Jökull - 01.01.2004, Side 62

Leó Kristjánsson
1)Information on the position of samples in Strauch’s (1963) stratigraphic scheme for Tjörnes was provided by J. Eiríksson
to A.I. Guðmundsson who recorded the sampling positions in L.K.’s field notebook in 1991. Supplementary information
on the Skeifá sites SA was provided by J. Eiríksson to L.K. in 2004. In Strauch’s notation which is based on previous map-
ping by G.G. Bárðarson (1925), the main distinct marine sediment units are designated by numbers, from 1 (lowest) to 25
(topmost, below the Höskuldsvík lavas). The present study covers the interval from unit 11 to 25. Terrestrial sediment units
are similarly assigned capital letters from A to J by Strauch, and sub-units are indicated by numbers increasing upwards.
Strauch (1963) uses the expression “Horisont” for both the units and the sub-units. Strauch’s scheme is mostly derived from
sections exposed at the coast; he does not indicate in detail how individual layers outcropping inland below the Skeifá lavas
fit into it and the situation there is also complicated by faulting (Strauch, 1963, p. 53).
SA 1 give reverse directions similar to Doell’s average
for his upper site I 37.
Those sediment directions that could be measured
with reasonable within-site agreement tend to be of
normal polarity. However, the three reverse sites di-
rectly underlying the Skeifá lava seem to be quite re-
liable and the lowest of these is at least 3 m below
the lava contact so that the reverse polarity is hardly
due to heating by the flow. The reverse polarity of site
SA 9 is much less certain, as one sample from there
(excluded from Table 1) was normally magnetized.
SUMMARY AND CONCLUSIONS
Remanence measurements from 30 sediment sites and
one lava within the Tjörnes beds (Figure 2, Table
1) indicate mostly normal polarities. The quality of
the Tjörnes beds for paleomagnetic research is clearly
much inferior to the sediments of the Breiðavík group
and Flatey. This is likely to be directly or indirectly re-
lated to the differences in grain size between the two
parts of the Tjörnes sequence. However, at 5 “normal”
and 5 “reverse” sites the 95% confidence angles (α95)
are less than 20◦, and all of these have virtual pole lati-
tudes of 60◦ and higher. They are considered to give a
fairly definite indication of the geomagnetic polarity
during the acquisition of the stable remanence com-
ponent. The sites with α95-values exceeding 20◦ are
not very trustworthy, especially those yielding mid-
latitude poles. Reverse polarities found in this study
below the Skeifá lava site and by Gladenkov and Gu-
rariy (1976) in overlying sediments, tend to support
the conclusion of Doell (1972) that the lava(s) were
erupted during a reverse geomagnetic (sub)chron. Ev-
idence for reverse polarities was not found in the
coastal sites HM which are supposed to span the same
units (Strauch s 11 to 14) as SA; a number of expla-
nations for this disagreement may be suggested. The
sites at KG and most of HK are of normal polarity,
while the top sampled unit of HK and the one mea-
sureable unit in HV are reverse. HV is overlain by the
normally magnetized lava flow I 38 of Doell (1972).
Accordingly, there is evidence for at least four ge-
omagnetic reversals having taken place during the de-
position of that part of the Tjörnes beds which was
sampled: two in the Skeifá profile (not seen in pro-
file HM), one between between Skeifá and Hallb-
jarnarstaðaá and one south of Höskuldsvík. The po-
larities in the sediments do not say much about their
age, but they are not in conflict with the sugges-
tion of Buchardt and Símonarson (2003) that the beds
between the Skeifá and Höskuldsvík lavas were de-
posited during the Gauss geomagnetic chron (3.58–
2.58 Ma ago). Progress in understanding the rema-
nent magnetization of the Tjörnes beds and its use for
stratigraphic purposes will clearly require extensive
additional sampling as well as major improvements
in sampling methods and laboratory techniques.
Acknowledgments
The study was suggested by Jón Eiríksson who se-
lected the sampling sites and provided information
on their stratigraphic position. Andrés I. Guðmunds-
son assisted in the collection of core samples. Har-
aldur Auðunsson carried out some of the paleomag-
netic measurements. The diagrams were made by
Geirfinnur Jónsson. The field work was supported by
the Science Foundation of Iceland (Vísindasjóður).
62 JÖKULL No. 54