Jökull - 01.12.1965, Qupperneq 19

GUDMUNDUR E. S I GVALDASON
UNIVERSITY RESEARCH INSTITUTE:
The Grímsvötn Thermal Area
Chemical Analysis of Jökulhlaup Water
The Skeidará jökulhlaups* have been describ-
ed thoroughly by Nielsen (1937), Rist (1955) and
Thorarinsson (1954). They are known to ori-
ginate in a volcanic caldera, Grímsvötn, in
the central part of Vatnajökull, Iceland’s largest
glacier. The jökulhlaups occurred about once
a decade, but since 1934 the period has been
5—6 years. Between jökulhlaups the Grímsvötn
caldera fills up with ice and water, but in the
jökulhlaup the water escapes. The jökulhlaups
were often followed by a volcanic eruption
but in the later jökulhlaups this has not oc-
curred. After the jökulhlaups, however, inten-
sive tliermal activity is observed on the caldera
floor, and along the caldera walls, but this is
soon submerged by the accumulating ice. This
source of heat apparently causes the melting of
the ice, which is then discharged as water
through subglacial channels during the jökul-
hlaups.
Associated with all jökulhlaups from the
Vatnajökull region is a strong smell of sulphur
dioxide and hydrogen sulphide, indicating the
close relationship of this phenomenon with geo-
thermal activity. Until the jökulhlaup of Sep-
tember 1965 very few chemical analysis of the
water hacl been made and the present oppor-
tunity was therefore especially welcome.
Mr. Sigurjón Rist took daily samples of the
waler over the periocl of peak discharge. I
would like to thank Mr. Rist for making the
samples available to the author. The samples
were collected into glass bottles, and some
time elapsed before they could be analysed.
Some of the analytical results are therefore not
entircly reliable, but in view of the overall
*) Jökulhlaup = glacier burst.
chemistry of these w’aters the error introduced
by this handling of the samples is not serious.
The chemical analysis are presented in Table
1. Considering the fact that this is glacier melt
water íhe high concentrations of several ele-
ments are rather unexpected. Duririg the period
which the samples cover, a distinct maximum
is found on Sept. 6th. According to Rist (per-
sonal communication) this is also the dav of
peak discharge. Assuming a uniform concentra-
tion of chemical elements in the subglacial
caldera lake, the levelling off to both sides of
the maximum is probablv due to mixing with
normal runoff water in the Skeidará drainage
area. The peak concentration will closely re-
semble the mean chemical composition of the
lakewater, since the peak discharge is very large
compared to normal runoff.
Thorarinsson (1953) calculated the annual
precipitation in the Grímsvötn caldera and
obtained a value of 2300 mm/year. With a total
intake area of 300 km2 this amounts to 0.7
km3/year. S. Rist (1955) estimated the total
water discharge during the jökulhlaup in 1954.
This jökulhlaup occurred after 6 years of accu-
mulation and the amount of water was 3.5
km3, averaging approximately 0.6 km3/year and
this is the same order of magnitude as the
precipitation. These figures, and less accurate
estimates of the total water discharge during
other jökulhlaups indicate a definite balance
between precipitation in the caldera intake
area, and the amount of water discharged in
the jökulhlaup. This suggests that thermal w'ater
has not entered the caldera from hot springs
with a recharge area lying outside the caldera
since the amount of thermal water necessary
to raise the chemical concentration of the
September 6th sample would be at least 1:4
JÖKULL 125