Jökull - 01.01.2015, Side 57

Seismicity beneath Þeistareykir, NE-Iceland
wave reflected phase off the crust-mantle boundary at
approximately 23 km depth.
Figure 1a shows moment-tensor solutions for
three deep events. Solution 1 (16–17 km depth b.s.l.,
seismograms in Figure 2) may represent a closing
crack. However, if we include additional uncertainty
in the source inversion, the solution pattern resembles
a double-couple solution with one nodal plane closely
aligned with the north-south direction. Solution 2 (15
km depth b.s.l.) seems to represent an opening crack
with a significant non-double-couple component. The
third solution (9 km depth b.s.l.), albeit constrained by
limited azimuthal coverage, suggests a double-couple
source mechanism with one nodal plane oriented ap-
proximately in the NE-SW direction.
We estimated the b-value within the area shown in
Figure 1a using the formula by Aki (1965) and Bender
(1983), which is based on the magnitude-frequency
distribution of earthquakes. We included data from
the IMO catalogue between 2009–2014. b-values be-
low 1 (0.76±0.13) are estimated and representative
for normal crust and volcanic areas. Similar values
were found elsewhere along the Icelandic rift zone
(e.g. Panzera et al., 2016). Here, magnitude binwidths
of up to 0.5 were tested, the magnitude of complete-
ness determined by the maximum curvature method
and data prior to 2009 disregarded due to the network
sparsity.
The center of the shallow earthquake cluster is
located at the west-northwestern part of the Bæjar-
fjall mountain at 2–5 km depth b.s.l. and just out-
side the limit of highly altered rocks mapped at the
surface (Ármannsson, 2014). The shallow earthquake
cluster correlates with seismicity reported by the SIL
network (Hjaltadóttir and Vogfjörð, 2011) and tem-
porary network (Vogfjörð, 2000). A relatively sharp
base of the seismicity and spatial event clustering has
also been found in Krafla and Askja beneath geother-
mal fields and above velocity anomalies indicating
the presence of melt (Schuler et al., 2015; Mitchell
et al., 2013). We also interpret the shallow events
beneath Þeistareykir to be linked to a shallow heat
source, where earthquakes might be triggered by hot
circulating fluids propagating along faults. For some
shallow and well-constrained events, we performed
source inversions to see whether they fit the double-
couple (shear-fault) source model. In Figure 1b, we
show three double-couple examples all with one nodal
plane closely aligned in NE-SW direction following
the fissure swarm.
Results of the TEM/MT inversions by Karlsdóttir
et al. (2012) do not show a low-resistivity anomaly
directly beneath our shallow earthquake cluster at the
west-northwestern part of Bæjarfjall, which would
have supported our interpretation that a shallow heat
source and hot geothermal fluids cause the increased
seismicity there. The closest low-resistivity anomaly
imaged lies beneath the southeastern part of Bæjar-
fjall at 5–6 km b.s.l. The area of deep earthquakes is
located outside the TEM/MT study area.
CONCLUSIONS
Microearthquakes located around Þeistareykir using a
local seismic array reveal that the seismicity is con-
centrated at 2–5 km depth b.s.l. beneath the northwest-
ern end of Bæjarfjall mountain. Earthquakes south-
east of Þeistareykir are scattered over a large area, at
8–20 km depth b.s.l. They are either linked with high
strain rates or reduced normal friction caused by melt
movement within the lower crust.
ACKNOWLEDGEMENTS
Seismometers of the Cambridge array were bor-
rowed from the Natural Environment Research Coun-
cil SEIS-UK under loan 891. The Icelandic Meteo-
rological Office kindly provided additional data from
stations DIM, GHA, GHS, KVO, MEL, REN, and
SKI. Funding was provided to RSW by a grant from
the Natural Environment Research Council. JS grate-
fully acknowledges support by the Swiss National
Science Foundation. We thank David Pugh for help-
ing to calculate fault plane and moment tensor solu-
tions. Generic Mapping Tools (Wessel et al., 2013)
were used to make the figures and the Python frame-
work ObsPy (Beyreuther et al, 2010) was used to ana-
lyze our data. Constructive comments by two anony-
mous reviewers helped improve the paper.
JÖKULL No. 65, 2015 57