Jökull - 01.12.1968, Page 33

evaluated such measurements from the rivers
Nidelven and Glomma in Norway. On the basis
of the technical literature available at that
time he assumed that the wind function was
on the form (v + 0.3) °-5. Devik found the fol-
lowing approximate expressions for the heat
loss by evaporation and convection:
s2 = 2.25 (v + 0.3)0-5 (ew - ea) (31)
s8=1.4(v + 0.3)o.5(tw-t.) (32)
where s is in Mcal km-2 sec-1, (Mcal = 106 cal),
e in mb, t in °C and v in m sec-1. The wind
velocity at the time of the heat loss measure-
ments was rather low, or 3.9 m sec-1 on the
average.
In newer evaporation formula the exponent
n is higher than 0.5 and in many formulas it
is 1.0. In order to investigate this question
measurements of the heat loss from calorimeters
with 0.09 m2 water surface were undertaken
at Tangafoss, a temporary meteorological stat-
ion in the Thjorsa Basin, in 1965. For air
temperatures from 0 to -t- 15 °C and p ~ 960
mb the following approximate expressions for
the rate of heat loss from the calorimeters by
evaporation and convection were obtained:
s2= 1.9 v6°-845 (ew-ea) (33)
s3 = 1.2 veo.8« (tw - ta) (33)
where vo is the wincl velocity in m sec-1 at 6
metres above the ground, e in mb, t in °C ancl
s in Mcal km-2 sec-1. A wind function of the
form (a + bv) could just as well be fitted to
the measured points and indeed this form is
most common in evaporation formulas. The
wind velocity, V6, during the measurements
varied from 2.5 to 13 m sec-1. It is not to be
expected that formulas derived from measure-
ments in calorimeters will apply to rivers and
most likely they give too high a rate of heat
loss. Nevertheless we have used the above ex-
pressions together with (20) for computations
of heat loss in the Thjorsa Basin, but of course
the formulas will be revised when material
from direct observations becomes available. The
formulas (33) and (34) give a considerably
liigher rate of heat loss than most other heat
loss formulas, for example the Russian formulas
used by Dingman, Weeks and Yen (1968). It
should be noted that where we have used (33>
and (34) the wind velocity is measured at the
rivers and that wind velocities are rather high
(monthly means of v«: 6 to 12 m sec-1). It
must be emphasized that different formulas for
the rate of heat loss by evaporation and con-
vection are not comparable if the actual wind
velocity over the water surface is not used.
PRACTICAL APPLICATIONS
AND EXPERIMENTAL CHECKS.
In investigations of river ice in the Thjorsa
Basin the heat loss formulas have been used
for climatological studies and calculations of
ice production and ice conditions.
For comparison of the ice-forming intensity
between various frost periods and from year
to year daily means of the heat loss from a
water surface at 0 °C have been calculated.
Daily means of the meteorological factors wrere
used and the calculations were done with an
electronic computer. With this method all the
relevant meteorological factors except preci-
pitation and drifting snow are taken into ac-
count.
Characteristic features of the ice conditions
of Thjorsa River and its tributaries are large
areas of open water which remain throughout
the winter. The formation of an ice cover is
here hindered by swift currents and inflow of
warm groundwater is also significant in some
reaches. During frost periods immense quanti-
ties of frazil ice are produced in the open
water and carried downstream to low velocity
reaches where it forms huge ice jams. In con-
nection with the hydro-electric development of
the rivers a knowledge of the frazil-ice dis-
charge is of great importance at some sites.
When the ice production is substantial, anchor
ice and border ice is only a minor part of the
total ice production and accordingly the frazil
ice discharge can be calculated from meteoro-
logical observations by rneans of the heat loss
formulas provided the size of the open-water
area upstream from the site is known. The
open-water areas can be determined from aerial
photographs but as this is rather expensive and
time-consuming we are trying to find a correla-
JÖKULL 18. ÁR 367