Jökull - 01.01.2010, Qupperneq 38

Brandsdóttir et al.
Conjugate faulting is most pronounced northeast
and southwest of the western fault (Figures 5 and 7),
similar to conjugate faulting at the southern tip of
the western SISZ2000 faults (Clifton and Einarsson,
2005; Hjaltadóttir and Vogfjörð, 2005). The north-
ern, shallower ends of both 2000 faults ruptured a
series of NNE-trending conjugate segments whereas
similar SW striking forks are only observed along
the southern, deeper, end of the western (Reykjafjall)
fault (Figures 10 and 12–16). Scattered activity west
and northwest of the northern end of the western fault
may be related to short N-S or NW-SE striking fault
strands. In spite of the two Ölfus events being of a
smaller magnitude than the SISZ2000 main events,
the aftershock zone of the 2008 events was of similar
length and depth as the June 2000 faults. In particu-
lar, the 16.5 km long 2000 Hestvatn fault (Hjaltadóttir
and Vogfjörð, 2005) is strikingly similar to the ∼17
km Reykjafjall fault, being divided into two sections,
deepening southwards.
Frictional failure on critically stressed faults has
been observed under the Coulomb failure criteria in
geothermal and volcanic areas following large earth-
quakes (i.e. Peng et al., 2010). The overall distribu-
tion of aftershocks along the main faults as well as
triggered activity across the EW zone, coincides with
the regions of highest Coulomb failure stress changes
for two paired N-S right-lateral strike-slip faults (De-
criem et al., 2010). Aftershock activity following the
Mw 6.5 June 2000 events also preferentially occurred
in the dilatational (northeast and southwest) quadrants
with postseismic deformation which extended about 5
km from the two main shock ruptures over a period of
two months (Árnadóttir et al., 2005). Whereas the
short-term deformation has been explained by poro-
elastic rebound due to postearthquake pore pressure
changes (Jónsson et al., 2003) a year-scale deforma-
tion is explained by afterslip at 8–14 km depth or
lower crustal viscoelastic relaxation (Árnadóttir et al.,
2005). Kinematic modelling based on geodetic mea-
surements further suggest that the SISZ is a complex
zone of N-S surface faulting driven by an E-W left-
lateral shear below 15–20 km depth with a deep slip
rate of 19 mm/yr (Árnadóttir et al., 2006). The over-
all distribution of the 2008 aftershock zone and fault
plane solutions are consistent with both strike-slip
and normal earthquakes being generated by oblique
movement with respect to the 103◦ spreading direc-
tion from NUVEL-1A concurrent with volcanic fis-
sure swarms within the RPRZ and WVZ. Further-
more, triggered events west of the main faults, may
indicate reloading by afterslip making it difficult to
infer the mechanisms responsible for the earthquake
triggering on the basis of short-term stress changes.
CONCLUSIONS
Using a new automated CMM technique a total of
19450 events, recorded on a 14-station network were
located during the period May 30–July 2. Filtering
based on SN ratios and location errors resulted in
7846 usable event locations.
Earthquake aftershocks delineate two major 12–
17 km long, right lateral, strike slip faults, which rup-
ture to greater depth (9 km) in the south than in the
north (1–6 km). Most of the located events lie along
the two main N-S faults, but the aftershock distribu-
tion reveals several smaller parallel faults as well as
conjugate NE-SW and ENE-WSW oriented faults.
A comparison of CMM and SIL locations revealed
a systematic westward shift of shallower events which
we attribute to variation in station distribution and
possible 3D variations in upper crustal structure.
The aftershock distribution along the two faults is
indicative of the main fault movement having been
in the centre of the fault, in agreement with the GPS
modeling of Hreinsdóttir et al. (2009) and Decriem et
al. (2010).
An increase in aftershock seismicity on May 31–
June 1, June 2–4, 6–7 and 8–9 is most likely caused
by short-term static stress buildup on adjacent faults.
Short-term viscoelastic response of the lower crust
may also be a contributing factor.
A third region of aftershocks marks a more com-
plex area of activity to the west of the main faults.
Some event distributions and fault plane solutions
align to suggest slip on smaller NS faults in this re-
gion. However, left lateral strike slip features cannot
be excluded in this zone, although probably with very
limited slip.
38 JÖKULL No. 60