Jökull - 01.01.2005, Qupperneq 27

Stratigraphy and paleomagnetism of Suðurdalur, Eastern Iceland
a reverse-to normal transition, and the top lava is con-
sidered to be the same as the olivine tholeiite com-
pound flow ST 1.
ST – 8 sites at the base of Guðmundsson’s (1978)
profile 43 on the northern side of the gully of Sturluá
were sampled, plus two sites in flow units ST 1A,B
above ST 1. Site ST 8 which is probably in the same
compound flow as ST 7, underlies a 25 m thick sedi-
ment and then a sequence of mostly tholeiitic lavas of
unknown polarity.
MK – The profile follows the stream Marklækur,
Guðmundsson’s (1978) profile 36 (Figure 3b). The
numbering of the lavas was carried out by Þórdís
Högnadóttir (pers. comm., 1993). The sampling
reaches up to the Múli plateau where additional ex-
posures are found, but they are widely spaced and
possibly incomplete or affected by faulting. A total
of 38 sites were sampled, one of which (MK 4) was
later found to be a dike running subparallel to the hill-
side; we assume that it obscures a normally magne-
tized lava. In addition, two successive olivine tholei-
ite flows or compound-flow units named MM 1 and
2 were sampled near the dirt road through the valley,
1 km north of MK. Due to the regional tilt, these un-
derlie the base of MK and presumably belong to the
same olivine tholeiite series as ST 1.
Lying stratigraphically above these profiles is the
fifth, KO of Figure 5, where magnetic measurements
were only made in the field. However, field and labo-
ratory polarities in profile MK agreed very well so we
expect the polarities in KO (where at least three rever-
sals occur) to be mostly correct. The lowermost part
of this profile is eight lavas outcropping in the Kelduá
river. To correlate between MK and KO we assume
that a series of tholeiite lavas (erupted somewhere to
the south) begins to interfinger with the porphyritic
flows of the ML suite around MK 48 and KO 2–7.
Sampling and measurements
At least five 25-mm drill-core samples were collected
at each site in profiles VV, VS and ST, four in MK.
They were oriented in azimuth by sighting on the Sun
or distant geographical objects. The remanence di-
rection and intensity of one 21–22 mm long speci-
men from each sample was measured at the Univer-
sity of Iceland, using an Institut Dr. Förster static four-
probe fluxgate magnetometer. In order to remove sec-
ondary magnetization components such as viscous re-
manence, all specimens were demagnetized by alter-
nating field (AF) treatment in a Molspin 2-axis tum-
bler. The peak demagnetizing fields were 10, 15, and
20 mT, and sometimes additionally 25 mT if the re-
manence direction had not reached a stable end-point
by 20 mT.
Main results
The lava flows were generally good material for the
remanence measurements. In profile MK there were
minimal problems with viscous remanence and other
secondary components, and field polarity measure-
ments agreed well with laboratory ones. East of Suð-
urdalur there was much less agreement, because con-
siderable viscous remanence was present. Eight sam-
ples from five flows in the latter profiles were rejected
due to poor within-unit consistency of directions.
Table 2 gives the “best” average directions from
each lava (i.e. where the scatter is lowest); in most
cases this is after the 20 mT treatment. The table also
shows virtual geomagnetic poles (VGPs), 95% confi-
dence circles for the directions, and intensities. The
remanence directions have been corrected for tectonic
tilt as follows, always assuming a downdip direction
of 280◦ East of North: VV, 9◦ in the lowest flows, de-
creasing to 5.5◦ at the top; VS 7◦; ST 7.5◦; MK 6.5◦
at the base (including MM), decreasing to 3.5◦.
The mean field direction has been calculated for
the collection of Table 2, after exclusion of the units
VV 3B, ST 1A, ST 8 (probable duplicate sampling of
the field) and VV 10 (poor stability). The mean di-
rection of the remaining 72 flows has a vector sum of
67.8, a declination of 6◦, and an inclination of 80.4◦,
as compared to a central axial dipole inclination of
about 77◦. Their virtual poles are shown in Figure
6. The mean virtual pole is at latitude 82.3◦, longi-
tude 351◦E and is therefore "near-sided" in contrast
to mean poles from several other studies in Iceland.
Wilson and McElhinny (1974) noted that the mean
pole from the Norðurdalur profiles Q to V is also near-
sided. Their explanations for these variations in mean
pole positions were somewhat speculative, and more
data are needed to resolve the matter.
JÖKULL No. 55 27