Jökull - 01.12.1983, Qupperneq 78

TABLE 1. Volumes of jökulhlaups in Skeidará since 1954.
TAFLA 1. Rúmmál Skeibarárhlaupa síðan 1954.
Year Volume km3 Months of accumulation Apparent accumulation per month (km3) Relative rate of accumulation Reference
1954 3,5 77 0,0454 100,0 Rist, 1954
1960 3,0 66 0,04545 100,1 Thorarinsson, 1974, p. 195
1965 3,0 67 0,04478 98,6 Thorarinsson, 1974, p. 200
1972 3,2 78 0,0410 90,3 Rist, 1973
1976 2,4 53 0,05428 119,6 Rist, 1976
1982 1,3 65 0,0200 — S. Rist, pers.comm. 1982
in the caldera and the ice remains afloat. The geo-
thermal system below will be saturated with water
and the geothermal water ascending to the surface
is replaced by cold water from above; it may also be
assumed that some water percolating into the
ground is lost from the system altogether.
Chemical geothermometry assumes that the
composition of the cold water percolating into a
geothermal system is irrelevant since temperature
alone in the system governs the composition of the
hot water through mineral equilibria. Therefore, the
independent variables controlling the composition
of the flood water are: temperature of the thermal
water, governing the solute chemistry and the de-
gree of dilution by melting, and climate, which affects
the proportion of surface melt water in the caldera.
The temperature of the thermal water can be afTect-
ed mostly in two ways, either by variable water
circulation or by variation in the heat source. The
pressure effects due to 100 m oscillations in the
caldera-lake suface would only be important initi-
ally after a jökulhlaup, resulting in some 25°C vari-
ation of the boiling point ofwater at 500 m depth.
In a steady-state system the composition of the
caldera lake remains the same throughout the
period of accumulation, its only variables being the
temperature in the geothermal system and the
proportion ofsuface melt water. Table 1 shows the
apparent rate of water - accumulation in the Gríms-
vötn system since 1948, as reflected by the volumes
of jökulhlaups, assuming that the lake is drained to
the same level every time. This has been the case for
all the hlaups except the 1982-one in which the level
of the ice surface only subsided by about half the
customary distance (II. Bjömsson 1982, pers.
comm.) The accuracy ofthe flood-volume estimates
is about 20% (S. Rist, pers. comm. to S. Thorarinsson,
1974). Evidently the rate of accumulation shows a
consistent decrease from 1948 to 1972, to pick up
again between 1972 and 1976. This would be
consistent with the notion of a minor eruption in
association with the \9l2-hlaup.
DISSOLVED LOAD IN RIVER SKEIDARÁ
The uniform chemical composition of glacial
rivers in Iceland (Armannsson et al. 1973; Rist 1974)
reflects the monotonous basaltic rock composition
and cold climate ofthe country. Chemical weather-
ing is virtually absent and chemical modification of
surface waters due to interaction with vegetated soil
is limited at the banks of the glacial braided rivers
which typically spread over the sandur outwash
plains below the glaciers. The two major sources
contributing the dissolved load of the glacial rivers
are thus (a) the basaltic surface rocks and (b) the
Table 2. Chemical analyses (wt. %) of suspended load from the jökulhlaup peak of Skeidará 1972 (a) and
from Grimsvötn 1934 tephra (b). (Analyst: N. Óskarsson).
Tafla 2. Efnagreiningar (pungaprósent) á aurburði Skeiðarár í hámarki hlaupsins 1972 (a) bornar saman við
samsetningu gjósku úr Grímsvatnagosinu 1934 (b).
Si02 ai2o3 tío2 Fe203 FeO MnO MgO CaO Na20 k2o p2o5 h2o
(a) 50.10 14.25 2.28 3.72 8.61 0.22 5.12 10.86 2.30 0.36 0.23 1.23
(b) 50.80 13.92 3.05 3.64 10.76 0.22 4.24 9.76 2.75 0.50 0.32 0.12
76 JÖKULL 33. ÁR