Jökull - 01.12.1989, Qupperneq 35

Fig. 1. Seismicity and plate boundaries in the Ice-
land area. Spreading north and south of Iceland
occurs along the Kolbeinsey and Reykjanes Ridges,
respectively. In Iceland the plate boundary is dis-
placed to the east by two fracture zones. Epicenters
are from the PDE listings of the USGS for the period
1960-1983.
Mynd 1. Skjálftar og flekaskil á Islandi. Gliðnun á
sér stað á neðansjávarhryggjunum, á Kolbeinseyjar-
hrygg norðan við landið en á Reykjaneshrygg sunn-
an þess. A Norður- og Suðurlandi hliðrast flekaskilin
tH austurs. Hliðrunarsvœðin einkennast af sniðgeng-
ishreyfingum, og þar verða skjálftar stœrri en ann-
ars staðar. Skjálftaupptökin eru merkt með deplum,
en gögnin eru fengin úr listum Jarðfrœðistofnunar
Bandaríkjanna fyrir árin 1960-1983.
assumed transcurrent motion of the plates on either
side of the South Iceland Seismic Zone gradually
builds up strain in the zone. The strained zone is at
least as wide as the seismic zone as delineated by
epicenters and surface faulting in Postglacial earth-
quakes (Fig. 2). Historical data indicate that the
strain is released in earthquake sequences with
recurrence intervals averaging 80-100 years (Einars-
son and others, 1981). A typical sequence begins in
the eastem part of the zone with an event of magni-
tude (Ms) about 7, followed by smaller events in the
westem part. The duration of a sequence can vary
from a few days to a few years. The last major
sequence occurred in 1896 so the next earthquakes
are expected with high probability within the next
20 years.
A modest effort in earthquake prediction research
has been initiated in South Iceland during the last
decade or so (Einarsson, 1985). Some of the experi-
ments have already shown very promising results, in
particular the radon monitoring program (Hauksson
and Goddard, 1981). In this paper we report the first
results of an experiment to monitor crustal move-
ment with repeated distance measurements in this
zone. Small geodetic figures were installed in three
different parts of the seismic zone, in the Flói district
in 1977, Holt district in 1979, and Ölfus district in
1981 (Fig. 2). The easternmost one is in the most
active part of the zone, which is also the area with
the highest probability for the next large event (Ein-
arsson, 1985; Stefánsson and Halldórsson, 1988).
The figures almost span the width of the zone of
strain accumulation as shown by epicenters of earth-
quakes and surface fracturing (Fig. 2), and may
therefore be expected to show a large part of the
plate motion between successive measurements.
Lines trending N-S and E-W are not expected to
show any change, but lines crossing the zone
obliquely should lengthen or shorten, depending on
which way they cross the boundary. The changes
should be of the order of a centimeter per year if the
strain accumulation is continuous.
These figures were remeasured in 1983-1984,
when the accumulated crustal movements were
expected to have reached measureable levels. At the
same time a more extensive geodetic network was
installed in the seismic zone,. enclosing thé previous
geodetic figures (Thorbergsson, 1985; Erlingsson
and Einarsson, 1985). The remeasurements show
that measureable movements have occurred, but
they do not show the regular pattem one might
expect along the boundaries of continuously moving
plates. Strain build-up along plate boundaries is
clearly not as simple as has sometimes been
assumed or hoped.
JÖKULL, No. 39, 1989 33