Jökull - 01.01.2010, Page 15

Fifteen years of CGPS in Iceland
!"#$
!"$$
!#$
$
#$
"$$
"#$
!
!
%$$# %$$& %$$' %$$( %$$) %$"$
"#$%
&'
!"#$
!"$$
!#$
$
#$
"$$
"#$
!
!
()%*+
!"#$
!"$$
!#$
$
#$
"$$
"#$
!
!
,$-*
Figure 7
Figure 7. Time series of the motion of site GFUM at
Grímsvötn relative to fixed Eurasia. A constant rate of
22 mm/yr has been subtracted from the vertical com-
ponent to account for glacial rebound. The time se-
ries are produced using Bernese V 5.0, as described
by Decriem et al. (2010). The vertical line notes the
onset of the 2004 eruption. The dashed horizontal line
at -100 in the north component is set to the position
of the station just before the 2004 eruption. A cor-
responding line for the east and vertical components
coincides with zero. – Tímaröð hreyfinga stöðvarinn-
ar á Grímsfjalli (GFUM) miðað við fastan Evrasíu-
fleka. Jafn hraði, 22 mm á ári, hefur verið dreginn frá
lóðréttum þætti mælinganna til að reyna að fjarlægja
landris vegna rýrnunar jökla. Lóðrétt lína sýnir upp-
haf eldgossins 2004. Lárétt punktalína í norðurþætti
táknar stöðu stöðvarinnar rétt fyrir eldgosið 2004, í
öðrum þáttum fellur samsvarandi lína saman við núll.
A few days before the 1 November 2004 Gríms-
vötn eruption, the GPS receiver unfortunately stopped
logging data due to a power failure. Power was re-
stored on 20 November 2004 and the resulting de-
formation showed a co-eruptive movement of about
0.2 m inwards to the caldera, and 0.2 m of subsi-
dence (Figure 7). There is uncertainty whether the
periods shortly before and after the eruption are af-
fected by ice accumulation on the antenna radome
since the monument was not heated at the time. Af-
ter the eruption, the magma chamber started inflating
again, rapidly for the first few months after the erup-
tion and then slowing to a constant rate. This behavior
is comparable to what was observed for the 1998 erup-
tion and subsequent inflation (Sturkell et al., 2003b,
2006).
Assuming that the station GFUM is located on the
Eurasian plate, we can estimate when the station dis-
placements surpasses its previous pre-eruptive stage,
as an indicator of how imminent the next eruption is.
Depending on which component is used (east, north
or up), the station reaches the same position as before
the 2004 eruption again anywhere between late 2007
and 2011 (Figure 7). If the station is within the plate
boundary, as suggested by Sturkell et al. (2003b), the
east component has surpassed its pre-2004 eruption
levels even earlier, but the north component is less bi-
ased by plate motion. In order to account for crustal
rebound due to the melting of the glaciers, an ad-
hoc rate of 22 mm/yr was subtracted from the vertical
component. Judging from the discrepancy between
the components of the time series for GFUM, it can
be inferred that the deformation sources for the 2004
eruption and subsequent inflation are not the same.
Upptyppingar-Álftadalsdyngja intrusion between
2007 and 2008
During 2007 to 2008, an intense earthquake swarm
associated with surface deformation occurred north of
Vatnajökull ice-cap, east of the plate boundary (Figure
1). A network of CGPS and seismic stations was in-
stalled not far from the activity in 2004–2005 to mon-
itor possible deformation and seismic activity at a wa-
ter reservoir being formed for a hydro-electric power
plant. The seismicity was unusually deep, 11–22 km,
and showed patterns that could be related to a mag-
JÖKULL No. 60 15