Jökull - 01.01.2004, Page 60

Leó Kristjánsson
11
12
13
14
Lava
15
16
17
18
19
20
21
22
23 Congl.
G
H
I
24
J
Lignite
25
Lavas
N?
?
N

NT*
N*
Doell
N
NT*
N*
N
N*
N*
N
N*
NT*
N*
R
?
R
T?
Doell
Mudstone
Serripes
Mactra
HM
KG
HK
HV
0 m
100 m
Figure 2. Stratigraphic scheme for the
Tjörnes sediments (after Buchardt and
Símonarson, 2003), showing the es-
timated range covered by the present
sampling locations. For explanation of
stratigraphic units (numbers and cap-
itals) see legend to Table 1. Paleo-
magnetic polarities (from Table 1 and
Doell, 1972: black is normal, white is
reverse, hatched is uncertain polarity)
are shown for the coastal sites HM, KG,
HK and HV. The inland sediment sites
in SA are left out as their correlation
to individual units of Strauch (1963)
is not clear. The polarities given by
Gladenkov and Gurariy (1976) are also
not included here. – Áætluð staðsetning
sýnasöfnunarinnar innan Tjörneslag-
anna samkvæmt skiptingu F. Strauchs á
lögunum í afmarkaðar einingar. Segul-
stefnur eru sýndar.
The irregular direction variations were reduced by re-
peating the 20 mT treatment and using the mean of the
results. About a half of the cores were demagnetized
also at 25 mT, and some beyond.
Mean directions for each site were calculated us-
ing the “kappa-maximizing” method of Watkins et al.
(1975): a computer program selects those direction
vectors (from the data available after 10, 15, 20 and
possibly 25 mT treatment) which yield the longest
vector sum. This will in many cases reduce within-
site scatter from that obtained at any one of the peak
fields. However, there is always a danger that the pro-
cedure will select a set of sample directions that are
clustered by chance. Values of 95% confidence angles
obtained are listed in Table 1; for comparison, repre-
sentative values for basalt lavas in Iceland are often of
the order of 6◦. The directions in the table have been
corrected for regional tectonic tilt which is estimated
to be 9◦ towards 310◦ East of true North.
Results
The mean intensity of natural remanence in the sed-
iments is 0.18 A/m, dropping to 0.08 A/m after 10
mT treatment (i.e. J100 in Table 1) and 0.04 A/m af-
ter 20 mT. Mean values of J100 for Icelandic lavas
are 3–4 A/m and the mean value for “good” samples
from the Breiðavík sediments (Eiríksson et al., 1990)
is 0.22 A/m. Initial volume susceptibilities, measured
in ten samples, were of the order of 0.01–0.015 SI
units, about half of those in basalts.
The poor within-site agreement for the lava flow,
which is a tholeiite having a crude pillow structure,
is disappointing: in case the discordant directions are
caused e.g. by regional alteration, it may be expected
to affect the sediments even more strongly. Strauch
(1963) and Doell (1972) consider two lava flows to
be present in Skeifá; our site is presumably the lower
flow I 36 which Doell (1972) also found to give unsta-
ble or scattered directions. Three of the samples from
60 JÖKULL No. 54