Jökull - 01.12.1991, Page 39

0
Pressure (bars)
10 20 30
Fig. 15. Conditions in an geothermal upflow zone near
the edge of the lake before (a) and after (b) a typical
jökulhlaup. The initial water depth is assumed to be
150 m and the vertical axes show the depth below the
lakefloor. Hita og þrýstiástand í uppstreymissvœði
jarðhitavökva við jaðar Grímsvatna fyrir (a) og eftir
(b) dœmigert Grímsvatnahlaup. Lóðréttu ásarnir sýna
dýpi mœltfrá vatnsbotninum.
boiling and increased buoyancy of the fluid (Fig. 15).
Altematively, vigorous upwelling of the hydrother-
mal fluid may be caused by magmatic intrusion into
the hydrothermal reservoir. This may have been the
case when the sinkhole, observed northeast of
Svíahnúkur Eystri on December 8, 1983 (Björnsson
and Kristmannsdóttir, 1984), was formed by the col-
lapse of an ice cauldron during a period of high seis-
mic activity in Grímsvötn after the eruption of 1983
(Bjömsson and Einarsson, 1990).
SEISMIC TREMORS
Björnsson and Einarsson (1990) suggest that a small
subglacial eruption occurred in Grímsvötn on August
21, 1984. Continuous tremors were observed on seis-
mographs for about one hour, and based on the rela-
tive amplitudes on different seismographs a source
could be located in the Vatnajökull area. After this
event seismic activity in the Grímsvötn volcano
ceased, having been high for some months. Aerial
photos from August 22 (Table Al) indicate that sur-
face melting was high but no signs of a recent erup-
tion were apparent. If a subglacial eruption occurred
on August 21 the volume of erupted material was
small compared to the 1934 and 1983 eruptions.
HIGH SEDIMENT LOAD
IN JÖKULHLA UPS
An unexpected peak in the sediment concentration
occurred near the end of the jökulhlaup in 1972
(Tómasson, 1974). No eruption was observed in
Grímsvötn (Björnsson and Hallgrímsson, 1976). Air
photos taken on August 10, 1972, 4 months after the
jökulhlaup show no signs of a recent eruption or
unusually high heat flow. Moreover, Einarsson and
Brandsdóttir (1984) consider an eruption at this time
very unlikely, as it would have been detected by seis-
mometers. A different explanation should therefore
be sought for the observed peak in the sediment load.
DARK LA YERSIN SNOWPITS
A thin brown layer was observed in snowpits at 1.5-2
m depth in Grímsvötn in the summer of 1954
JÖKULL, No. 41, 1991 37