Jökull - 01.06.2000, Page 51

Geirfinnur Jónsson and Leó Kristjánsson
NEW SURVEY – JOINT PROCESSING
In 1990 the authors began an aerial survey of the
magnetic field along the south coast of Iceland. As a
part of that survey, two lines and a loop were flown
across the Katla caldera (Jóhannesson et al., 1990)
beneath Mýrdalsjökull. Positions were determined
by a Loran-C navigation system with the estimated
position error not exceeding 3-400 m. Total-field mea-
surements were made at 5 or 6 s (around 300 m) in-
tervals. The flight altitude was higher than previously,
2000-2100 m. Initial processing of data was similar
to that of Sigurgeirsson (Kristjánsson and Jónsson,
1996).
Cross-over checks between Sigurgeirsson’s data
and the 1990 data showed a mismatch which may be
minimized by manually shifting some of his lines east
or west by 0.5-1.5 km. This is reasonable because
over the glacier there are few if any landmarks, and
constant speed of the aircraft was apparently assumed
in the plotting of Sigurgeirsson’s measurements, alt-
hough ascending and descending must have affected
the speed considerably. Figure 2 shows the field on
Sigurgeirsson’s and our own flight lines after correcti-
on for the mismatch. Our corrections extend beyond
the glacier area where they may not be appropriate.
It is obvious from Figure 2 that the field variati-
ons along flight lines are on a much shorter scale th-
an the line spacing. Before gridding the field, stand-
ard procedure would be to filter and decimate the
data along flight lines so that point spacing would be
comparable with line spacing. That, however, would
reduce the information in the data considerably. We
have, therefore, decided to grid the original data
which shows more detail, sometimes appearing as a
pearl-chain structure along the flight path. This has
to be kept in mind during visual inspection of the
gridded representation of the magnetic field. We used
the (Golden Software) Kriging method for gridding
the data to a 50 by 50 km grid with 1 km node spac-
ing. The field intensity is displayed in color in Figure
3, and as illuminated relief in Figure 4.
GEOLOGICAL SETTING AND OTHER
GEOPHYSICAL DATA
The region under study lies in the southernmost part
of the eastern volcanic zone of Iceland. The zone is
associated with a broad magnetic high of 20 km width
trending 45-50  east of north. This high, which is pre-
sumably due to basalt volcanics emplaced during the
Brunhes geomagnetic chron (  0.78 Ma), disappears
at the south coast. The region is characterized by frag-
mental rocks of the palagonite formation of subglacial
or subaqueous origin. Radiometric age determinati-
ons (K. Wiese, personal communication 1994) indica-
te that the adjacent Eyjafjallajökull volcanic center is
about 1 Ma old.
The presence of an approximately 14 km long and
11 km wide caldera, trending NW under central south
Mýrdalsjökull was indicated in the geological map of
South Iceland (Jóhannesson et al., 1990). Björnsson
et al. (1994, 2000) have mapped the caldera in detail
by radio-echo sounding. Its depth is 600 to 750 m,
deepest in the north, while shallower and more rug-
ged in the south. Björnsson et al. (1994) suggest that
the variable morphology reflects different levels of
volcanic activity. The highest topography occurs at the
caldera rim where peaks protrude the glacier surface
as nunataks.
Seismic undershooting on a NNW-SSE line
through Mýrdalsjökull (Guðmundsson et al., 1994)
revealed anomalously low velocity and an S-wave
shadow beneath the caldera. They suggested this was
caused by a shallow magma chamber in the crust,
about 5 km across and reaching down to about 1.5 km
below sea level. Earthquake activity in the Mýrdals-
jökull area (Einarsson and Brandsdóttir, 2000) orig-
inates in two main areas, one within the topographic
caldera, the other farther west, at the periphery of the
glacier. It is suggested that this activity is related to
two magma chambers within the volcano, the western
one being younger and less developed.
A positive Bouguer gravity anomaly of up to
40 mgal is found over Mýrdalsjökull. Guðmundsson
(1994a,b) suggests that the main anomaly is due to
a gabbroic intrusive complex which may not reach
above 2.5 km depth. A large gabbroic body should
be expected to generate a positive magnetic signature
50 JÖKULL No. 49