Jökull - 01.12.1953, Page 19

ípfejBi
Fig. 8. The south-western-
most part of the Gríms-
vötn depression on 18
Aug. 1953.
Suðvestasti hluti Grims-
vatnalœgðarinnar séður úr
lofti 18. ágúst 1953.
Aerial photograph
by S. Þórarinsson.
height of the damming ice barrier. Besides, the
glacier bursts have been irregular during the
last decades. During the 1940’s we had, instead
of one glacier burst of normal magnitude, three
smaller ones, which, however, did approximate-
ly discharge together the same quantity of water
as one normal glacier burst. This irregularity is
probably at least partly due to the recent
thinning of the damming glacier. According to
a seismic sounding carried out by A. Joset and
now characterized as reliable (Joset 1952, p. 43
and Pl. XI) the height difference between the
firn surface and the bottom of the SW part of
the depression ab. 1 km off the Grímsfjall cliff
was ab. 580 m on April 7, 1951. We know that
the ice surface in the S part of the depression
was ab. 400 m lower than Svíahnúkur eystri or
in ab. 1300 m height after the glacier burst in
1934 (Nielsen 1937, Jonas 1948). If the seismic
sounding is reliable, which I am inclined to
doubt, the consequence is that the volume of
the depression beneath the 1400 m level is at
least 7 km3, and probably considerably more.
Another consequence is that either is the ice in
the depression at least 500 m thick or there is a
permanent lake in the depression, dammed by
a subglacial rock barrier, which seems less prob-
able. The establishing of the depression being
so deep further means that the craters of the
1934 eruption were situated on a steep slope
more than 500 m above the bottom of the de-
pression. But it is really very difficult to explain
tlie plane and horizontal surface of the ice in
the SW part of the depression and its regular
rising between the glacier bursts, without assum-
ing that the ice is floating on water, the level of
which is continuously raised between the glacier
bursts. Furthermore it seems impossible that
more than a small amount of the total discharge
of the glacier burst in 1934 can have been due
to melting by the eruptions in the visible craters
then active ( cf. Thorarinsson 1953). And even
if it is possible that other craters were then active
at lower levels within the depression, beneath
the ice cover, it seems unlikely that a great part
of the meltwater discharged by the glacier burst
can be due to their activity. It seems thus most
probable that the main part of the melt-water
of a glacier burst is accumulated in the depress-
ion between the eruptions and is due partly to
ablation and partly to continuous subglacial
melting caused by thermal activity.
According to Eythórsson’s and Rist’s measure-
ments in April 1951, the surface of the damming
ice barrier was 105 to 125 m higher than the
water level in the depression at that time.
Assuming that the water level in Grímsvötn
was raised 50 m from its height on 1 July 1953,
17