6. Discussion

From the data presented in sections 2-5 it is shown that the Odden feature in the early part of the winter (December to May) primarily consisted of local ice formation and decay. The short term development was closely correlated with local (Jan Mayen) temperature and wind conditions. It is also shown that the prevailing winds at Jan Mayen on the long run will transport ice into the Odden area, and thus there can be substantial portions of multi-year ice from the EGC in the Odden area. In particular since the mechanism that supports a bay of open water behind the classical Odden is missing this winter, ice may pass the 'Nordbukta' area without considerable melting. The summer Odden on the other hand is shown to consist of primarily old ice being advected into the area on a time scale of one month, and not the ice that was created during the May Odden advance.

Bourke et al, (1992) described the Jan Mayen surface current as a meander to the East Greenland Current. Surface drift patterns from ice floes and buoys showed the eastward menadering of the generally southwards drifting ice, but no observations at the surface of drift into the central Greenland Sea was reported. However, one float (at 500 dbar) drifted along very much the same track as the huge floe observed during this study. It is argued that the surface current field only in a very narrow region between 7330'N and 74N will carry ice into the Greenland sea, whereas to the south it will follow the meander pattern that eventually leads abck to the East Greenland Current. The reason why surface drift has not been observed before could be that we very seldom see such large floes that can be easilly identified in this area, or that the prevailing winds from the North on most occasions will push the surface floes into the area of the meander which subsequently will take them back to the East Greenland Current.

The last pieces of ice in Odden disappeared in mid August, but the large quantity of fresh water that was transported into the area as ice remains, and forms a layer of fresh water on the surface of a large part of the Greenland Sea.

The 1996 Odden development in some ways resembled the development in 1981, and in that year the Odden started developing again in early November. Also in 1996, we saw Odden development in November and December.