Jökull - 01.12.1990, Qupperneq 141

(Bjömsson, 1979), at least until recently (Sigurðsson,
1989).
The purely glacial component of the rivers, which
flow out from the Langjökull glacier, is not exactly
known, but hardly exceeds 100 m3/s. That would
account for approximately 3,000 mm/year of precip-
Hation on the glacier. A mean precipitation of 4,500
ttim/year on the glacier would leave only 50 m3/s for
the glacial groundwater. A change in the glacier bal-
ance could be included in the presumed value for the
precipitation. In the previous discussion some num-
bers have been mentioned for the probable glacial
groundwater flow in the various basins. If the rough
numbers are summarized (all in m3/s) one obtains the
followingdistribution: NtoSeyðisálessthan5,NWto
Hvítá í Borgarfirði less than 10, SW to Þingvallavatn
20-40, S through the mountains less than 10, SE to
Hvítá eystri perhaps 20, which yields a total of 50-80
m3/s.
The mean infiltration ratio on the glacier could
then be 30^-0 %. The permeability of the bedrock is
certainly not restrictive to the infiltration as is most
clearly seen in the dry surrounding areas. The per-
meability and the structure of the glacier itself are
probably the dominating, restrictive factors, permit-
tmg only some part of the groundwater in the glacier
to penetrate down in the underlying bedrock. Even so,
a mean infiltration of 1,500-2,500 mm/year seems
to take place, demonstrating clearly the importance of
the permeability of the bedrock.
HYDROGEOLOGY OF THE HOFSAFRÉTT
Region
The highland region of Hofsafrétt, at 600-800 m
a-s.l. N of the glacier Hofsjökull, is highly interest-
mg for the investigation of glacial groundwater runoff
(Fig. 19). It is essentially composed of stratified basalt-
lavas formations, ranging in age from Recent lavas at
the margin of the glacier to the Tertiary rocks in the
deep valleys, incising its northem front. The western
Part is cut by a fissure zone, running just west of N
and very likely connected with a caldera respectively
central volcano hidden below the glacier (Bjömsson,
1988). Hofsafrétt lies in aprecipitation shadow behind
Hofsjökull to the south and the plateau of Nýjabæjar-
fjall to NE.
A great part of the precipitation, falling as snow
during the winter, is swept off in the snow melt in
spring to early summer. The infiltration is therefore
probably very small and still reduced by the low per-
meability of the tillites and moraines, which cover
wide areas, delaying the infiltration and allowing a
strong evapotranspiration to take place in summer.
Springs issue on the northem slopes of the plateau,
partly in Late Quaternary but partly in Early Quater-
nary rocks. In the westem part the springs are closely
connected to fissures but in the eastem part they are re-
lated to stratified aquifers at relatively shallow depths.
No longtime serial records are available for the dis-
charge in the springs. It was measured 1977 (Einars-
son and Hannesdóttir, 1977), and sporadic measure-
ments and estimates have been made since then. The
total groundwater runoff can be assessed from the
records of the hydrological gauge at Goðdalir, which
point to a probable minimum of near to 6 m3/s. The
total flow in summer from the fissured westem part is
estimated to be 4-5 m3/s, but the area of the icefree
drainage basin is 400-500 km2. The winterflow seems
to be not much less, so far as observed. The springfed
runoff in summer from the non-fissured eastem part
is estimated to be probably 3-4 m3/s, but it seems to
dwindle to a trickle in winter.
Assuming a mean groundwater outflow of 2 m3/s
as the presently best guess for the eastem part, it corre-
sponds to an infiltrationrate of only 100 mm/year. The
permeability of the western part is probably better, so
that the infiltration could there be somewhat higher.
Even so, a surplus of 1.5-3 m3/s seems there to be
present, which then most likely has a glacial origin,
accounting for 1/3-2/3 of the total runoff.
The different hydrogeological nature of both parts
of the plateau is clearly reflected in the chemistry of
the groundwater as seen in some 30 samples from
the region collected 1984-1988. The chloride content
is everywhere low, 1.9-2.3 ppm in the southernmost
springs in the fissure zone, 2.4-3.8 ppm in the northem
part of the fissure zone, 1.5-2.0 ppm in the NE towards
Nýjabæjarfjall and 1.9-2.6 ppm in the rest of the east-
ern part (higher evapotranspiration?). The chemistry
in the eastem part shows some diversity, which in most
JÖKULL, No. 40, 1990 137