Jökull - 01.12.1992, Blaðsíða 8

indicate that the October eruption was not an iso-
lated event and that some magmatic activity contin-
ued into 1962. The nature of this activity is unclear
but the earthquake swarms are similar to the swarms
recorded in the Krafla volcanic system during defla-
tion events, when magma, from a subcaldera magma
chamber, was laterally intruded into the Krafla fissure
swarm. A shallow magma intrusion within the Askja
caldera could cause widening within a limited region
and create new faults. Such an intrusion can also be
expected to increase the geothermal gradient and lower
the groundwater level.
DISCUSSION
Volcanic activity in the Northem Volcanic Zone
during the last three centuries has been confined to
the Krafla and Askja volcanic systems. The Krafla
volcanic system was active in 1724-1729 and a major
rifting episode has been in progress in Krafla since
1975. Three rifting events have taken place in the
Askja system during the last twelve decades, in 1874-
1876,1921-1933, and 1961-1962.
Scientific knowledge gained during the Krafla in-
trusive and eruptive activity casts new light on the
course of earlier events. For example, reports of felt
earthquake activity just prior to eruptions in Sveinagjá
and in the Krafla fissure swarm (Gjástykki) are strik-
ingly similar. Earthquakes were felt in the Mývatn dis-
trict just prior to or in the beginning of the Sveinagjá
eruptions on March 18, March 23, and August 15,
1875 but otherwise reports do not mention felt earth-
quakes during these eruptions. Drawing analogies
to seismic and various geodetic measurements dur-
ing the recent Krafla eruptions, we suggest that these
eruptions were preceded by intrusive activity during
which a magma propagated along the fissure swarm,
away from the shallow magma chamber beneath Askja
caldera towards the surface. The propagation was ac-
companied by strong seismic activity caused by dif-
ferential stress at the tip of the intrusion whereas the
seismicity decreased by orders of magnitude once the
intrusion surfaced in an eruption.
Reports of earthquake activity being strongest in
the beginning of the Askja rifting episode, (judging
from Guðmundsson, 1932) are in good agreement with
our current understanding of stress accumulation at di-
verging plate boundaries. The spreading at the bound-
ary is discontinuous in time and most of the extension
occurs in brief rifting episodes. The Krafla rifting epi-
sode was dominated by intrusive activity from 1975
to 1979, during which 10 intrusive events occurred
accompanied by minor eruptive activity (Tryggvason,
1984; Björnsson, 1985). The intrusive/extrusive ratio
changed after 1980, when most of the magma escaping
the magma reservoir surfaced in eruptions. Evidently
dike formation during the intrusive activity absorbed
the extensional capacity which had accumulated on
the plate boundary since the last rifting episode. The
Krafla fissure swarm was unable to accommodate fur-
ther intrusions and the excessive magma upwelling
into the magma chamber beneath the caldera had to
be extruded at the surface. Reports from people who
visited the Sveinagjá eruption sites in 1875 indicate
that the lava production increased with each eruption
(Thoroddsen, 1958) which could be taken as an in-
dicator of decreasing extensional capacity with time
within the fissure swarm.
The intrusion period during the Askja 1874-1876
rifting episode was shorter than during the recent
Krafla rifting episode. Considerable intrusive activity
took place during the first few months of the Askja rift-
ing whereas the Krafla fissure swarm accommodated
most of the intrusive activity during the first four years.
Apart from the Plinian eruption in the Askja caldera,
frequent eruptions in Sveinagjá dominated the Askja
activity. The Sveinagjá eruptions were more frequent
and may have had different lava extrusion rates com-
pared to the recent Krafla eruptions. Reports indicate
that the Sveinagjá eruption sites commonly shifted
along the fissure swarm and that high lava extrusion
rate continued for days whereas the Krafla-Gjástykki
eruptions had high lava extrusion rates in the first few
hours and soon concentrated in a few craters along a
small stretch of the original fissure where the eruption
would stay until the end.
The extrusion period during the Askja 1874-1876
rifting ep'isode was also shorter than during the recent
Krafla rifting episode. Eleven eruptions took place
in Sveinagjá during a 10 month period in 1875 and
the eruptions may have lasted 2 years. The erup-
6 JÖKULL, No. 42, 1992