Jökull - 01.01.2010, Blaðsíða 16

Geirsson et al.
matic intrusion (Jakobsdóttir et al., 2008; White et al.,
2009).
Two CGPS stations (SAUD and BRUJ) showed
clear signs of the intrusion (Figure 3). The lack of
vertical motion detected at the stations, relative to the
horizontal velocity change of about 30 mm/yr, was
interpreted as a dipping dike of a volume of 0.05
km3 was intruding into the crust (Jakobsdóttir et al.,
2008). This model was further supported by episodic
GPS measurements and modeling of InSAR observa-
tions (Hooper et al., 2009). The deformation ceased
in April 2008 following a culmination and a change
in character of the seismic activity (Geirsson et al.,
2009).
The episode has been interpreted as a magma in-
trusion into the lower parts of the crust (12–20 km
depth), and the earthquakes are thought to be caused
by brittle fracturing of the host rock where magma
is intruding. This episode provided the first real test
of the ability of using data from the CGPS network
to follow a magma intrusion. The CGPS network
proved very valuable for assessing the state of the in-
trusive episode along with seismic and other deforma-
tion data.
Load induced deformation
Glacier retreat
Warming climate is causing the glaciers in Iceland to
melt by thinning of the ice caps and retreat of the
edges (Björnsson and Pálsson, 2008). The melting re-
duces the load on the Earth’s crust, resulting in crustal
uplift around and under the glaciers. This uplift has
been captured by the CGPS network (Figure 4), and
importantly also by two GPS campaigns in 1993 and
2004 that covered the whole of Iceland (Árnadóttir et
al., 2009). The observations show that a broad area
in central Iceland is being uplifted by more than 1–2
cm/yr. This result, along with models of the history of
the ice load deduced from glaciological observations,
has been used to infer crustal structure and viscoelas-
tic properties of the crust (Árnadóttir et al., 2009).
Interestingly, the present uplift rates observed by the
CGPS network deviate somewhat from the 1993 to
2004 episodic campaigns, with higher present rates of
uplift, which may reflect increased melting rates of
the ice-caps since mid-1990. In addition to the high
uplift rates, annual cyclic variations in the CGPS site
elevation are also observed (Geirsson et al., 2006), co-
inciding with annual snow accumulation and melting
of the ice caps (Grapenthin et al., 2006). Grapenthin
et al. (2006) used these observations to infer the elas-
tic properties of the crust, since viscoelastic response
can be ignored for annual frequencies.
The Hálslón water reservoir
In September 2006, a new 25 km long and 2 km
wide lake, with a maximum depth of nearly 200 me-
ters, was formed north of the Vatnajökull ice-cap for
a hydro-electric power plant (Figure 3). The Icelandic
National Power Company had a network of seismic
stations and three CGPS stations installed by IMO in
the area to follow possible deformation and seismicity
caused by the formation of the lake. The CGPS defor-
mation monitoring was augmented by episodic GPS
measurements in the area (Ófeigsson, 2008). Three of
the episodic sites were upgraded to continuous sites,
coming into full operation in 2008. The CGPS sites
show a maximum of 15 mm observed subsidence due
to the initial filling of the lake. The observed deforma-
tion during the initial filling of the reservoir seems to
rather indicate outward movement (Ófeigsson, 2008).
The reservoir is subject to large annual variations in
lake level, around 50 m in a normal year, with a high-
stand in September. Seasonal variations are observed
in the CGPS time series, but they are subtle and prob-
ably partly counter-balanced by the opposite phase in
load changes at Vatnajökull.
Other transient deformation
Krísuvík uplift-subsidence episode
A rapid increase in seismic activity accompanied with
a significant surface uplift in the Krísuvík region (Fig-
ure 1) was detected in 2009. This is the first time such
activity has been documented on the Reykjanes Penin-
sula. Continuous GPS measurements started in Krísu-
vík in February 2007 when the station KRIV was in-
stalled. A velocity anomaly is apparent in the time
series from KRIV since late 2008 or early 2009 (Fig-
ure 8). The station has experienced both increased
southward motion and uplift followed by a period of
reversed motion. During the period of uplift, seismic
16 JÖKULL No. 60