Jökull - 01.12.1972, Blaðsíða 52

Fig. 3. Some meteorological instruments in
1967. The mountain Jökulborg.
Mynd 3. Nokkur yeðurathugunartœki sumarið[
1967. Jökulborg i baksýn.
pared with the period of intermittency due to
e.g. variations in the cloudiness, as pointed out
by Swinbank (1964). On the other hand it
should be a good deal less than the time inter-
vals which correspond to the frequencies of the
synoptic components in the spectrum of the
weather parameters considered. An averaging
time of öne to some few hours seems to be
plausible and it corresponds to the observed
spectral gap for the considered parameters (see
Lumley and Panofsky 1966, p. 43).
The micrometeorological observation site was
placecl near the centre of the glacier in the
accumulation area at 1100 m height (see Björns-
so?i 1971). In this area the surface is almost
horizontal in the E-W direction but inclined
to N by about 10° over 500 m. Due to the
surrounding orography and the varying in-
clination of the glacier surface as a whole the
present observation site was far from an ideal
one for measurements of the turbulent fluxes.
The results can therefore not be interpreted as
strictly physical but should be considered more
like an estimate in the statistical meaning of
the word.
INSTRUMENTATION
AND OBSERVATIONS
Wind
The average velocity of wind was registrated
by a cup-anemometer (Woelfle, Type 1482)
placed at a height of 2 m above the snow sur-
face. Daily means have been published by
Björnsson (1971; fig. 5 and 6). The roughness
parameter was measured by taking wincl pro-
files by 3 cup-anemometers (Fuess, 212.2) ex-
posed at 0.5 m, 1.0 m and 2.0 m height. The
anemometers were compared inter se on a hori-
zontal board. To avoid systematic errors due to
errors in the calibration, the anemometers were
interchanged systematically during the observa-
tion period.
Wind profiles were measured several times
every day over a time interval of approx. 30
minutes. Calculated values of the roughness
parameter zo frorn eq. (17) and eq. (21) proved
to be in a rather poor agreement. Near-loga-
rithmic profiles with wind speed over 1 m/s
were therefore selected from a u-log z paper.
Theoretically the wind profile is almost loga-
rithmic for small heights. Therefore the rough-
ness parameters were calculated by extrapolating
the logarithmic wind profile through the low-
est two points in these profiles. For the heights
zi = 0.5 m and Z2 = 1.0 m the roughness para-
meter is given by
U2
ln zo =-----------ln l/a
u2 — ui
where the wind speed ui corresponds to the
height zi and u2 to z2.
During the summer 1967 the roughness para.
meter was found to vary only slightly and the
surface was essentially not changing during the
whole period of observation. The arithmetic
mean zo = 1.5 mm was used in the calculations.
In 1968 the roughness parameter showed a
systematic increase from 2 to 5 mm during the
observation period. For average velocity above
1 m/s the observed roughness parameter was
large enough to maintain a fully turbulent
current.
Température and liumidity
The air temperature and relative humidity
were observed with artifically ventilated psy-
chrometers of the Assmann type and recorded
on a thermo-hygrograph in an instrument screen
(type Linke) at 50 cm and 200 cm height (only
at 2 m 1967). The instrumental error of the
termo-hygrographs was found by comparison
with maximum and minimum and dry and wet-
50 JÖKULL 22. ÁR