Jökull - 01.01.2001, Page 45

Paleomagnetic studies in Skarðsheiði, SW-Iceland
characteristic of the curves is a Curie point exceed-
ing 540 C, indicating fairly pure magnetite (oxidized
in SJ 18–1). However, in some of the samples one
sees that a component with a lower Curie point (350–
500 C) is being irreversibly converted (presumably
exsolved) on heating. In many lava samples from
other Tertiary areas in Iceland which are unstable to
AF treatment, the thermomagnetic curves are more
complex (L. Kristjánsson, unpublished data) and in-
clude components with Curie points below 350 C.
It should be noted that the thermomagnetic behav-
ior of a sample is not decisive when it comes to evalu-
ating whether a particular remanence direction is a re-
liable indicator of the primary geomagnetic field. The
effective titanomagnetite grain size, grain shape and
interactions between grains (or lamellae) may be just
as important factors in the preservation of the primary
remanence through time as the chemical composition
of the major magnetic components. Our evaluation
of the reliability of the remanence direction in a lava
is based mostly on its stability to AF treatment and
its consistency between the different samples. The
chemical changes which occur during heating in our
Skarðsheiði samples (Figure 7) indicate however, that
they are not very good material for paleomagnetic
field–intensity measurements.
SUMMARY AND CONCLUSIONS
Skarðsheiði has been considered to belong mostly to
the R5 and N5 polarity zones of Einarsson (1957). In
his Ph.D. thesis on the area, Franzson (1978) assumed
that these zones correspond to the upper Gilbert and
lower Gauss chrons respectively. The boundary has
in recent years been estimated to be 3.58 M.y. in age
(Cande and Kent, 1995). Intermediate geomagnetic
poles in several successive flows at this transition
were first reported by Wilson et al. (1972) in the
North–Eastern part of Skarðsheiði. They are shown
in diagrams by Wilson et al. (1972, Figure 5) and Da-
gley and Lawley (1974) but the detailed results have
not been published. The presence of intermediate
poles was confirmed by the more extensive sampling
of Kristjánsson and Sigurgeirsson (1993) in their pro-
file SH in Villingadalur.
In this paper we present a composite profile SI/SJ
mapped through the exposed succession of South–
Western Skarðsheiði, and from a short profile SS far-
ther east. We describe paleomagnetic work on 51
units in SI, 41 in SJ, 17 in SS, and 12 in another
short profile SW across Villingadalur from SH. The
paleomagnetic work includes extended demagnetiza-
tion treatment of some samples, and thermomagnetic
measurements. The paleomagnetic directions are gen-
erally stable and represent the primary thermal rema-
nence of the lavas; this is demonstrated by the ex-
cellent within–flow directional agreement after AF
cleaning treatment in the laboratory, also in flows
where other magnetic properties of the samples vary
considerably. Flows yielding low–latitude VGPs have
relatively low values of remanence intensity, as well
as the most altered flows in the lower part of profile
SI (Table 1), but their reliability appears to be quite
satisfactory.
Along with the mapping by Franzson (1978) at
various locations in Skarðsheiði, our results show that
the R5–N5 boundary transects it at altitudes varying
between about 500 and 800 m. In the upper part of R5,
feldspar–porphyritic lavas are common. The polarity–
transition zone becomes thinner towards the south and
west. In Skarðsheiði, we have also documented the
presence of two subchrons, probably the Cochiti in the
lowest accessible exposures (SI 1–14; lower boundary
not seen) and the Mammoth at the top (SJ 34–42; up-
per boundary not seen). Additionally, some major ge-
omagnetic excursions (which may well represent in-
completely recorded subchrons) are noted, the main
ones being at SI 16 (close to a polarity zone bound-
ary) and 28B, 32, and 47 within R5.
We present a summary of polarity results from
Akrafjall and Skarðsheiði where the lowest clearly
normal lavas in the N5 polarity zone have been
aligned (Figure 8). Also shown are positions of
some distinct lava groups and conglomerates which
may possibly be correlated. The thickness of the
porphyritic series (PF) in profile C is unknown but
according to Franzson (1978) it is about 200 m in
Skessuhorn (Figure 2). Reversely magnetized com-
pound flows (CP) similar to some of those in Akra-
fjall are also seen at various localities in Skarðsheiði
(Franzson, 1978) but not in our profile SI.
JÖKULL No. 50 45