Jökull - 01.12.1990, Side 163

sional microearthquake. Only two small events have
been large enough to be located.
Recent volcanic activity
beneath vatnajökull
The volcanic systems below Vatnajökull have been
quite active in historical times. Available informa-
hon on eruptions is compiled in Table I. The table is
to a large degree based on Þórarinsson (1974) who
compiled historical records on eruptions and jökul-
hlaups from Vatnajökull. The largest additions are
from Steinþórsson (1978), who analyzed ash layers
from an ice core drilled in Bárðarbunga, and Larsen
(1982), who studied ash layers from soil sections in
Jökuldalur, NE of Vatnajökull. Their work includes
cvidence for several eruptions that are not mentioned
in the written records.
The soil in JÖkuldalurcontains numerous ash lay-
ers from the period 890-1717, which were sampled
and analyzed. On the basis of their chemical composi-
hon, they were then traced to theirrespective volcanic
system. Nine layers were found to originate in the
Bárðarbunga system, among them the ash layers of
1477 and 1717 that previously had been attributed to
Kverkfjöll. Six layers had the chemical characteristics
of the Grímsvötn and the Kverkfjöll systems. Most of
them showed greater similarity to Grímsvötn, only
the layer from 1619 resembles Kverkfjöll somewhat.
The 415 m long ice core from Bárðarbunga contained
htany ash layers from the period since 1650 accord-
tng to Steinþórsson’s (1978) correlation. Besides the
Very characteristic Katla layers used for the correla-
hon, two distinct groups of ash layers were found, one
resembling Grímsvötn in composition, the other had
the more primitive character of the Bárðarbunga sys-
tem. In the light of the chemical characteristics of the
volcanic systems defined by the work of Jakobsson
(1979) and Larsen (1982) the analyses of ash layers in
the Bárðarbunga ice core can be reinterpreted. Thus
the layers of 1697, 1706, 1707, 1711, 1716, 1717,
1720, 1739, 1766, 1769 are here interpreted to origi-
tate in the Bárðarbunga system. It should be noted,
however, that the chemical characteristics of the Loki-
Fögrufjöll system are unknown. Some of these events
could therefore have taken place there.
Table I contains 63 certain and 13 questionable
volcanic events. The uneven distribution with time
indicates that many events have been missed, partic-
ularly in the early centuries. Since 1700, 53 events
have been documented, giving an eruption frequency
for this area of 18 events per century. We suspect
that further investigations of historical documents and
tephra layers in ice and soil will reveal more eruptions.
No volcano in Iceland has had a higher eruption
frequency than Grímsvötn (Þórarinsson, 1974). At
least 29 of the eruptions listed in Table I are asso-
ciated with the Grímsvötn system. At least one of
them accompanied a major rifting event, i.e. the
Skaftá fires 1783-1784. It involved a huge lava flow
from the Laki fissure, with a simultaneous eruption in
Grímsvötn (Sigurðsson and Sparks, 1978; Þórðarson,
1990). Eruptions apparently limited to the caldera re-
gion of Grímsvötn have occurred at least 10 times,
possibly as often as 18 times. There are also many
reports of volcanic eruptions involving other parts
of the volcanic system, to the N and NE, and the
swarm southwest of Grímsvötn towards Háabunga and
Þórðarhyma (1684-85, 1716, 1725-26, 1753, 1783-
84,1823,1867,1873,1883,1887,1902-03 and 1933-
39).
We find no evidence in the data for behaviour that
could be called steady-state. On the contrary, there
seem to be considerable fluctuations in the activity of
Grímsvötn, both in the eruptive and geothermal activ-
ity. The present quiescent interval following the events
in 1933-39 stands out. The events of 1941,1945 and
1954 involved only small, if any, magmatic activity,
and the events of 1983-84 were only small eruptions
(see Björnsson, 1988, p. 96-99). Thus the volcanic
activity of the last fifty years is low compared to the
century prior to 1939. These observations are in line
with the suggestion of Björnsson (1977, 1983) that a
significant shift in magmatic activity from Grímsvötn
to the Loki Ridge occurred in the mid-fifties. A com-
parable quiescent interval followed the Skaftá fires
1783-84 (i.e. to 1816 or 1823, 32 or 39 years).
Since the 1960’s, there has been a considerable re-
duction in geothermal activity in the Grímsvötn area.
At the same time, the area of the lake has diminished,
JÖKULL, No. 40, 1990 159