Jökull - 01.01.2019, Qupperneq 36

Historical accounts of pre-eruption seismicity of Katla, Hekla,
Öræfajökull and other volcanoes in Iceland
Páll Einarsson
Institute of Earth Sciences, University of Iceland, Sturlugata 7, 101 Reykjavík; palli@hi.is
https://doi.org/10.33799/jokull2019.69.035
Abstract — Detecting unusual activity leading to the outbreak of a volcanic eruption is of vital importance
for the short-term warning to the local population of an impending eruption. The varied volcano types of Ice-
land and range of ambient conditions at which eruptions occur offer an unusually wide spectrum of volcanic
phenomena and volcanic hazards during the initial phase of eruptions. A recent study of eruptions in Iceland
during the last four decades of instrumental observations has revealed that all the eruptions had a detectable
precursory seismic activity, that under favourable conditions can be used to issue short-term warnings to the
surrounding communities. Considerable documentary data also exist for pre-instrumental times extending sev-
eral centuries back in time, that can be compared to the instrumental experience. This is true in particular
for two of the most active volcanoes, Katla and Hekla, that are sufficiently close to the populated areas of the
country. All seven confirmed eruptions of Katla since 1625 were preceded by felt earthquakes, beginning one to
nine hours before the eruption was detected and two to over twelve hours before a jökulhlaup from this partly
sub-glacial volcano reached the inhabited areas. The behaviour of Hekla is quite different. Large eruptions
from the main edifice of Hekla since 1510 were generally accompanied by rather weak seismic activity. Earth-
quakes are usually felt only minutes before the first explosion occurs, in the 1947 case even several minutes
after the first explosion of the volcano. Eruptions of the Hekla volcanic system outside the main edifice are,
on the other hand, accompanied by considerable seismic activity, and the precursor times may be more than
three hours, even much longer. The two historical eruptions of Öræfajökull, in 1362 and 1727, were apparently
preceded by felt seismicity, sufficient to alarm the local population.
INTRODUCTION
It has been common knowledge in Iceland for cen-
turies that volcanic eruptions are preceded by earth-
quakes. Bishop Hannes Finnsson (1739-1796) com-
piled some of the historical, written documents on
eruptions and other catastrophes, and concluded by
suggesting that instruments such as barometers and
compass needles might be useful for the monitor-
ing of the activity of volcanoes. He also suggested
that ". . . observing the intensity and direction of seis-
mic shocks. . . " could be used to warn of impending
eruptions (Finnsson, in Thorarinsson (1967)). Even
though Finnsson was educated in Copenhagen where
he was influenced by many prominent natural scien-
tists of the Age of Enlightenment, it is remarkable that
these statements were made more than a century be-
fore seismographs were invented.
Ever since the settlement of Iceland in the ninth
century Icelanders have had to deal with the activity
of their thirty, or so, active volcanoes and live with
the threat of volcanic hazards such as ash fallout, lava
flows, and jökulhlaups (Gudmundsson et al., 2008).
The high rate of eruptions is due to a combination of
two processes, i.e. a divergence of two major lithos-
pheric plates taking place at a hotspot overlying an
upwelling mantle plume (e.g., Bjarnason, 2008). The
high magmatic activity generates a crust that is 2–4
times thicker than the normal oceanic crust produced
at adjacent segments of the Mid-Atlantic Ridge, the
Reykjanes and Kolbeinsey Ridges (Brandsdóttir and
Menke, 2008). The thickened crust allows the plate
boundary to rise above sea level, exposing the rift
volcanoes that are aligned along the boundary, such
as Grímsvötn, Bárðarbunga, Askja, and Krafla (Fig-
ure 1). Volcanoes also exist that are not directly re-
JÖKULL No. 69, 2019 35