Jökull - 01.12.1969, Qupperneq 99

count every day, when even a small amount
of pack is sigthed somewhere at the coast. This
rnakes a great difference in many years, parti-
cularly with light ice. It seems likely, that
even two or three months of such light ice
per year might not have been noted in the
colder period of the nineteenth century. For
the period 1591 — 1780 standardization of the
ice records was aimed at as to make tliem
comparable with Thoroddsens records for
the years 1781 — 1915. Since the Meteorological
Office of Iceland was established in 1920, the
ice records have been much more detailed
(Fig. 3) and in a continuous graph of the ice
incidence as in Figs. 1, 5 and 6, it was there-
fore inevitable to reduce the observed ice
amount of the last decades considerably. This
has to be remembered in looking at all these
diagrams, but it must also be admitted that
this fact reduces considerably the confidence
of these ice records.
The third type of data is shown schemati-
cally in the lowest grapli of Fig. 1. This gives
the number of severe years in every decade.
The definition of such severe years is that
there lias been starvation, people clying from
lack of food, or that the drift ice has reached
the SW-coast of Iceland. In order to make the
graph continuous tlie number of severe years
is also defined for those decades when no such
year occurs. If a severe year occurs in the next
decade before or after, the period between
severe years is thus considered to be 20 years,
and therefore this number is \A for the decade
in consideration. The scale of the severe years
is not linear in the drawing. It was considered
possible to draw this graph for a much longer
period than the graph for decadal incidence
of ice. When historical information is scanty,
one may expect unusual events such as severe
starvation or drift ice surrounding the country
to be recorded, even if the annual amount of
ice is not mentioned. Sometimes it is also evid-
ent that the record of severe years is complete
for a. long period. Thus it is written that
famine occurred about 976 AD, and that an-
other famine occurred 80 years later. However,
in the 15th century there is a bad gap in the
written history, and therefore that part of the
graph was omitted in Fig. 1.
Now we shall look for the correlation be-
tween these three factors in Fig. 1. Using this
correlation we attempt to extrapolate the
temperature and ice graph for most of the era
of human settlement in Iceland. The result is
shown in Figs. 5 and 6.
fsamánuftir á ár Ársh i f 3
S n
u J JL
atug U LT -n
u
u
o 1 A 1 r^in
r T lIL ,1 rrT I Z
1 —rr n r TTrr- Mim: Lf nrii ir r
II IT r J
TT —r
Jt— 11—7 —r~ Hartar i á ár atug
Vo ] 1/10
1/12 J ir
900 -KX 30 ttí X) 12 00 1300 14 30 1500 1600 1700 18 00 1 900
Fig. 1. A schematic graph of the information used in the paper. Top: Mean temperature of
Stykkishólmur and Teigarhorn: Center: Decadal ice incidence in months. Bottom: Decadal num-
ber of severe years.
JÖKULL 19. ÁR 95