Jökull - 01.12.1977, Qupperneq 77

Fig. 4. Flow summation
curve íor all the jökul-
hlaups in Skaftá from
1955 to 1977.
Mynd 4. Samanlagt
vatnsmagn jökulhlaupa
d hverjum tlma frá 1955
til 1977.
Consider the hypothesis that thc burst water
is nielted by volcanic eruptions. The hypothesis
can be tested by estimating the volume of the
volcanic materials which are required for pro-
ducing the burst water. A minimum value for
this volume of volcanics can be obtained from
a model in wliich all the thermal energy of the
eruption is instantaneously transferred to melt-
tng of ice. According to this model of perfect
and instantaneous heat transfer the volume of
volcanic materials, say hyaloclastic materials
with the density of p = 2.5 g/cm3, whiclt would
have been produced during the jökulhlaups,
has varied from 3 • 10° m3 in 1973 to 15 • 10° m3
in 1964. The total volume of the hyaloclastic
materials erupted during 22 years would be
about 130 • 10° m3. This corresponds to a 130 m
thick layer of liyaloclastics piled up on a base
of 1 km2 area.
No field observations support the conclusion
that volcanic materials of the order of 100 • 108
m3 have been piled up beneath the ice cauldrons
since 1955. A more realistic model of heat
transfer would give a still higher estimate
of tlie volume of volcanics. Moreover, a tho-
rough investigation of the seismological activity
in Vatnajökull indicates that none of the jökul-
hlaups in Skaftá since 1955 are likely to be
related to volcanic eruptions. (Ragnar Stefáns-
son, pers. comm., 1973, Páll Einarsson, pers.
comrn., 1977). Finally, one can point out that
the increased runoff rate from the cauklrons
since 1955 does not indicate that a lava mass
is cooling down.
Consider the alternative that the burst water
is continuously melted at the glacier surface and
at a subglacial geothermal area. The total
volume of lturst water was 2.2 • 109 m3 in 22
years, or about 100 • 106 m3/year. The surface
ablation can account at the most for 20 • 106
m3/yr; that is, an average ablation rate of 400
mrn/yr over a water basin of, say, 50 km2.
Therefore, at least 80% of the burst water is
melted by the geothermal area. The strength
of the geothermal area should be of the order
of 800 MW. This power is an order of magni-
tude less than the estimate for the entire Gríms-
vötn geothermal region. But the average flux
density through an area of, say, 10—20 km2
would be of the same order of magnitude as the
estimate for the Grímsvötn and Kverkfjöll geo-
thermal regions (Bodvarsson 1961, Björnsson
1974, Friedman et al. 1973).
The present author has suggested that the ice
cauldrons are situated within the Grímsvötn
geothermal region (Björnsson 1974). According-
ly, jökulhlaups from both the ice cauldrons and
Grímsvötn woulcl be fed by a common geo-
thermal field. Estimated hydrological data sup-
port this suggestion. The runoff by jökulhlaups
in Skaftá has increased since 1955. At the same
time, the runoff from Grímsvötn has decreased.
The average runoff frorn the Grímsvötn area
alone in the present century has been estimated
as 660-10° m3/yr, or 0.66 km3/yr (Björnsson
1974). But since 1955 the total average runoff
front Grímsvötn and the ice cauldrons equals
JÖKULL 27. ÁR 75