Jökull - 01.06.2000, Qupperneq 5

Ground deformation at Katla
small surface displacement over a large area while
an expanding shallow magma chamber will cause
greater surface displacement in a small area near the
chamber. In the case of Katla, which is covered by
a glacier, measurements on solid foundations are not
feasible within about 10 km distance from the vent
that erupted in 1918. The available techniques at the
time of initial deformation measurements near Katla
in fact excluded measurements other than precision
levelling. Electronic distance measurements were at
that time in a developing stage and their capacity
poorly known; besides, the required instrumentation
was expensive and difficult to apply in areas of no
roads.
For precision levelling to be an effective met-
hod in observing ground deformation, the locations
for permanent markers have to be selected very car-
efully. Elevation differences between markers must be
small and stability of the ground where markers are
placed must be very good. For each optical levelling
tilt station, a relatively flat area, several hundreds of
meters across, is needed, where competent bedrock is
exposed. The station should be easily accessible by a
motor vehicle. A search for acceptable locations for
such stations was made in the summer of 1967, and
three locations were selected where optical levelling
tilt stations were constructed (Figure 2).
One of these locations is in Höfðabrekkuheiði
about 14 km ESE of Katla, near the old road which
was abandoned after the June 25, 1955 jökulhlaup
destroyed the old bridge across the river Múlakvísl.
The station consisted of 10 markers cemented into tuff
(móberg).
Another location in Kötlukriki is about 11 km due
east of Katla, about 2 km south of the mountain Sand-
fell. The tilt station consists of 12 permanent markers
cemented in pahoehoe lava which have come from
beneath the glacier Mýrdalsjökull, probably from the
volcano Katla.
The third location is near the river Jökulkvísl,
about 13 km ENE of Katla. Thirteen permanent mar-
kers were cemented into pahoehoe lava which may
have come from the Eldgjá eruption fissure.
All these tilt stations are located more than 10 km
away from Katla, or rather the eruption site of the
1918 Katla eruption. Because of this distance, only
very small tilt is to be expected if inflation is caused by
magma movement toward a shallow reservoir below
the 1918 eruption site. However, the location of a
hypothetical magma chamber beneth Katla was not
known when these stations were constructed, and
also the magnitude of possible ground deformation to
precede a Katla eruption was not known.
OBSERVATIONS
The optical levelling tilt stations which were
constructed near Katla in 1967 consist of 10 to 13 per-
manent markers each. These markers form arrays of
irregular shape, making it possible to observe ground
tilt of any azimuth. The distance between adjacent
markers is usually about 50 m or less. Measurements
were made with utmost care, using best available
optical levels with plane plate optical micrometers and
invar measuring rods and observing procedure which
minimised both accidental and systematic errors.
From 1967 to 1973, observations were made once
each year. No observations were made for 13 years
thereafter, but they were resumed in 1986.
It soon became clear that the station on Höfða-
brekkuheiði was unreliable. The tuff eroded rapidly
and one marker became loose before observation of
1986. One or more other markers showed irregular
movements, probably because of frost action.
The other two stations appear quite reliable and
observed vertical displacement of the permanent
bench marks are small and agree quite well with
general, although gentle, tilting of the ground.
One additional tilt station was constructed in 1992
on Fimmvörðuháls, to the west of Mýrdalsjökull,
about 20 km west of the 1918 eruption site. This
station was tilted significantly down in an easterly
direction between observations of September 1993
and September 1994. No spectacular tilt was obser-
ved at that time at the tilt stations east of Katla,
but those stations are closer to the Katla volcano
than the Fimmvörðuháls station, and should be more
sensitive to processes related to Katla. Therefore, this
observed tilt at Fimmvörðuháls is not considered as
related to the Katla volcano. An earthquake swarm
of June 1994, located about 7 km north-east of the
JÖKULL No. 48 3