Jökull - 01.01.2004, Page 58

Leó Kristjánsson
in the Tjörnes sequence are presented so that they can
be of use in future applications of paleomagnetism to
stratigraphic work in the peninsula.
Some previous paleomagnetic studies at Tjörnes
These studies have been reviewed by Eiríksson et al.
(1990) in their paper describing paleomagnetic results
from lavas and sediments in the Breiðavík group. The
earliest work on the magnetism of the sediments by
Dutch investigators in the 1950’s and early 1960’s,
when large specimens (“hand samples”) were col-
lected, was only reported in somewhat vague terms
and will not be discussed here. Detailed sampling of
lavas in Tjörnes for laboratory measurements of re-
manence directions was first made by Doell (1972).
Doell found normal polarity in the lowermost lava
above the Tjörnes beds and reverse polarity in the up-
per one of two pillow lavas in the Skeifá stream within
these beds. The other Skeifá flow was unstable.
The only published results known from the
Tjörnes sediment beds below the Höskuldsvík lavas
are contained in a brief account by Gladenkov and
Gurariy (1976, see Figure 3 of Eiríksson, 1981a and
Figure 3 of Buchardt and Símonarson, 2003). They
seem to have found mostly normal polarities below
palagonite tuffs in the topmost (Serripes) zone of the
sediments but mostly reverse polarities near the base
of that zone above the Skeifá lavas. Reverse sed-
iments were also found below lignites in the mid-
dle of the Mactra zone of the sediments and on both
sides of the boundary between the Mactra and Tapes
zones. No information was provided on experimen-
tal techniques, the number of samples collected, their
stability or within-site agreement. F. Strauch (writ-
ten comm., 1982) collected 295 oriented cores from
sediments in Tjörnes in 1975. Paleomagnetic direc-
tion measurements were made on these but apparently
never published.
Eiríksson et al. (1990) carried out an extensive pa-
leomagnetic study on the sediments and lavas of the
Breiðavík Group as well as from a deep borehole on
Flatey island 25 km west of Tjörnes. The sediment
samples were collected from mudrock, sandstone and
tuff. Their remanence vectors turned out to be reason-
ably easy to measure in the laboratory, even with rel-
atively insensitive equipment. The directions of these
vectors were also often quite stable, especially in fine-
grained units. At the time, the data were not inspected
for evidence of rotational remanence (RRM). This is
an axial remanence component which sometimes ap-
pears in rock samples during alternating field (AF)
demagnetization in tumblers, and increases systemati-
cally when the peak field of the treatment is increased
(Stephenson, 1980). Upon rechecking of the original
data, RRM is not seen to cause any significant error in
the direction measurement on most of the samples.
Eiríksson et al. (1990, Figure 4) found several
reversals in the Breiðavík Group. The sediment
sites within that group were of reverse or uncertain
magnetic polarity, except two normal-polarity ones
(FG 5 and RB 3/3A). Thermomagnetic measurements
as well as the magnitude of coercive force in the
Breiðavík sediments, demonstrated that the magnetic
mineral in the sediments is magnetite. It is considered
that the remanence is of thermo-chemical or thermo-
viscous origin and introduced during or soon after
deposition; this is supported by measurements on a
late Quaternary interbasaltic sediment in Reykjavík
(Kristjánsson, 2003, site MI S).
THE PRESENT STUDY
Following the relatively successful magnetic mea-
surements in the Breiðavík group, it was decided to
attempt sampling at five locations (Figures 1 and 2) in
the lower part of the Tjörnes sequence.
Sampling and sample preparation
Core samples were collected from one lava flow in
the Skeifá stream, and 30 sites in the sediments, with
a portable gasoline-powered drill after removal of
weathered material. The four samples (occasionally
five or six) taken were distributed over a few meters
laterally. Each site is assigned to units in Strauch’s
(1963) stratigraphic scheme where possible (Table
1, final column). The sediments are soft and frag-
ile, and their grain size is very coarse compared to
the Breiðavík sediments sampled by Eiríksson et al.
(1990). Cores are about 2.4 cm in diameter whereas
2.5 cm cores are normally obtained in low-porosity
basalts with the same bits. The orientation scratch is
58 JÖKULL No. 54