Jökull - 01.12.1957, Qupperneq 6

surface channels which also would have to cover
a large part of the region.
There are no observable signs of any such
system. It is therefore unlikely that the water
could by steady state conditions take up such
a large fraction of the available heat flow. Even
a ratio as high as i/io would appear question-
able.
There is another difficulty. In some of the
low-temperature areas in the region the hot
water is issued at temperatures over 80° C. At
normal and steady state conditions the circula-
tion depth would have to exceed 3 km. This
depth appears relatively large in view of the
small horizontal extension of the region.
According to Einarsson’s results the north-
western peninsula has not been subjected to
volcanism for a few million years at least. This
time would suffice for the complete cooling of
all local intrusives in the upper 5 km. The
thermal activity can therefore not be caused by
a local and shallow transient effect of recent
volcanism.
The presence of the many thermal springs
may thus be regarded as an indication for
generally abnormal subsurface temperature con-
ditions on the north-western peninsula.
Well-lemperatures. Almost all wells in Iceland
that exceed the depth of 50 metres have been
drilled in thermal areas for the purpose of the
production of hot water or natural steam. There
are, in fact, only three apparently non-thermal
locations where wells deeper than 90 metres
have been drilled and that appear to give some
information on the temperature-depth relation
as unaffected by significant local thermal ano-
malies or volcanism.
The location of these wells is shown in Fig. 1
and the data are given in Table I. The
temperature-depth relation is almost linear in
all three wells.
Well (1) is located in Þykkvibœr on the
south-western lowland. The upper 50 metres of
the well penetrated dense basaltic sand wliereas
the lower 40 metres penetrated regular basalt.
Well (2) is located at the farm Arnarholt on
the small peninsula Kjalarnes and at the mouth
of the glacial fiord Hvalfjördur. Well (3) is
located at the farm Tindar on Fellsströncl and
is also at the mouth of a small glacial valley.
Both the latter wells penetrated basalt.
Samples of drill cores for a systematic study
of the thermal conductivity were not available.
The conductivity of the rocks penetrated will
therefore have to be inferred from data on
similar rocks elsewhere.
The thermal conductivity of basalts is gener-
ally listed as 0.004 to 0.005 cal/sec °C cm (8).
The conductivity of 4 samples of almost un-
altered basalts from wells in south-western Ice-
land has been found to be 0.0039 to 0.0045 in
the same units. On the other hand, tuffs which
have a similar chemical composition as the
basalts have a lower conductivity, mainly the
thermally altered tuffs. Values as low as 0.0023
have been measured on a heavily altered sample
having a density of 2.00. These low values are,
however, most probably confined to zones of
lieavy thermal alteration. The writer is indebt-
ed to Dr. F. Birch, Dunbar Laboratory, Harvard
University, for having carried out these mea-
surements.
The present indications are that the average
vertical thermal conductivity of the basalt plat-
eau lies between 0.0035 and 0.0045 cal/sec°C
cm depending on the degree of alteration. A
figure of 0.004 will be adopted here, and has
been used in order to compute the heat flow
in Table I in the case of wells (2) and (3). The
average figure for well (1) has been estimated
at 0.0035 because of the sand horizon at the
surface.
TABLE I E.'itimated Estimated
No. Location Depth Ternperature average heat heat flow
metres gradient conductivity microcal/
°C/m cal/sec °C cm sec cm2
(i) Þykkvibær 90 0.093 0.0035 3.3
(2) Arnarholt 210 0.161 0.004 6.5
(3) Tindar (2 wells) 105 0.110 0.004 4.4
Mean ......................... 4.7
4