Jökull - 01.01.2013, Blaðsíða 110

F. Fuchs et al.
measured events suggests a common source mecha-
nism and might point to much tighter spatial cluster-
ing than observed by the epicenter distribution. The
spatial distribution of the epicenters and depths of
approximately 10 km indicate a magmatic origin for
the seismic activity. Lava flows around the Snæfells-
jökull have been mapped by Jóhannesson (1981). The
southern cluster of seismic events falls into the area
that was active during the most recent central caldera
eruption of Snæfellsjökull which covered the south-
ern flanks of the volcano. Similarly, seismicity in the
north-eastern cluster might originate from the magma
source that was active during the 4000-5000 BP erup-
tive phase. Few seismic events were located north-
west of the volcano where young flows associated
with local craters have also been found. Still, a denser
seismometer network is necessary to investigate the
spatial clustering of hypocenters with more detail.
Two models for the geometry of the plumbing
system of the Snæfellsjökull have been proposed by
Kokfelt et al. (2008) based on isotope data of vol-
canic rocks. The first model suggests a stratified and
long-lived magma chamber of conical shape, while
the second model proposes a simultaneously evolving
plumbing system made of smaller and isolated mag-
matic bodies of at least 3-4 km3 each. Both models
have in common that older and more evolved magmas
are situated at shallower depths beneath the central
summit, while younger magmas are located at greater
depths below the flanks of the volcano. We identify
three potentially separate sources of seismicity on the
Snæfellsjökull volcano, with two main clusters on the
southern and north-eastern flanks, respectively and
few events located in the lower lands north-west of the
volcanic edifice (Figure 2). The origin of seismicity is
in agreement with both models proposed by Kokfelt
et al. (2008) and suggests that the plumbing system
of the Snæfellsjökull volcano extends from about 8
to 13 km depth. However, uncertainties on epicenters
and depth are too large to rule out either of the two
suggested scenarios. Based on seismic and geodetic
data, magma chambers beneath active Icelandic vol-
canoes are confined to the upper crust (Brandsdóttir
and Menke, 2008; Mitchell et al., 2013).
The absence of high frequency components in the
waveforms and the swarm-like appearance of seis-
micity suggest that seismic activity could be associ-
ated with fluid processes. However, seismicity that
has been associated with fluid-induced mechanisms at
e.g. Eyjafjallajökull (Tarasewicz et al., 2012) or Up-
ptyppingar (White et al., 2011) volcano shows higher
frequencies and is termed to be of volcano-tectonic
origin. Similar low-frequency bands have been ob-
served at Askja volcano but were interpreted as a re-
sult of absorption of high frequencies by Soosalu et al.
(2009). Additionally, the frequency content of the sig-
nals we measure underneath Snæfellsjökull is too high
and the signal duration too short to be purely fluid-
related such as long period events (Chouet, 1996).
Regarding the thin crust of approximately only 15–
20 km under the Snæfellsnes, our estimated hypocen-
tral depths may fall within the normally ductile part
of the crust. White et al. (2011) argue that brittle fail-
ure within the ductile regime can be generated by high
strain rates induced locally by magma movement. We
therefore propose that seismic activity underneath the
Snæfellsjökull volcano reflects hybrid events that in-
volve brittle failure in combination with fluid-related
processes. Although scattering and absorption of the
seismic waves may contribute to the absence of higher
frequency signals, we favour the scenario where seis-
micity is related to fluid-induced brittle failure in the
deeper part of the crust.
Sigurdsson (1970) suggested that seismic and vol-
canic activity in the SVB could be increasing as
present day spreading rates are comparable to the late
Pleistocene ones which might have formed the en-
echelon alignment of volcanic structures (Árnadóttir
et al., 2009). In this case one would also expect tec-
tonic events at shallower depths that follow the orien-
tation of the inferred transcurrent fault. However, the
origin of recent volcanism and its distribution along
the Snæfellsnes peninsula is poorly understood and
the proposed fault remains hypothetical. The instal-
lation of continuous GPS stations associated with a
larger array of permanent seismic stations could help
to determine the current tectonic deformation of the
Snæfellsnes peninsula and to reveal the origins of its
rejuvenated volcanism.
110 JÖKULL No. 63, 2013