Jökull - 01.12.1983, Qupperneq 79

maritime precipitation, the chemistry of which is
controlled by the amount of ocean spray and the
elevation of condensation.
The relative contribution of the two sources can
be determined if the ion input from the surface rocks
and the averaged precipitation composition for the
drainage basin can be estimated. To account
qualitatively for normal Skeidará discharge an
abrasion experiment was performed on basalt of a
composition similar to that of the suspended load in
the river during its 1972 burst (Table 2). Prelimin-
ary results from that experiment indicate that the
low Si02 of normal Skeidará discharge (No. 1,
Table 3) is reached quite early in the process. The
hydrolized ions increase at almost constant ratios
during the abrasion until the point ofsaturation for
opaline silica is reached at approx. 100 ppm Si02; at
lower silica values the silica/alkali ratio of the water
can be regarded as constant.
The variable composition of precipitation from
diíTerent elevation is not well known but the gener-
ally accepted compositional trend shows the
oceanic ratios of Cl/Na, Na/Ca, and C1/S04 to
decrease drastically with elevation because HCl
remains preferentially to sulphate in the gas phase.
The composition of precipitation in the Skeidará
drainage basin is unknown as yet but supposedly it
ranges between typical coastal- and high- altitude
chemistry. The sodium concentration in the
abrasion experiment reached about 4 ppm'at 11.2
ppm SiO0. Normal Skeidará water, containing 11.1
ppm Na, derives accordingly about 40% of its
sodium from abrasion. The very low chloride
content of the river (3 ppm) indicates that high-
altitude precipitation dominates the drainage basin
which also is evident from the relatively high
sulphate. According to the experiment 47% of the
sulphate in Skeidará is dissolved from the rocks,
which leaves the computed precipitation contrib-
ution with lowCl/Na and C1/S04 ratios. The con-
clusion is that the normal Skeidará water derives
about half its sodium and sulphate from the precipit-
ation, which is mostly ice of high-altitude origin. It
is, however, to be noted that thc calculated average
sodium of the precipitation feeding the Skeidará
drainage basin is four times higher than that mea-
sured for the Bárdarbunga ice core (Sci. Inst.,
unpubl. results).
The purpose of the foregoing discussion is to
justify the subtraction of solute concentrations in
normal Skeidará water to reconstruct the chemistry
of the Grímsvötn reservoir from the analyzed jökul-
hlaup water. It is suggested here that even if the
jökulhlaups from Grímsvötn are immense in their
proportions the dissolved load of the water due to
abrasion of the sediment load remains approxi-
mately the same, and that the observed increase
represents an innate characteristic of the ílood
water. Therefore the chemical analyses of the hlaup
water can be corrected for the ”normal“ component
by simple subtraction.
SOLUTE CHEMISTRY OF THE
JÖKULHLAUPS 1972 AND 1982
A detailed sampling program of Skeidará was
undertaken prior to and during the 1972 burst. The
first signs of chemical change appeared as a slight
but significant increase in chloride and pH value in
February 1972, but a more rapid increase in sodium
and carbonate on March 2 marks the beginning of
the jökulhlaup proper a day after the initiation of
hydrogen sulfide odours in the SE lowlands (Thor-
arinsson 1974). These chemical indications were
apparent a week before detectable increase in the
river discharge. The water rapidly assumed a
sodium rich (ca. 200 ppm) basic (pH 10) compos-
ition with high carbonateand sulphate (No.2, Table
3). This water type made up only a few percent of
the burst and was succeeded after March 18 by the
burst water proper (No 3, Table 3) the composition
of which remained remarkably stable till the end of
March, constituting about 95% of the flood volume.
Following the course of events to the end the last
signs of water in Skeidará disappeared in the last
week of April. The main mass of the burst undoubt-
edly has a strong geothermal component and in the
following we attempt to define its chemical source
and to delineate the processes at work producing
the two distinct water types of the burst.
Fig. 2 shows the 1972-jökulhlaup in terms ofdis-
charge per time unit (Rist 1973), with the three
water bodies under discussion indicated. The slow
initial mixing of sodic carbonate water with the
normal Skeidará (Fig. 3) is interesting in two re-
spects: firstly, the Iow Mg- and Ca-contents of the
discharge indicate that the dissolved load due to
abrasion during a 600-fold increase in discharge
remains unchanged — otherwise the low alkaline
earth composition innate to the sodic water would
have been swamped by products of abrasion
(compare Nos. 1 and 2, Table 3). Secondly, there
are striking chemical indicadons in this water,
JÖKULL 33. ÁR 77