N, lVE-55"E). The 1970 1975 1980 1985 1970 1975 1980 1985 YEAR Fig. 2. Annual ice-covered area in the Barents Sea and the adjacent part of the Arctic Ocean (68N-83"N. 10E-55"E). as observed for the period 1966-1988 at the end of April (upper) and at the end of August (lower). The database is the same as for Fig. 1 . seasonal variation of the sea ice distribution shows a maximum/minimum extension in late April/ August, respectively. The ice-covered area in late April was during 1973-1976 about 700,000 km2 and about 1150,000 km2 in 1969 and 1979, revealing a vari- ation of as much as 400,00~500,000 km2 in the annual maximum extension over a period of four years. It is noted that the variability in the maxi- mum extension is far greater than the variability in the minimum extension. The Barents Sea south of 80"N has been com- pletely ice free for one or two months in 1972, 1976, 1984, 1985, and 1986. The long-term vari- ation from the turn of the century indicates that the extension at the end of the melting season has been reduced markedly (Sear 1988). A cor- responding yearly reduction of 5.4 2 2.7 km?/ year is found from correlation studies for the last 23 years (1966-1988) from the area represented by Fig. 2 (late August). Thus, the minimum ice extension in our area has been reduced by as much as 40% during this period. The maximum extension (late April), however, shows no signifi- cant change over the same period. The increasing difference between the annual minimum and maximum extension indicates that a reduction/ increment of the ice thickness/melting may have taken place in the Barents Sea over the two last decades. Polynyas One of the main areas for polynya occurrence in the western Barents Sea is found in the inner part of Storfjorden. When northerly winds prevail, newly formed ice is continuously transported southwards resulting in an opening outside of the rim of land-fast ice. Because of the considerable ice formation that takes place, there occurs a continuous brine production which results in ver- tical convection and the formation of cold, dense bottom water (Midttun 1985; Quadfasel et al. 1988). Polynyas are also formed on the lee side of the smaller islands in the archipelago. These are consequently not stable but change position with the wind direction. Hopen, Kong Karls Land, Storoya, and Kvitoya are areas where polynyas may frequently develop. Fig. 3 shows the prob- ability of encountering open water (polynyas) in parts of these areas. Ice-free areas are observed during the winter season 10%-30% of the time, 62 Torgny Vinje R Anitnd S . K v a m b e k k Storfjorden Hopen WEEK FIR. 2. The probability of encountering polynyas of an) sizc at four various locations in thc Barents Sea for the period 1966- I Y X X . The dalahase I S t h e same as for Fig. 1 and the probability does not change significantly during the winter. (There is, of course. a marked peak during the melting season.) As north-northeasterly winds a r e most frequent over the area during the cold season (Gorshkov & Faleev 1980). a steady divergence occurs in the bng Karls Land 20 J O 4 0 50 WEEK ice fields with a corresponding formation of leads (Fig. 4). T h e total open areas in the ice fields of the Barents Sea may therefore be considerably larger than the area of the polynyas. In addition, we also have on a smaller scale t h e effect of the twice daily tidal divergences in the ice field, in this case particularly over shallow banks or near the islands. Composition Thickness As mentioned above, t h e ice fields may constitute quite a complex picture, containing ice of various age and origin. As the ice drift in the Barents Sea is mainly wind-driven, the ice exchange with the neighbouring seas is to a large extent determined by the atmospheric stress field. T h e Barents Sea may on an annual average be a n ice source for the Arctic Ocean and an ice sink for the Kara Sea over periods of about a decade (Zacharov 1976). However, on the average, there is a seasonal variation indicating a reversed ice exchange with the neighbouring seas during t h e warmer season (Vinje 1988). The dominant ice flow in the Arctic is the Transpolar Ice Drift Stream which pours from 4000 to 5000 k m 3 ice per year into the Greenland Barents Sea drift ice characteristics 63 Fig. 4. Infrared NOAA A V H R R image showing lee polynyas and fracture patterns in the Barents Sea during a period with northerly winds, 27 February. 1987. Sea (Vinje & Finnekisa 1986). (This corresponds to a fresh water supply to the Greenland Sea of the same order as the Amazon discharge.) In comparison, the average annual transport from the Kara Sea to the Barents Sea is on the order of 500km3, and from the Barents Sea t o the Arctic Ocean only 3&40km3 per year (Vinje 1988). Although the latter two figures are small, their variations play an important role for the stability of the water column as well as for the local heat budget, particularly because of the shallowness of this sea. The thickest ice encountered in the Barents Sea is generally imported from the Arctic Ocean. This influx of perennial or multi-year ice may be extra large during conditions with persistent northerly winds in the region. This will affect the heat budget of the area t o a great extent, resulting in a larger extent of sea ice the following summer season. Such a succession of events took place. for instance. in 1988-89 when the ice fields at the end of July 1989 showed a 60-years’ extreme southern extension of multi-year ice (Volkov & Vinje 1990). The thickness of the land-fast ice at Hopen reached a maximum of 173 cm at the end of April during the extreme cold season of 1968-69 (Vinje 1985). Based on t h e relationship between the ice thickness and the accumulated degree-days at this site, we have calculated the long term average ice 7 a b k I Ice t h i c k n c \ \ [ c n i ] JI bariou, l a t i t u d e 3 in t h e Baren15 S e a hascd o n the o h s e r v e d r c l a t i o n \ h l p h e r w e e n ice t h i c k n e ~ a n d accurnul;itcd dcgrc.c-da!\ f o r t h e p e r d lYh7-1YXO ( a f t e r Vinje 19x8) L a t i t u d e 75% 7h'S 77:2 78's 7 Y Z N 80's T h i c k n e s s ( c r n ) Y 5 105 1 3 ) 1311 145 1711 thickness that may be formed at various latitudes (Table 1). The degree-days are derived from daily averages of temperature and are the sum of the number of degrees Celsius below zero experi- enced each day during the period of sustained frost. Table 1 illustrates the effect of the considerable meridional temperature gradient in the area. We have noted that the extreme at Hopen. 76.5"N, was 173 cm. Comparing this figure with Table 1 gives us an indication of the maximum deviation we may expect from the mean (60 cm). The multi- year ice drifting in from the Arctic Ocean is generally 2-4 m thick. In addition there is a con- siderable formation of ice in the lee polynyas to the south of the islands during conditions with northerly winds. Typically. 20 cm of new ice may be formed per day during cold spells. This may result in an accumulation of about 6 m of newly- frozen ice per month in the steadily unveiled water in a polynya. Martin & Cavalieri (1989) calculate a new ice formation of 1&17m in the polynya near Frans Josef Land for the whole freezing season. If northerly winds prevail for some time. the result may be long lee-belts of thinner ice. bordered by thicker winter ice or multi-year ice from the Arctic Ocean (Fig. 3 ) . Thus the polynya effect may influence the ice thickness distribution in the Barents Sea to a considerable extent. A n additional ice producing agent is the tidal effect which twice daily opens and closes the space between the ice floes. This may cause an accumulation of frazil ice along the rim of the ice floes. illustrated bv the often observed raised ice floe edges. Floe size and frrsr ice distribution Wave effects produce small-sized ice floes of 10- 50 m in the outer 10-50 km of the ice margin. This active disintegration zone is most narrow in areas with great water depths, while it becomes par- ticularly broad i n areas where the waves are increased by a rising shelf break, for instance on Spitsbergenbanken. T h e wave energy is rapidly dissipated and further inward from the active wave zone only the long waves with periods of 15-25 s penetrate; accordingly. t h e floe size also increases from several hundred metres to several kilometres. During the freezing season off-ice winds will cause divergence and refreezing between the floes. thus producing larger floes of ice breccia (ice floes consisting of smaller floes of different age). In addition, a formation of ice bands extending perpendicular t o the main edge gen- erally occurs. T h e r e a r e a number of hypotheses and models indicating how these bands a r e pro- duced. but so far n o single hypothesis has been generally accepted. The formation of fast ice in the Barents Sea occurs in sheltered areas. T h e most important places are the inner part of the Storfjorden. the triangle between Nordaustlandet. Kong Karls Land and Barentsoya, a n d t h e archipelago of Frans Josef Land. A smaller, frequently recur- ring. fast ice area is found over t h e shallow shelf just south of Kongsoya. Ridging Ridging events may take place with onshore winds. Rubble fields, possibly d u e t o this effect, are often observed in the N W part of the Barents Sea. and ice piles 10-15m high have been observed o n the shore of the islands. T h e height of the ridging is also illustrated by the fact that fairly large accumulations of ice blocks may be observed on t o p of 7-15 m high grounded icebergs (Loset et al. 1988; Vinje e t al. 1989). Ice ridging also occurs in convergences caused by wind, cur- rents. and land constraints. The most frequent ridge height observed from the meteorological station o n H o p e n is 2 m . T h e most frequent ridge density in t h e ice fields of the Barents Sea as observed from a long term project using aircraft is about 10 km-l. T h e frequency distribution is fairly broad. with a maximum of 26 km-' (Vinje 1985). Ice bottom topography The ice bottom topography has been studied extensively using a scanning sonar ( Johnsen 1989). The measurements indicate that t h e multi- year ice has a relatively smoother bottom top- Barents Sea drift ice characteristics 65 . . . . --.- --.- . . . . /-. / / I I / / / / I 2 / Fig. 5. Typical bottom topography features of multi-year ice as mapped with a scanning sonar at 78.8"N and 29.7"E September 1989. The dimensions of the box are 40 m x 40 m x 10 m . The flat areas in the corners are areas with no data due to ship or ridge shadows. The grid spacing is 2 m . ography than the first-year ice in the marginal ice zone. This may reflect the effect of thermo- dynamic processes having acted over a longer time period for the former ice type. A typical sample of a multi-year ice bottom topography is given in Fig. 5 . An ice field is considered to have a very high oil pooling capacity due to capture in the leads, under and upon the ice floes. Johnsen 8~ Vinje (1987) estimated that the domes under the ice floes 10-90 km from the outer ice edge could trap as much as 0.08 m3/m2 on an average. Martin et al. (1976) found in a laboratory test that 0.004 m3/ m2 oil could be pumped upon pancake ice. Pan- cake ice is a very common ice type in the outer 10 km of the ice margin during the freezing season. Terrigenous and biogenous material in sea ice When the freezing starts in September/October, the coastal water masses contain a maximum of suspended material delivered by the run-off from a variable extent, dependent upon the con- centration of suspended material and the intensity of the freezing. The occurrence of sediment-laden ice is quite common in the Arctic Ocean (Nansen 1904,1906; Thiede 1988) as well as in the Barents Sea. ElverhGi & Solheim (1983) report that the material in the latter area is both of local and external origin. This is in accordance with the influx of ice to this region from other shelf areas in the Arctic. As mentioned above, the greater part of the potentially sediment-laden ice import generally comes from the Kara Sea. However, temporary deviations from the normal flux pat- tern indicate that considerable amounts may at times also come from other parts of the Siberian shelf. This was the case in 1989 (Volkov & Vinje 1990). Another phenomenon which was observed on Spitsbergenbanken in March 1987 is the accumu- lation of material in ice ridges which may act as traps for material brought up under the ice by turbulent currents over shallow banks. When these newly formed, nonconsolidated ridges are land and glaciers. The material in the upper water . broken up, mud-accumulation between the ice layers will be embedded in the new frozen ice to blocks and a number of dark brown ice surfaces 66 Torgriv Vinje & A n u n d S . Koambekk along the track of the ship may be observed. T h e observation of scattered muddy ice floes in the Barents Sea ice fields is not a rare event. The accumulation of bottom material trapped i n newlv formed ridges is a phenomenon that surely occurs during freezing in all shallow areas. When these ridges are consolidated during the following summer. storage of sediments as well as of nutrients, bacteria. or other biological material such as eggs and larvae will occur. Being trans- ported over long distances. this storage may form a basis for new biological growth in new areas during the melting season. then mainly in the marginal ice zone. The potential of an ice field to carry trapped terrigenous or biogenous material is indicated by the above-mentioned great number of ridges observed per km. Garrison e t al. (1989) observed that a high concentration of organisms i n sea ice may be the result of accumulation of frazil ice. The highest concentration of organisms should accordingly be observed along the down- uind border of a polynya or along the edge of ice Aoes where we will have an accumulation of frazil Ice. Another structure-forming agent is the melting and freezing that takes place in perennial ice. Freezing takes place at the bottom d u e to the conducted radiative heat loss during winter. Melt- ing mainly takes place on the surface due to the positive radiation balance during the summer. This repeated melting and freezing over a period of some years results in an upward transport of sediments. nutrients. and biological material. This may explain the frequent observations of very muddy surfaces or marked layering of dark material in the column of perennial ice. Ice drift Tidal or inertial effects play an important role for the daily movement of the ice, particularly over the shallow shelves, in passages, or along the coast. T h e movement of the ice fields d u e t o tidal or inertial effects is clearly illustrated by the trails formed by grounded icebergs (Fig. 6). T h e dia- meters of the loops vary between 5 km and 15 k m , depending o n the corresponding water depth and F i g 6 . Trails after grounded icebergs on Spitsbergenbanken caused by tidal or inertial effects in the moving ice field. Landsat TM image 1 June. 1988. The loop diameter is about 15 km. Barents Sea drift ice characteristics 67 produce double, mushroom-shaped eddies. This has repeatedly been observed near the southern tip of Hopen (Vinje 1985). Information on the ice drift in the Barents Sea has been collected mainly from ice drift buoys since 1975. To illustrate the variance caused by the passage of lows, the drift of three buoys deployed at the same time and longitude but at different latitudes can be considered (Fig. 7). The wind stress. If the number of grounded icebergs is large, the grinding of ice floes may have an effect on the floe size distribution in the area. The twice-daily tidal wave causes a periodic divergence and convergence over shallow banks, near the coast or in narrow passages. This is a well-known and marked phenomenon which has been taken advantage of for a long time by ice navigators. The tidal wave may also at times 80N 75N 7 0 N O E 20E 40E 60E 20E 40E 7 5 N 7 0 N Fig. 7. Drift of ICEXAIR-buoys in 1986/87. The buoys were parachuted on ice Roes 16 December 1986 at 76.8"N. 78.5"N. and 80.O"N along 40"E. OFF indicates when the buoy became afloat. Note the great variety in the drift directions. The dates in the figure are day/month. (After Johnsen & Vinje 1987). 68 average simultaneous ice velocity in the outher margin is observed t o be about 0.2111s-I with a mean net displacement of 0.06 m s - ' towards southwest (Vinje 1988). According to buoy drift observations, the ice velocity amounts to 1-2% of the geostrophic wind speed (Johnsen & Vinje 1987). This is in agreement with wind-induced ice edge displacements such as those observed o n satellite images (Vinje 1977). T h e wind effect on the ice increases towards the margin, probably because of the increased form drag d u e to wave effects (Stokes drift) as well as to the scattering of ice floes. Torgny Vinje Ce Anrinti S . Kiinnibekk References Elvcrhei. A . & Solheim. A . 1983: The Barents Sea Ice shcet- a sedimentological discussion. Polar Res. I . 23-42, Garrison. D L . . Close, A R. & Reimnitz. E. 1989: Algae concentrated by frazil ice: Evidencc from laboratory cxperi- mcnts and field experiments. A n r . S r i . I (1). 31S316. Gorshkov. S . G . & F a k c \ . V . 1. (cds) 19x0: Atlac ofrhe oceans The Arcric Oceari. USSR Ministry of Defensc. Helland-Hansen, B. & Nansen. F . 1909. Thc Noruegian Sea. Fiskdir. Skr Ser. Haciirnders. 2 ( 2 ) . 1-360. H o d , A. 1962: lsfjcll p i kysten av 0st-Finnmark Norsk Geo- grafrsk Tidsskrifr S 4 , 2 2 b 2 3 8 . Johnsen. A . S. 1989: Relations between top and bottom ice topography using a scanning sonar P r o c . 10th Con/ P o r t . and Ocean EngfrIt'ering under Arcrrc Condirtoils 1. 7 7 4 6 . Tekniska Hsgskolan i Lulei. Sverigc. Johnsen. A . S & Virile. T 1987: Havisunders~kelseri Barcnts- havct. A KUP-proslekr 1.3. Norsk Polarinstitutt. Oslo. 97 pp. Lemkc. P 1980: Application of the inverse modelling technique to Arctic and Antarctic sea ice anomalics. Hamburger Geo- physrkalische E i r u e l s c h r i f r e n , Reihe A . Wissenschofrltche Abhandlungen. Heft 4Y. 95 pp. Loeng. H . . Nakkcn, 0. & Rakncs, A . 1983: Loddas utbrcdelsc i Barentshavct i forhold t i 1 temperaturfcltct i pcriodcn 1 9 7 6 1983. Fisken OR Hover I . 1-17. Leset. S . . Vinje. T . . Levas. S . M . . Johnsen. A ' S . , Jenscn. H.. Erlingsson. B. & Vefsnmo. S . 19x8: I D A P 88 Vessel deplownenr. Volume 2 . Field ohseruarrons and a n a l w s . Norwegian Hjdrotechnical Laborator?. Trondheim. 138 pp Martin. S . & Cavalicri. D J 1989: Contribution of the Siberian shelf paltnyas 1 0 thc Arctic Ocean intermcdiate and deep uater J . Grophis. R e s . Y4 I C Y ) . 12725-12738 34artin. S . Kauffman. P. & Wclandcr. P E. 1976. A laborator! \tuJ! ot the di3persion of crude oil within scil ice grown in a ~ a \ c field Proc. 27Ih A l a s k o Sci. Con/: 2 . 261-787. Fdirhanks. Alaska. hlidttun. L. 1985: Formation of dense hottom water i n the Barcnts Sea. Deep-Sea Res. 32. 1233-1231. N a n w . F . IYOJ. The bathynctrical Iedtures of the North Polar Scac. w i t h ti discussion of the contincntal shclvcc and prcvioub owllatiom of the shore-line. Pp. 1-232 i n Nansen F. (cd.): 7 l i r .Yorwegian Norrli Polar Expedrriorr 1x93-lXY6. Scienrific resulrs 4 ( 1 3 ) . J . Dvhwad. Christiania Nanscn. F. IYM: Protozoa on the ice-floes of t h c North Polar Sea. Pp. 1-22 i n Nanscn F. (ed.): 7 h e Nornvgiarr North Polar €.rpedirion IXY.3-18Y6. Scietirific resu/r.s 5 ( 1 6 ) . J . Dyhwad, Chriahania. Quadfascl. D . . Rudela. B & Kurz. K . 1988: Outflow of dense water from a Svalbard fjord into the Fram Strait. Deep-Sea Re, 35, I I J , ~ 1 1 S I ~ . Sztcrsdal. G . 8: Loeng. H 1983: Ecological adaption of repro- duction in Arctic cod. Pp. 13-35 in Gode. 0. R . & Tilaeth, S . ( c d s . ) : Proc. Souier-Norwegian S y m p . . Reproduction arid recrrurtnem of Arcrtc r o d . Inatitutc of Marine Research, Bcrgcn Scar. C . B. 1988: An index of sea ice variation5 i n the Nordic Seas. 1. Climarol. 8 . 33%355. Thicdc. J . ( c d ) 1988: Scientific Cruise Report of Arctic Expedition ARK IV/3. Reporrs on Polar Research ( A W l J 4 3 . 237 pp. Vinje. T. IY7&1982: Annual reports. Norsk Polarrrisrirurr A r b o k . Oslo. Vinjc. T. 1980: On the extreme sea ice conditions observed in the Greenland and Barents seas in 1979. Norsk Po/aririsritur! A r b o k 1979. 5%65. Oslo. Vinje. T . 1977: Sea ice studies in the Spitsbergcn-Greenland area. L A N D S A T Report € 7 7 - 1 0 2 0 6 . U . S . Dcpt. of Com. N a t l . Info. Service. 59x5 Port Royal Rd., SpringRcld. V A . V1nje.T. 1985: Drift. composition. morphology anddistribution of the sea ice ficlds in the Barents Sea. Norck Pohrinsrirutr Skrifrer 17YC. Oslo. 26 pp. Vinje. T. 1988: Dynamics and Morphology of the Barcnts Sea ice fields. Proc. Yrh Coiif. Port a n d Ocean Engineering w i d e r Arrric Condirrons I . 26S-268. University of Alaska Fairbanks. Alaska, U.S.A. V i n j c . T . & Finnekisa. 0 . 1986: The ice transport through thc Fram Strait. Norsk Polarinsrirurr Skrifrer 186 39 pp. Vinjc. T . . J c n x n . H.. Johnsen. A . S . . Laset. S . . Hamran. S . . L e ~ a s . S . M . & Erlingsson. B. IYXY: IDAP-XY R I V L A N C E deplo! menr. V o l u m e 2: Field obseroations arid analysis. Norak Polarinstitutt. Oslo. 74 pp. \'olkov. V & V i n j c . T . (cds) 1990: Sovict-Norwcgian Occano- graphic Programme lY8&lYY2. Cruise reports 1989. Norsk Polarrrisrirrirr Rapporrserre 5 7 . 5 5 pp Zacharob. V . F 1976: Cooling of the Arctic and the ice cover of the Arctic Scas. (Trsl.: Norsk Polarinstitutt) A A N l l T r u l f ~ l 337 Gidrometcoizdat. Leningrad. 96 pp.