Jökull - 01.12.1991, Qupperneq 40

(Þórarinsson, 1955). The thickness of the winter
accumulation was of the order of 4 m, suggesting that
the layer was dispersed over the area in mid winter.
Fresh ash layers within Grímsvötn have a black
colour when observed in snowpits and crevasses.
Hence, the brown colour indicates that the layer was
created when fragments of hyaloclastite tuffs where
dispersed by a steam explosion.
SHORTINTERVALS BETWEEN JÖKULHLA UPS
The unusually short time interval between the jökul-
hlaups of 1939, 1941, 1945 and 1948 has been cited
as indirect evidence for eruptions in the Grímsvötn
area (Jóhannesson, 1983; 1984; Grönvold and
Jóhannesson, 1984). As far as we are aware, no
observations were made in Grímsvötn in 1939.
During the jökulhlaup in 1941, however, Grímsvötn
were inspected from air by P. Hannesson (Þórarins-
son, 1974). He recorded no unusual activity in
Grímsvötn, only crevasses around the periphery of
the depression indicating that the central ice shelf
was subsiding. The jökulhlaup in 1948 reached a
maximum on February 23 about a month after its
start. On the morning of February 20 a fallout of
gray-brown ash was recorded on a fishing boat off
the SE coast, about 120 km to the SE of Grímsvötn
but ash was not detected elsewhere (Þórarinsson,
1974). Grímsvötn were inspected from the air on
February 22, 1948, (Figs. 13b, c) but no signs of vol-
canic activity were observed (Þórarinsson, 1974).
Whatever the nature of the gray-brown ash detected
off the SE coast, its origin was not in Grímsvötn.
Björnsson, (1988, p. 84), however, suggested that
increased geothermal activity following the eruption
of 1938 caused the frequent jökulhlaups in the
1940’s.
Effects ofthe 1938 eruption on
jökulhlaup frequency
The main volcanic event in Grímsvötn since 1934
was the fissure eruption in 1938 and the high fre-
quency of jökulhlaups in the period 1938-1948 may
be explained by melting of ice at the emption site.
The volume of the ridge created in the eruption is of
the order of 0.4 km3 and assuming that it is made of
hyaloclastites, its mean density may be close to 2000
kg/m3 which puts the total mass of the erupted mate-
rial to 8.0- 10n kg. The maximum amount of melting
during the eruption would occur if all the magma
were quenched as basaltic glass. The heat released
during the eruption would then be given by
E=mC(T0-T!) where m is the total mass of the empt-
ed material, C is the specific heat capacity of the
glass, T0 is the initial temperature of the magma and
T, is the ambient temperature at the eruption site
(Tj=0°C). By using T0=l 150°C (typical for tholeiitic
magma, Williams and McBimey, 1979) and C=1000
J/kg°C (Allen, 1980), the total heat released would be
9.2- 1017 J. Using the latent heat of fusion for water,
Lw=3.35-105 J/kg, the total mass of ice melted during
the eruption is 2.7-1012 kg or 2.7 km3 of water,
which is similar to the volume of the depression
formed in the glacier surface in 1938 (Fig. 9). If pil-
lows make up a significant part of the rock volume
the rate of melting during the emption would be
somewhat less as the heat would be released over a
longer period of time.
The quenching of the magma and the formation of
glass results in the latent heat of fusion of the magma
not being released upon cooling but gradually during
the alteration of the glass to palagonite. According to
experience from Surtsey, the palagonitization was
well advanced in a large fraction of the rock volume
12 years after the eruption that formed the hyalo-
clastites (Jakobsson and Moore, 1982). Steinþórsson
and Oskarsson (1986) estimated that the heat released
during palagonitization of basaltic glass is about
4.2- 105 J/kg. Thus, the total energy released in the
palagonitization of a mass of 8.0-1011 kg is 3.4-1017 J
which would melt 1.0-1012 kg of ice which is equiva-
lent to a water volume of 1.0 km3. The total volume
of meltwater would therefore be 2.7+1.0 km3 = 3.7
km3. Crystallization and cooling of the feeder dyke
will also cause melting. For example, a feeder that is
2 km high, 8 km long and 2 m wide (i.e. 0.032 km3)
would give away heat that melts about 0.4 km3 of
water. The total volume of water melted gradually
over several years is therefore of the order of 1.4 km3.
The meltwater formed above the ridge after the
eruption of 1938 either drained continuously to
38 JÖKULL, No. 41, 1991