Jökull - 01.12.2007, Page 54

Þorbjarnardóttir et al.
Figure 7. Mainshock-aftershock sequence southwest of Kistufell in northwest Vatnajökull. All single-event
earthquakes (green) of the sequence, from 24 September to 2 October, are shown. Relocated events (orange)
with relative error less than 500 m in latitude and longitude and 700 m in depth are shown. Most of these
occurred during the first three days of the sequence. The yellow star shows the location of the magnitude 3.7
mainshock. – Skjálftaröð suðvestan við Kistufell. Allir venjulega staðsettir skjálftar (grænir hringir), frá 24.
september til 2. október, eru sýndir. Afstætt staðsettir skjálftar (appelsínugulir hringir) með skekkju í lengd og
breidd minni en 500 m og 700 m í dýpi eru sýndir. Gula stjarnan sýnir staðsetningu meginskjálftans, sem var
3.7 að stærð.
at 10–13 km depth along a 1.1 km near-vertical plane
striking slightly to the northeast. Focal mechanisms
suggest predominant left-lateral movement along the
fault plane with a slight normal component.
Earthquake activity in Bárðarbunga is mostly con-
centrated beneath the northeast flank of the volcano
(Figure 1), a few kilometers south of the Kistufell
earthquake sequence. In 2006 the largest earthquake
beneath Bárðarbungawas of magnitude 3.5, occurring
in April. Seismicity was fairly continuous throughout
the year, but earthquake sequences were absent.
In April and September 2006, microearthquakes,
ranging in magnitude from 0.2 to 2, were observed in
connection with jökulhlaups from the Skaftá ice caul-
drons (Figure 1). On 21 April, flooding was detected
at the Sveinstindur gauging station, operated by the
National Energy Authority (see http://www.vatn.is).
Between 22 and 29 April at least 12 earthquakes took
place within an 8 km radius of the eastern cauldron.
The September jökulhlaup, from the western caul-
dron, reached Sveinstindur on 27 September. At least
seven earthquakes were detected near to the source of
the flood on 23 September. Daily GPS measurements,
which started in the summer of 2006, show that the
surface of the western cauldron began to subside on 24
September (Tómas Jóhannesson, pers. comm., 2007);
however, it is likely that a leak was established on 23
September or earlier. We interpret the shallow seis-
micity on 23 September as icequake activity caused
by increased subglacial water pressure and enhanced
sliding of Skaftárjökull in response to the onset of the
jökulhlaup.
52 JÖKULL No. 57