Jökull - 01.12.1972, Blaðsíða 54

The arithmetric mean of several measurements
in the middle of the day, when the insolation
usually was strongest, was used in the calcula-
tions. These values represent the change in the
albedo due to changes in the surface and. were
fairly similar for the whole glacier which was
covered with snow for the whole period. For
the observation period of 1967 and 1968 the
average albedo values were 0.63 and 0.54 re-
spectively. This marked difference was due to
more dirty snow surface in 1968 and lieavier
snowfall during the summer of 1967. The ab-
sorbed short-wave radiation is given in Fig. 5
and 6.
Cloudiness in octas of the visible sky was
estimated every hour during the day time and
sometimes in the night. Mean values during the
night were taken equal to the arithmetic mean
of two values before and after a breake in the
observation series. Cloudiness was only taken
equal to 7 to 8 when there was clense overcast
of a radiation temperature exceeding that of
the melting snow. Results found by eq. (3) are
given in Fig. 5 ancl 6.
Daily observations were taken of the snow
surface lowering at the ablation stakes and the
snow density was measured regularly cluring the
observation period. No measurements of the
free water content could be done.
DISCUSSION OF THE
CALCULATION PROCEDURES
Calculations of the turbulent fluxes were
only made on days when the air temperature
was positive more than 15 hours and when the
turbulence conditions satisfied the requirement.
L > 5 m; which corresponds to Ri < 0.12 at 2 m.
Sources and sinks connected with freezing of
surface layers could not be taken into considera-
tion because measurements of the free water
content of the snow could not be performed.
An attempt was made to calculate the tur-
bulent energy fluxes for 3 hr. intervals. These
calculations revealed that the requirements
assumed for the turbulence conditions were not
fulfilled for a good deal of the 3-hr. intervals.
Day values for the fluxes could therefore not
be found for every day by summing up 3-hr.
values. On the other hand it appeared that 24-
hr. values could be computed directly from the
52 JÖKULL 22. ÁR
mean daily gradients for almost every day.
These values can however not be quite re-
presentative for processes which seem to take
place under so stable conditions that the re-
quirements for application of the theory are
not fulfilled. Further, without regard to the
stability conclitions, 3-hr. gradients should be
better expressive than mean daily gradients for
the non-linear processes considered. Comparison
of values calculated by these two procedures
revealed that the directly computed 24-hr.
values were on the average 10 to 15 percent
higlier than the sum of 3-hr. fluxes. As a
curiosity it can be mentioned that the sum of
3-hr. enthalpy fluxes, calculated without any
limits for the stability conditions, was found to
be 1.15 times that given for 24-hr. values dur-
ing the observations period in 1968.
In the view of the criticism of the directly
computed 24-hr. values one can not assume that
they give a physically correct picture of the
processes. Exact outcomes for single days can
not be expected. On the other hand they are
believed to be mutually comparative to the
extent that they give a qualitative picture of
the long tirne variation.
During the sumrner of 1967 the air tempera-
ture and the vapor pressure were only measur-
ed in height of 2 m above the surface and at
the height zo = 1.5 mm they were assumed to
be to = 0.1 °C and eo = 6.1 mb. This estima-
tion of the integration coefficients zo, z'o, z"0
in eq. (12) and (13) failed however and the
calculated values for the fluxes were highly
overestimated, especially on days with high wind
speed. Therefore results from 1968 are mainly
used in the following discussion.
Daily values of energy balance components
are graphed in Fig. 5 and Fig. 6. Relative im-
portance of energy balance terms in 1968 to-
gether with measured and calculated ablation
and daily course of climatological elements are
shown in Fig. 7. There is as expected a con-
siderable discrepancy between daily values for
calculated and measured ablation. Table 1 con-
tains a summary of these results for various
periods. Calculated and measured ablation is
shown to be in good accordance for the periods
considered.