Jökull - 01.12.1969, Síða 30

definitely cold (1966) and four exceptionally
cold (1963, 1965, 1967 and 1968).
Sea temperature was by far the lowest dur-
ing the last period, and this applies not only
to the surface layers, but also to the inter-
mediate layers all over the North Icelandic
coastal area.
Considering the complexity of meteorological
and oceanographical factors, an analysis of the
possible causes of the year to year fluctuations
may be very difficult. Sonte of the factors most
likely to affect the conditions north of Iceland
during the summer season will now be con-
sidered. These are:
1) Variations in the distribution of drift ice.
2) Variations in the concurrent meteoro-
logical conditions.
3) Effects of meteorological conditions in the
past.
4) Variations in the properties and relative
proportions of water masses formed in the
area north of Iceland.
5) Variations in the properties, extension and
current velocities of Atlantic water in the
Irminger Sea.
Although the distribution of drift ice must
largely be considered the effect rather than
the cause of climatic changes, one must keep
in mind that the drift ice ntay exert great local
influence by cooling limitecl areas.
Up to the last few years, drift ice has rarely
been observed in the coastal area in recent de-
cades except for occasional scattered floes in
the western part. Direct effects of drift ice
would therefore primarily be expected off the
northwest coast and in the area west of Gríms-
ey. The cooling effect, however, must be less
in winter, when the surface temperature is al-
ready low, than in spring or summer when the
surl’ace layers have warmed up. Besides, drift
ice normally occurs more l’requently in spring
than in other seasons. Comparison between
frequency of drift ice ancl sea surface tempera-
ture iti the western part of the area in summer
during the years 1894—1939 revealed a distinct
negative correlation (Stefánsson 1954). The cor-
relation between the temperature at Grímsey
in spring (March—May) and the frequency of
ice in the North Icelandic coastal area is in-
dicated in Fig. 8. The number of days with
26 JÖKULL 19. ÁR
ice near the coasts of Icelancl is based partly
on Thoroddsen’s ice chart (Thoroddsen 1917)
and partly on information published by the
Danish and Icelandic meteorological institutes.
The ice curve must be taken with great re-
servations. As explained elsewhere (Stefánsson
1954) the estimated ice frequency is at best
only a rough approximation of the actual ice
distribution. Nevertheless, the similarity of the
two curves definitely supports the hypothesis
that sea temperature ancl ice frequency are
related.
Although variations in sea surface tempera-
ture can thus be partly attributed to the dis-
tribution of ice, the drift ice does not explain
the variations taking place in ice-free years,
nor does it explain variations in the quantity
of Atlantic water in the subsurface layers north
of Iceland. Owing to the low density of polar
water or melted water from ice, its effect will
generally not be significant in the deeper-lying
water strata. Scattered ice floes are not likely
to slow down a major ocean current, but on
the other hand the appearance of ice may be
the result of changes in the circulation, e.g.
increased penetration of polar water to the
Icelandic coastal area.
There exists a small but significant correla-
tion between the sea surface temperature and
the concurrent air temperature. This associa-
tion may be both direct and indirect. Thus
there may be a transport of sensible heat from
the sea to the air, whereas drift ice will cool
both the sea and the air. Furthermore, there
exists a positive correlation between sea sur-
face temperature and concurrent southerly
winds (Stefánsson and Guðmundsson 1969).
This can be explainecl by the fact that warm
weather is normally associated with southerlv
winds.1) On the other hand, there is little
correlation between temperature or salinity in
the deeper layers and the concurrent air tem-
perature or wind.
It has been establishecl (Stefánsson 1962) that
the hydrographic conditions at intermediate
depths (50—200 m) north of Icelancl depend
1) This may not always apply to the inshore
area, where southeasterly winds can lead to up-
welling of cold water in summer (cf. Stefáns-
son and Þórðardóttir 1965).