Jökull - 01.12.1989, Qupperneq 5

to the north caldera. The subglacial lake covers the
greater part of the main caldera and at high water
level it has in the past extended into the north cal-
dera. The area covered by the lake is the area where
geothermal activity is most intense.
Surveys aimed at mapping the subglacial topogra-
phy in the Grímsvötn area date back to 1951, with
the seismic work of the French-Icelandic expedition
(Eyþórsson, 1951, 1952; Joset and Holtzscherer,
1954). More seismic reflection work was done in
1955 by the Icelandic-French expedition (Þórarins-
son, 1965). However, these surveys failed to provide
satisfactory estimates of the elevation of the bottom
of the subglacial lake. Gravity surveys were carried
out in 1960 and 1961 to obtain more information on
the subglacial topography of the area (Pálmason,
1964; Sigurðsson, 1970), but interpretation of the
gravity data was difficult due to lack of control
points where ice thickness was known and Bouguer
anomaly could be determined.
Since 1977, the subglacial topography of large
parts of the Vatnajökull ice cap has been defined in
considerable detail by radio-echo soundings
(Bjömsson, 1986, 1988). One part of this work was
a detailed survey of the Grímsvötn area, carried out
at the same time as the seismic survey. However, as
radio waves of the frequency used in the soundings
(1 -5 MHz) do not penetrate water, other methods
were needed in order to obtain information on the
caldera floor beneath the lake.
The seismic reflection survey described in this
paper complemented the radio-echo soundings by
obtaining information on the subglacial lake and the
main caldera in Grímsvötn. The survey objectives
can be listed as follows:
1. To determine the elevation and topography of the
floor of the main caldera in Grímsvötn in as much
detail as possible.
2. To measure the area and the volume of the subgla-
cial lake and in that way obtain data for the
assessment of the flood potential of the area.
3. To obtain information on the structure of the cal-
dera, in particular whether it is composed of
several smaller calderas.
4. To obtain information on the material deposited
on the lakefloor by volcanic eruptions and sedi-
mentation, i.e. the caldera infill.
In this paper the fieldwork and processing of the
data are described and the results of the survey are
presented.
FIELD PROCEDURES
Three seismic reflection lines were surveyed,
using a method giving continuous single-fold cover-
age. Line 1 (4800 m) bears east-west and line 2 and
3 run north-south (2520 m and 3000 m). The ends of
the lines reach some distance on to the slopes that
border the ice shelf to north, east and southwest
(Fig. 2).
The source-receiver arrangement used was. in
essence a split spread. However, two shots were
fired in each shothole, one for each part of the
spread, as all the geophones were placed to one side
of the hole for the first shot and then moved to the
other side for the second shot. The length of the
receiving spread was 230 m, the 12 geophones being
placed at 20 m intervals with a minimum offset of
10 m. The second geophone spread for each
shothole was used as the first spread for the next
shothole on the line (Fig. 3).
A single vertical (P-wave) geophone with a
natural frequency of 8 Hz was used for each chan-
nel. During the course of the survey the weather
was sunny and hot, giving rise to wet and slushy
snow conditions on the surface of the glacier. Plac-
ing the geophones at about 30 cm depth in the snow
seemed to give satisfactory coupling.
Small dynamite charges (150-400 g) were used as
seismic sources. The charges were detonated at the
bottom of 30 m deep holes, drilled by a hot water
drill.
The seismic recording system used was a 12 chan-
nel Geometrics Nimbus 1210F. A 80 Hz highpass
analog filter on the Nimbus was used for all record-
ings apart from the first two shotpoints where a
50 Hz highpass was used. 'Tests showed that the
application of the filter produced sharper reflections
and improved signal to noise ratio. Each record was
1024 ms long with a sampling interval of 1 ms.
JÖKULL, No. 39, 1989 3