Jökull - 01.01.2001, Qupperneq 39

Paleomagnetic studies in Skarðsheiði, SW-Iceland
were made on one specimen from each sample, with
an Institut Dr. Förster four–probe static fluxgate mag-
netometer. The remanence was measured before and
after alternating field (AF) demagnetization at 10, 15
and 20 mT peak field in a two–axis tumbler.
Extended demagnetization curves for five typi-
cal samples are shown in Figure 4. The directional
change between successive steps is shown where it
exceeds 1.5 . The 10 mT treatment is generally suf-
ficient to remove all or almost all of the secondary
viscous remanence present; the measured remanence
direction in a specimen rarely changes more than 3
degrees of arc between the 15 and 20 mT steps. Good
to excellent within–unit directional agreement is ob-
tained, as indicated by the 95% confidence radius

 being larger than 9 in only four units. Me-
dian destructive fields for the natural remanence are
very variable, but they exceed 20 mT in about 60% of
the samples from our sampling profiles (including the
profiles SS and SW, see below).
Treatment was continued to 25 mT or higher fields
for 32 flows from our profiles, in most cases due to mi-
nor directional instability of the primary remanence
in some of the samples but occasionally due to the
presence of lightning–induced remanence. In some
samples carrying a weak and soft primary remanence
component, anhysteretic and/or rotational remanence
may be picked up during AF demagnetization. These
samples also have a strong tendency to pick up new
viscous remanence immediately after the treatment.
This causes fluctuations in the measured remanence
directions at strong demagnetization fields (see e.g.
sample SI 33–3 in Figure 4) which we have attempted
to reduce by performing the treatment twice and aver-
aging the results.
Units where between–sample directional agree-
ment was unsatisfactory, were resampled at a site sev-
eral m to tens of m from the original sampling site. Al-
together about 24 samples were not included in statis-
tical calculations. Table 1 lists the number of samples
used from each lava flow, the best paleomagnetic di-
rection for that flow, and its corresponding virtual ge-
omagnetic pole (VGP). The directions have been cor-
rected for tectonic dip (see Appendix). The table also
gives the 95% confidence radius for the mean direc-
tion of each flow, its mean remanence intensity after
10 mT AF treatment, and its polarity.
Figure 4. Remanence intensity after AF demagnetiza-
tion in five samples from Skarðsheiði, including one
sample from a transitional flow (SS 17) at the R5–N5
boundary. An increase in intensity at the first step in
the reverse flow SI 33 (and also slightly in SJ 34) is
due to removal of a viscous overprint of Brunhes age.
Directional changes between steps are shown if they
exceed 1.5 . – Niðurstöður riðstraums–afseglunar á
fimm sýnum úr Skarðsheiði. Aukning segulmögnunar
í tveim sýnum verður þegar nýleg seigjusegulmögnun
með „réttri“ stefnu er fjarlægð á undan upprunalegri
segulmögnun með „öfugri“ stefnu.
Paleomagnetism of profiles SI and SJ – results
In the lower part of profile SI, hydrothermal alter-
ation may have affected the remanence of lavas to a
greater extent than in most lava profiles hitherto sam-
pled in Iceland. The average remanence intensity (af-
JÖKULL No. 50 39