Jökull - 01.12.1969, Page 98

An Estimate of Drift Ice and Temperature
in Iceland in 1000 Years
PÁLL BERGTHÓRSSON,
THE ICELANDIC METEOROLOGICAL OFFICE, REYKJAVÍK, ICELAND
ABSTRACT
A fairly good correlation is found between
annual temperature and the incidence of drift
ice in Iceland. This relation is used to estimate
the temperature in Iceland in the period 1591—
1816. Furthermore, the temperature and ice
incidence, in the years 1591—1969 (estimated
and measured) are found to be correlated with
the frequency of severe years, defined in a
special manner. This makes it possible to esti-
mate the temperature and ice incidence in the
period 930—1591 A.D.
The graphs of temperature and ice seem to
be in good accordance wilh the results of ana-
lysis of 01fi in a glacier ice core at Camp
Century, Greenland.
INTRODUCTION
The following investigation was presentecl
in a preliminary form at a conference on the
climate of the eleventh and sixteenth centuries,
held in Aspen in Colorado, June 16—24, 1962.
A short review of tliese preliminary notes was
published by Tatsuro Asai (1967). The pro-
cedure of investigation is here slightly modifi-
ed, but the conclusions are generally the same
as in the original paper.
MATERIAL USED
Fig. 1 gives a general scheme of the material
used in this investigation.
The first source is temperature measurements
dating back to 1846. As a temperature index
the annual mean temperature of two stations,
Stykkishólmur (65°05'N 22°44' W) ancl Teigar-
horn (64°41'N 14°21'W) was used. The ob-
servations started in 1874 at Teigarhorn, and
according to the experience since then the
94 JÖKULL 19. ÁR
most probable mean of the two stations for the
period 1846—1873 can be obtained by adding
0.1° C to the annual temperature in Stykkis-
hólmur for that period. This mean tempera-
ture should be rather representative for the
country, Stykkishólmur being affected by the
mild Irminger current at the west coast, but
Teigarhorn being rather under the influence
of the cool East-Iceland Current. A more de-
tailed running mean of this mean tempera-
ture is shown in Fig. 5.
The seconcl source, shown broadly in the
middle of Fig. 1, is the annual number of days
affected by ice on the coast, counted in months.
A more detailed running decadal mean of the
annual ice inciclence is shown in Fig. 5, dating
back to about 1590. For the years 1781 — 1915
the data collected by Thoroddsen (1916—1917)
were used, witliout any amendment, as read
from a diagram in his book “Árferði á íslandi
í 1000 ár”. For the period 1591—1780, Thor-
oddsen considered it impossible to draw such
a diagram for every month. On the other
hand, it seems possible to estimate from his
book the annual ice incidence, partly based on
indirect information regarcling general weather
in Iceland. In general this means that one
has to increase the apparent incidence of ice.
One example may serve to justify this state-
ment. For the year 1758 the annals tell there
was no drift ice at the coast, ancl that this
had hardly occurred in memorable time. Even
if no ice is mentioned in the annals for 1751,
1752 ancl 1753, it is therefore hardly justiíied
to consider them to have been ice-free. A word
of warning must. however be said here about
this method of counting the days with incid-
ence of ice. It is a matter of definition, wheth-
er we only count days with rather heavy ice,
seriously affecting navigation, or whether we