Jökull - 01.12.1992, Qupperneq 40

V.G.P. LATITUDE
Fig. 7. Plot of virtual geomagnetic pole (V.G.P.) lati-
tude (degrees, positive north) versus aggregate strati-
graphic height in the composite profile described in the
text. Poles marked x correspond to a95 > 23.5°in Ta-
ble 1. —Breiddargráða segulpóls í samsetta sniðinu,
úr 1. töflu. Skörun sleppt.
were collected per lava during initial sampling. They
were oriented with respect to the sun or geographic
sightings. Four samples were taken from many lavas,
especially those that were porous or crumbly. The pro-
file TL and TM 1-23 were sampled in 1977, the upper
part of TM and TN 18-66 in 1984. The shorter pro-
files KL, KM and KM were sampled in 1985, and the
top and bottom of TN in 1986. No samples were taken
from TO and TP. The total number of units sampled
was 204. During the latter expeditions, additional
samples were collected from some flows which had
previously given unsatisfactory results.
All remanence measurements were made at the
University of Iceland with an Institut Dr. Förster four-
probe static fluxgate magnetometer. Measurements on
samples collected in 1977 were made before alternat-
ing field (AF) demagnetization, and after AF treatment
in 100 and 200 Oe peak fields, using a two-axis tum-
bler device. The samples collected in 1985 were also
demagnetized at 150 Oe, and those collected in 1984
and 1986 were demagnetized at 100,150,200 and 250
Oe fields. The secondary viscous remanence which is
generally present in the samples to some extent, has in
most cases been eliminated by 100 Oe treatment. At
this and higher fields each lava generally gives good to
excellent agreement between sample directions, and
the improvement in the results by the extended AF
treatment is not decisive. The least satisfactory results
came from the lower part of TM and from KN, where
the agreement between polarities measured in the field
and in the laboratory is also poor. The primary rema-
nence of lavas in these profiles may have been affected
by thermal activity originating in the Stakkfell central
volcano.
The directional results from each unit were com-
bined by Fisher statistics, and the mean direction se-
lected for further processing is the one having the
smallest 95% confidence limits, a95 (all samples de-
magnetized at the same field). Eight samples were
rejected, as well as six lavas which gave scattered or
unstable directions, their best a95 exceeding 60°. In
another 9 flows, a95 values were larger than 23.5°. In
the remaining 189 flows, the root-mean-square of a95
is just over 7°. In one of the rejected flows, and a few
others especially from profile TL, there were signs of
isothermal remanence due to lightning in the data be-
fore demagnetization. However, in most of these the
lightning effects appear to have been removed by 100
Oe AF treatment.
Paleomagnetic results. The paleomagnetic direc-
tions which are given in Table 1, have been corrected
for tectonic tilt. The estimated downdip direction for
all the profiles is 260°east of true north, and the tilts
are as follows: 28°for profile TL, 20°for KL and KM,
12°for KN, 16 to 12°(decreasing upwards) for TM
and 12-6°for TN. These are unusually large tilts, with
correspondingly large uncertainties which may reach
38 JÖKULL,No. 42, 1992