Jökull - 01.01.2005, Qupperneq 111

Seismic activity in Iceland during 2004
−21.5˚ −21˚ −20.5˚ −20˚ −19.
63.7˚
63.8˚
63.9˚
64˚
64.1˚
M 0 1 2 3 4
H
He Ho
Á
Figure 3. Earthquake epicenters in the SISZ. Á denotes Ásahverfi, He the Hestvatn fault, Ho the Holt fault and
H Hveragerði. – Jarðskjálftar í Suðurlandsbrotabeltinu.
Vatnajökull and the November 2004 eruption of
Grímsvötn
The Vatnajökull ice cap, located in south-eastern Ice-
land, overlies several active volcanoes and geothermal
areas. Seismicity beneath the ice cap has increased
since mid 2003 (Figure 5a). From August to October
occasional pulses of seismic tremor, lasting roughly
half an hour, were detected at station grf, located on
Grímsfjall nunatak at the southern caldera rim (Fig-
ure 6). These pulses were interpreted as signs of in-
creasing geothermal activity in the Grímsvötn caldera
(Vogfjörd et al., 2005b). Furthermore, GPS measure-
ments showed that Grímsfjall had already surpassed
the elevation level attained before the last eruption
in 1998 due to magma accumulation (Sturkell et al.,
2005), indicating that an eruption was likely to occur
soon.
Additionally, there were signs of an imminent
jökulhlaup from Grímsvötn because the elevation of
the caldera lake had risen to the highest level since
the November 1996 jökulhlaup (Finnur Pálsson, pers.
comm., 2004). Considering this in connection with
the increased seismicity, uplift and inferred geother-
mal activity, an eruption was considered very likely
at Grímsvötn. Seismic activity increased further after
mid-October and on 01 November three earthquakes
of magnitude ∼3 occurred between 05:00 and 07:00
UTC in the caldera. Constant microearthquake activ-
ity followed throughout the day, but the events were
all too small to locate until at 19:30 when a M lw
2.5 earthquake occurred. By 20:00, earthquakes were
occurring constantly. Furthermore, seismic tremor
increased significantly around 20:00, suggesting the
beginning of a volcanic eruption (Vogfjörd et al.,
2005b). Earthquakes from October to 01 November
were relocated using a double-difference method to
increase accuracy (Slunga et al., 1995) (Figure 6).
Their locations form a northerly striking cluster, about
1 km east of the eruption site. This result was also
JÖKULL No. 55 111