Jökull - 01.12.1969, Qupperneq 11

Temperature in Stykkishólmur 1846—1968
ADDA BÁRA S I G F Ú S DÓTT I R
THE ICELANDIC METEOROLOGICAL OFFICE, REYKJAVÍK
A weather station has been operated in
Stykkishólmur (65o05/N 22°44' W) a village on
the southern coast of Breidafjördur from Nov-
ember 1845. This valuable series of observa-
tions offers an opportunity to study tempera-
ture variations at this place for more than 100
years.
Before temperature measurements, which
extend over as long a period as this one, are
used to investigate climatic fluctuations one
has to consider other factors than real tem-
perature changes that may for instance affect
the values of the monthly mean temperature.
The weather station in Stykkishólmur has
been moved several times within the village
and the exposure of the thermometers has also
varied. Up to 1951 wall screens were used, ancl
their height above the ground was not always
the same. From Nov. 1951 the station has been
equipped with a free standing hut and the
thermometers have been approximately 2 met-
ers above the ground. Contemporary measure-
ments were made in the free standing hut and
the wall screen latest used, for over a year.
The result was that the free standing hut was
slightly warmer than the wall screen during
winter but colder in the summer. The cliffer-
ences for single months are shown in Table 1.
These corrections are in this paper applied
to temperature measurements made in the wall
screen used 1942—1951, but no correction was
applied to older measurements in other wall
screens at different places in the village, be-
cause comparison of wall screens and free
standing huts at several weather stations have
not given liomogeneous results. (Veðráttan, Ars-
yfirlit 1962 p. 122 and Ársyfirlit 1966 p. 114).
No attempt was made to compare the dif-
ferent sites of the station, but this long record
of temperature measurements is without cloubt
not strictly homogeneous.
The monthly mean temperature lias in the
past been computed in different ways from
observations made at different hours. Since
1942, observations have been made 8 times a
day at 3 hour intervals. Since then the monthly
mean temperature has been computed as the
aritmetic mean of these observations and
should represent the true monthly mean. In
order to remove inhomogeneities due to differ-
ent methods in computing the monthly mean,
observations from 1956 to 1965 were used to
estimate the deviation of tlie various observa-
tion times from the true monthly mean. These
deviations are in this paper used to compute
new monthly means for the years 1846 to June
1873, and to correct the mean from July 1873
to 1941.
The monthly means should thus be comput-
ed in a comparable way for the whole period,
if one takes it as granted that the claily ampli-
tude has been the same in different screens
on different walls in the village which in the
reality cannot be expected.
The monthly mean temperature computed
in this way is here used to study temperature
TABLE 1
Stykkishólmur Weather Station
Temperature in free standing hut h- temperature in wall screen, °C
Febr. March April May June J uly Aug. Sept. Oct. Nov. Dec.
0.3 0.1 0.0 0.0 -0.1 -0.1 -0.1 -0.1 0.0 0.1 0.2
JÖKULL 19. ÁR 7
Jan.
0.3