Jökull - 01.12.1983, Blaðsíða 75
Chemical Monitoring of Jökulhlaup Water in Skeiðará
and the geothermal system in Grímsvötn, Iceland.
SIGURÐUR STEINÞÓRSSON
Science Institute, University of Iceland, Dunhaga 3, 107 Reykjavík.
NÍELS ÓSKARSSON
Nordic Volcanological Institute, University of Iceland, 101 Reykjavík
ABSTRACT
The glacial river Skeidará was continuously monitored
for solute chemistry during the years 1971 and 1972 in
which period ajökulhlaup of 3.2 krrd volume took place in
the river. In addition, other glacial rivers in Iceland issuing
from subglacial volcanic or hydrothermal areas have been
monitoredfrom time to time. The article describes the solute
chemistry of the rivers Skeidará and Skaftá, offering an
interpretation of the hydrothermal systems beneath Vatna-
jökull responsible for the jökulhlaups. In conclusion some
points are made regarding the prediction of jökulhlaups by
the chemical monitoring of glacial rivers.
The dissolved load of the rivers Skeidará and Skaftá can
be traced to three sources, i.e. solutes contained in the
precipitation, those derived from the dissolution of the
suspended load, and those having a hydrothermal origin. The
analyses of the burst water can be corrected for the non-
hydrothermal components permitting evaluation of the com-
position of hydrothermal Jluids.
The analyses allow an estimalion of (1) the temperature
of the Grímsvötn geothermal system in theyears prior to the
jökulhlaups 1972 and 1982, (2) the dilution of the hydro-
thermal fluid by melt water and (3) water seeping into the
system from other sources, and (4) the power of the system.
The calculated temperature of the geothermal componenl
of the 1972 hlaup was 10TC, that of the 1982 hlaup
192PC. The thermal energy in Grímsvötn was 4160 MW
príor to 1972 and 5590 MWprior to 1982; this change could
be the result of a minor eruption in Grímsvötn in 1972.
It is shown that the Grímsvötn system is dominated by
hydrothermal water whereas that of the Skaftá source is
dominated by steam.
INTRODUCTION
Jökulhlaups, or catastrophic bursts of glacial
rivers, are fairly common phenomena in Iceland.
The mechanism for triggering the hlaups may entail
either a purely hydrostatic floating of the ice barrier
of an ice-dammed lake, or a volcanic eruption may
be involved. Likewise, the flood water itselfmay be
derived from three sources: precipitation of surf-
ace water percolating through the glacier ice, melt-
ed ice, and geothermal water from subglacial ther-
mal areas. Instances ofall these possibilities are well
known in Iceland and have been described by vari-
ous authors. In particular, Thorarinsson (1974, 1975)
has made a comprehensive compilation of the jökul-
hlaups from Grímsvötn in Vatnajökull and from
Katla in Mýrdalsjökull, and Bjömsson (1974, 1976,
1977) has elucidated the mechanism of accumul-
ation and flooding.
Three subglacial volcanoes have been active in
Iceland in this century: Thórdarhyma in the SW
slope of Vatnajökull erupted in 1903; Katla,
beneath Mýrdalsjökull, erupted in 1918 and possi-
bly gave a minor eruption in 1955 (Thorarinsson
1975), and Grímsvötn beneath Vatnajökull erupted
in 1922 and 1934 (Thorarinsson 1974). Furthermore,
Tómasson et al. (1974) have suggested that the Jökul-
hlaup in Skeidará in 1972 may have been associated
with a minor eruption in Grímsvötn. Skeidará, the
main river draining Grímsvötn, has flooded on aver-
age every 6 1/2 years since 1934 without there being
visible evidence ofvolcanic activity. In the Skeidará
floods water of all three types enumerated above is
present: there is a sizeable meteoric component in
the llood water, there is almost always a flood
associated with an eruption, but the majority of the
floods occur without an eruption (e.g. Thorarinsson
1974).
Particularly, the Katla-floods in South Iceland
are extremely catastrophic, bursting on suddenly
with full force with a minimum of pre-warning. In
previous centuries Katla has erupted fairly regul-
arly, twice every century (Thorarinsson 1975). The
prediction of an eruption and the associated jökul-
hlaup is of considerable security moment because of
JÖKULL 33. ÁR 73