Jökull - 01.12.1974, Qupperneq 31

Fig. 2. Dissolved solids in Skeidará and Þjórsá
in 1964 and 1965.
Mynd 2. Uppleyst steinefni í Skeiðará og
Þjórsá 1964—1965.
ate water of geothermal origin. Fig. 2 shows the
amount of dissolved solids in Skeidará prior to,
during and after the jökulhlaup in 1965. As
indicated by this diagram, leakage occurs from
Grímsvötn both before and after the jökul-
hlaup in such great quantities that the total
amount of dissolved solids is considerably higher
than normally. The above clearly indicates that
the amount of water stored in Grímsvötn can
be considerably less than the precipitation on
the catchment area of Grímsvötn.
DISCHARGE MEASUREMENTS
Measuring discharge in rapidly changing al-
luvial channels as occur in jökulhlaups on
Skeidarársandur poses great difficulties. These
difficulties arise from the fact tliat an empirical
formula is not known l'or this kind of channels
as the Manning equation which is intended for
stable river beds is not strictly valid here.
To calculate the discharge in a river two
parameters have to be known, i. e. the cross
sectional area and the average velocity. Multi-
plication of these two factors gives the dis-
charge. Neither of these factors can be directly
measured in the case of Skeidarárhlaup.
The average velocity is found in two dif-
ferent ways. One by measuring the surface velo-
city and assuming a relationship between sur-
face velocity and average velocity. This rela-
tionship is assumed to follow that for stable
channels as other data are not available. The
other by calculating the average velocity from
the Manning formula. The gradient of the
river is measured and an estimate is made of
the average depth and the Manning coefficient.
The cross sectional area of the river is still
more difficult to estimate as it is a product of
two factors: the width which is relatively easily
measured and the mean depth or the hydraulic
radius which can only be estimated. One way
of estimating the mean depth is to use the
elevation of the river bed after the flood as a
base. But because of the rapid changes in bot-
tom topography during the flood this reference
is not at all necessarily applicable. Rist (1973)
approached the problem by assuming a base-
line, about 0.6 m lower, as representing the
river bed during peak flood, whereby he arriv-
ed at a value as much as 40—80% higher than
the value derived by after-flood topography.
S. Rist’s estimate of the accuracy of discharge
measurements as being ± 20% seems rather op-
timistic and probably tends to give too high
results. This is especially the case with the peak
flow. The above assumption that half the floocl
water flows in 3 days may therefore not be
valid.
Some of the emerging water does not have
its origin in Grímsvötn, but from melting of
the glacier by the flood water on its way onto
Skeidarársandur. This melting is produced by
the lieat capacity of the flood water, which
probably is around 4°C in Grímsvötn and also
by heat generated from friction in the ice
channel in the 1300 m drop from Grímsvötn
to Skeidarársandur. Together these factors can
yield additional water amounting to 8% of the
flow of the storage water.
The reason for this discussion of flood volume
is the necessity to use it as a base for calculat-
ing the sediment discharge. In view of the dis-
cussion above the discharge in this jökulhlaup
is estimated as 2 km3 and this figure will be
usecl for the calculation of sediment transport.
SEDIMENT TRANSPORT
The investigation of the sediment transport
of Skeidará was carried out in three stages.
JÖKULL 24. ÁR 29