Jökull - 01.12.1960, Qupperneq 20

EYSTEINN TRYGGVASON :
Earthquakes, jökulhlaups and subglacial eruptions
INTRODUCTION.
There are many records o£ volcanic eruptions
in the glaciers Vatnajökull and Mýrdalsjökull
connected with great jökulhlaups (glacier
itursts). There are also many records of similar
jökulhlaups from Vatnajökull without any con-
nection with visible volcanic eruptions. In case
of Grímsvötn eruptions, the jökulhlaups gene-
rally begin several days before the visible erup-
tion. This has led to the opinion, that the
jökulhlaup, by reducing the pressure on the
subglacial surface, is the primary cause of
Grímsvötn eruptions (Thorarinsson 1953). I£
this is right, the water must be stored over
long periods, several years in case of Gríms-
vötn, to escape suddenly when the amount of
water has reached some critical point. There
are evidences of such a process in Grímsvötn:
A large depression in the glacier, which can
serve as water reservoir, and thermal activity
continuously melting the glacier ice, in addition
to surface melting.
There are serious obstacles against generaliz-
ing this opinion to cover jökulhlaups from
Mýrdalsjökull. Here jökulhlaups are very rare,
exept in connection with eruptions in the sub-
glacial volcano Katla. The glacier surface cloes
not indicate any marked subglacial depression,
where the water can be storecl between jökul-
hlaups.
punktur og um leið lokapunktur þróunar þeirr-
ar jöklafrœði, sem nefna mœtti íslenzka i þeirri
merkingu, að hún byggðist að mestu leyti á
reynslu íslenzkra manna af islenzkum jöklum.
Þessi jöklafrœði þolir fyllilega samanburð við
þá, sem þróaðist samhliða og óháð í Alpalönd-
unum og sú islenzka var löngum feti framar i
þróuninni. En með 19. öldinni breytist þetta. Is-
lenzk jöklafraði staðnar, en þróunin í Alpalönd-
unum verður brátt örari, og smátt og smátl
verður fræðigreinin alþjóðleg. A síðustu áratug-
um hefur ísland aftur orðið vettvangur jökla-
rannsókna. — en það er önnur saga.
Seismic observational data support the opin-
ion, that jökulhlaups frorn Grímsvötn are the
primary cause of Grímsvötn eruptions, but
these data indicate a different relationship
between jökulhlaups and eruptions in Mýrdals-
jökull. Here the most probable physical rela-
tion between these two events is, that the heat
energy of the eruption produces the water of
the jökulhlaup by melting the glacier ice.
OBSERVATIONAL DATA.
We have seismograph records from Reykjavík
at times o£ several jökulhlaups from Icelandic
glaciers. During four of these jökulhlaups, small
earth tremors were recorded with epicenter in
or near the point of origin of the jökulhlaups.
Table I shows all recorded earthquakes, which
can be correlatecl with the jökulhlaups. During
the Skeidará-hlaup of 1913 or jökulhlaups dur-
ing the period 1926—1960, not mentioned in
Table I, no earthquakes were recorded. The
sensitivity of the seismographs in Reykjavík was
however very low prior to 1951, so earthquakes
of magnitucle less than 3i/2 to 4 were not de-
tected on the seismograms. After 1951 earth-
quakes of magnitude about 3 can be detected,
if the epicenter lies in Grímsvötn, and earth-
quakes from Mýrdalsjökull of magnitude as low
as 2]/2 are recorded, if conditions are favour-
able.
DISCEJSSION.
In a search for a physical relation between
earthquakes ancl jökulhlaups, the writer as-
sumes, that only the following five explana-
tions of this relationship are possible.
1. A subglacial volcanic eruption causes the
melting of the glacier ice, and the earth-
quakes are associated with the eruption.
2. The jökulhlaup is caused by emptying o£
an ice dammed lake, and the pressure re-
lease caused by the escape of a great mass
18