Fróðskaparrit - 01.01.2002, Blaðsíða 103

ARKTISKA FJALLAVEÐRIÐ í 2000 Á SORNFELLI
101
data are quality controlled and stored in
a database. Plots of the most important
parameters are shown on the website
http://www.metsupport.dk/data/sornfelli.
These plots are updated hourly. Data are
also published on the website of the Faro-
ese Office of Public Works,
http://www.lv.fo/egl_en/real_sornf/
sornfnow.html,
which also has photographs from the prepa-
ration of the site during 1999.
The instrument cylinder is continuously
monitored by 4 internal temperature sen-
sors and 15 other control parameters, en-
abling remote inspection of the overall op-
eration. Since the start not many data have
been lost. This is mainly due to the remote
inspection, which enables dealing remotely
with issues like turning on more heat in the
lid if snow accumulated there, checking the
lifting ntechanism, checking that the đrain
in the bottom of the cylinder is open, and
being able to call for local assistance, sup-
plied by LV, and the military personnel at
Sornfelli, when needed.
Air temperature
Four different air temperature values are
recorded at the Sornfelli meteorological
station. A shielded and an unshielded mean
air temperature from the platform sensors
(Fig. 5), which are granted 9 minutes for
acclimatization, then one minute of record-
ing before the platform is lowered into the
cylinder. A shielded 30 min mean air tem-
perature from the temperature sensor fixed
on the external boom (Fig. 5), and a shield-
ed I min mean air temperature from the
same sensor, recorded during the same
minute as the platform sensors record.
In the year 2000, one minute mean tem-
perature values from all three sensors
recorded simultaneously, showed the un-
shielded sensor to be warmest, with a
MAAT of 2.18°C, while the shielded sen-
sor fixed on the boom was coldest with a
MAAT of 1.76°C. The shielded sensor on
the platform had a MAAT of 1,83°C.
These three measurements are combined
into one corrected annual air temperature
(Fig. 6) using the following two rules: 1)
Use the coldest sensor. 2) Use data from the
coldest platform sensor, if the shielded
fíxed sensor is in the interval +0.2°C to
-0.8°C. The first rule both prevents the use
of data from the shielded platform sensor,
when it is not sufficiently cooled in calm
weather, and generally prevents the use of
data from the unshielded sensor. The sec-
ond rule prevents using the fixed sensor,
when icing keeps the temperature stable
during freezing or thawing.
Using these simple rules the shielded
platform sensor contributed 61 % to the data
of the corrected air temperature in 2000,
while 37% came from the fixed outside
sensor, and only 2% from the unshielded
platform sensor. The corrected air tempera-
ture (Fig. 6) for year 2000 has a MAAT of
only 1.71°C. The annual amplitude was
only 8.7°C (Table 2), based on monthly val-
ues. The warmest month was August with
6.5°C, while April was coldest with -2.2°C
(Table 2). Analysing the distribution of data
used from the different sensors for the con-
structed air temperature curve, shows that
the unshielded sensor was used only during
calm nights, when both of the shielded sen-