Jökull - 01.12.1972, Blaðsíða 51

ing variations for the wincl-, temperature- and
humidity-fields.
For a horizontal surface and the x-axis direct-
ed along the wind direction the equation of
motion, the energy equation and the diffusion
equation for water vapor can be written re-
spectively (Lumley and Panofsky 1964)
3u "sT + ðu u_a?T “ ” 1 3P 1 3t p 3x p 3z (30)
3T 3t + 3T U "3xT = “ 1 3H a P CP 3z (31)
3m + 3m 1 3Hj (32)
3t U 3x _ p Lv 3z
As an illustrating example we can choose the
following typical values: Hd = 10~3 cal/cm2 s,
Hj = 6 • 10-4 cal/cm2 s, r = 1 dyn/cm2, u =
5 m/s and p = 1.2 • 10~3 g/cm3. Assuming the
geostrophic wind to be 10 m/s the constant
pressure gradient along the wind direction can
be estimated to I 3P/3x I '= 8 • 10-5 dyn/cm.
The following restrictions can then be made:
3u
aT
< 10-1 cm/s2
3T
öT
+ u
ST
< 7 • 10-4 deg/ s
3m
+ u
3m
< 1.7 • 10-? 1/s
Assuming steady state these inequalities give
per 1 km:
du
- - <0.2 m/s,
dx
dT
---- < 0.14 deg,
dx
dm
dx
< 0.3 • 10-4
(corresponds to 0.05 mb). Assuming, on the
other hand, horizontal homogenity we could
only tolerate following variations during one
yfSiilimii
Fig. 2. A view to north from the glacier down
the Bægisárdalur-valley. Fog covers the bottom
of the valleys, Bægisárdalur, Hörgárdalur and
Thorvaldsdalur. Some meteorological instru
ments, 1967, in the front.
Mynd 2. Séð niðnr af jökli norður Bœgisárdal
og Þorvaldsdal, en þvert á þá gengur Hörgár-
dalur. Þoka hylur dalbotnana. Fremst sjást
nokkur veðurathugunartœki sumarið 1967.
hour: 3.3 m/s, 2.3 deg and 0.55 • 10-® variation
in the relative humidity (or 0.8 mb).
A related problem is the question how ex-
tensive the quasi-homogeneous surface must be,
before the effects of the transient zones dis-
appear ancl stationary profiles are established
which are representative for the underlaying
surface. The minimum distance from the
disturbances in the surface to the observation
site seems to be for temperature profiles several
hundred times the lieight of the profiles and
somewhat less for the wind pofile (see Munn
1966, Dyer 1963, Swinbank 1963, Elliott 1958).
An observation site placed in the centre of a
melting snow surface can on many accounts be
an ideal site. The main difficulty of applica-
tion of the theory though usually arises from
the inclination of the glacier surface. No at-
tempt will be made to take this important
problem into consideration in the present
paper.
It seems to be difficult to give an exact an-
swer to the question of how long averaging
time intervals should be used in the non-linear
equations for the vertical fluxes considered (see
Cramer 1967). To secure effectively steady state
conditions the run period should be long com-
JÖKULL 22. ÁR 49