Jökull - 01.12.1983, Blaðsíða 17

hydrograph. It rose in three days and receded in
two weeks. At the same time an eruption occurred
tn the Grímsvötn area and one of three craters was
sighted in, as being locatedjust north ofGrímsvötn
(Thorarinsson 1974: fig. 21 p. 82). Melting of 2 km2
due to eruption of some 0.2 km3 of lava to the
glacier base could explain the peak. The rate of
melting stayed abnormally high until 1873, when
the largest historic eruption in Grímsvötn occurred
(Thorarinsson 1974). However, this estimate for the
years 1867 to 1873 might be false, as the minimum
in the first period after 1873 may suggest that the
water level fell deeper than usual in the jökulhlaup
of 1873.
No explanation will be suggested of the two
other peaks. The jökulhlaup in 1897 had a typical
hydrograph but it interrupted the regular ten-years
period between jökulhlaups as only 5 years had
passed since the jökulhlaup in 1892. The jökul-
JÖKULHLAUPS FROM GRÍMSVÖTN.
Fig. 2. a) Estimated volume of water in jökul-
hlaups. b) Computed rate of melting by the sub-
glacial heat source at Grímsvötn, (left), and the
heat ílux, (right).
2. mynd. a) Mat á rúmmáli vatns í Grímsvatnahlaupum.
b) Stöplarit sýnir mal á ísbráðnun vegna jarðhita milli
hlaupa og afl varmagjajans (MIV).
hlaup in 1948 was unusual, as it ran for two
months (Thorarinsson 1974).
VVe may conclude the discussion of Fig. 2 as
follows: Ice in Grímsvötn is melted firstly by ther-
mal fluid which brings heat up from a magma body,
second when the magma itself erupts up to the
glacier bed. If all peaks in Fig. 2b were caused by
injection of magma to the glacier base, the ice melt-
ed as a result would be 10% of the volume ofjökul-
hlaups during the last 120 years. Further, 70% were
melted by the geothermal fiuid, whereas 20% are
surface ablation. Magma which erupts straight
through the ice cover causes negligible melting.
INFLOW OF MAGMA
TO THE GRÍMSVÖTN AREA
The natural calorimeter at Grímsvötn can be
used to estimate the inflow of magma from which
the heat derives. We can assume that approxima-
tely 10% of the heat is released from cooling of lava
which has been injected to the glacier bed and 90%
from magma which solidifies in the upper crust. A
thermal power of 5000 MW at Grímsvötn is equi-
valent to the heat flux from 45 x 1 OTn’yU1 of
magma which solidifies and cools down to 400°C,
and 5 x 106m3yr"' of magma which cools down to
about 0°C. To this inflow we must add volcanics
which erupt through the ice. They cause negligible
melting and, therefore, are not estimateed by the
calorimeter. This addition is on the average 1.5 x
106m3yr~' according to Thorarinsson (1967), who
estimated the total volume 1.5 km3 of tephra from
Grímsvötn during the last 1100 years.
The inflow of magma to the Grímsvötn area can,
therefore, be estimated as 52 x lO^nTyr-1 on the
average during the last 120 years.If the peaks on
Fig. 2b reflect transport of magma up to the glacier
base this would mean 10% (5 x lÖ’m!yr_1) of the
magma, whereas 3% erupt as tephra through the
glacier. But the major part, 87% (45 x lO’m'yr-1
solidifies in the upper crust, where this mass sinks
down again and contributes to the extension of
Iceland. The lateral extension of Iceland is on aver-
age 2 cm yr“ *. The inflow of magma corresponds, on
the other hand, to a vertical wall 10 km long and 10
km high which expands 0.5 m yr-1. Transport of
sediments in jökulhlaups may at the most reduce
this volume by 10% (Björnsson 1982). The greatest
part must sink down again at the same time as
magma intrudes into almost horizontal sills. Walker
JÖKULL 33. ÁR 15