Jökull - 01.01.2010, Qupperneq 159

Paleomagnetic observations in SW- and S-Iceland
Figure 8. Stratigraphy of profiles HR and FO, sim-
plified from an annotated drawing by Á. Geirsdóttir
(pers. comm., 1996), with additional input from Th.
Thordarson (pers. comm. 1996, 2001), Á. Guðmunds-
son (pers. comm., 1992) and the author. The magnetic
polarities of the lavas are mostly based on laboratory
measurements but partly on fluxgate measurements in
the field. Thin sedimentary layers are assumed to have
the same polarity as adjacent lavas. The numbers on
the left-hand side refer to cumulative thicknesses (in
meters), not altitudes. – Jarðlög í sniðunum HR og
FO, ásamt segulstefnu í þeim. Súlan er einfölduð úr
teikningu frá Áslaugu Geirsdóttur og Þorvaldi Þórð-
arsyni. Tölur eiga við samanlagðar þykktir, ekki hæð.
The scheme presented in Figure 8 is a tentative
one, as exposures in the Fossnes profiles are some-
what incomplete. Much additional work is therefore
needed before the history of interacting volcanism and
glaciations in this area can be fully understood. As-
suming with reference to the stratigraphic and dating
results of Kristjánsson et al. (1998) that this sequence
was emplaced in the Matuyama reverse chron, Fig-
ure 8 appears to record at least three normal-polarity
events in a 500 m thick profile. The presence of three
magnetic events is in contrast to for instance Figure 3
where only one reversal is found in a 350 m thickness
of lava and clastic rocks. This indicates a higher rate
of eruptive activity in FO/HR than in the WB profile,
provided that the normal-polarity zones in FO/HR are
not due to stratigraphic complexities.
SECULAR VARIATION RESULTS
It is notable that few mid–or low-latitude virtual geo-
magnetic poles are found in the presumed Pleistocene
lavas in the three areas studied here. The VGP in
only one out of some 90 lavas of Appendix 1 (ex-
cluding probable duplications in KI and in FO/HR)
is positioned in a latitude lower than 40◦ N or S,
and two VGPs lie between 40◦ and 50◦ in latitude,
while in lava sequences of more than 5 Ma age in Ice-
land around 20% of poles are found below 50◦ (cf.
Kristjánsson, 2008). This is one aspect of a long-term
change occurring in the character of the paleomag-
netic field as recorded in Icelandic lavas. This change
has also caused the scatter (a.s.d.) of VGP positions in
published lava collections (Kristjánsson, 2008, 2009
and additional data) about their respective means to
decrease significantly during the last 15 Ma (Figure
9). Results of this kind can provide valuable con-
straints on models of the behavior of the geomagnetic
field through time.
SUMMARY AND DISCUSSION
The present paleomagnetic study of three small areas
within lava series of less than 3 Ma age in southwest
and south Iceland has similar aims as some previous
laboratory paleomagnetic studies of Icelandic lavas in
this age range, published for instance by Doell (1972),
Wilson et al. (1972), Watkins et al. (1975), Kristjáns-
son et al. (1980, 1988, 1998), Eiríksson et al. (1990),
Udagawa et al. (1999), and Helgason and Duncan
(2001). Primarily, they have attempted to locate geo-
magnetic excursions and polarity-zone boundaries,
which might be radiometrically dated and used in
stratigraphic correlations between igneous units. Such
correlations are especially difficult in areas affected
by glaciations during eruptions, due to the hetero-
geneity of products, lack of overlap between units
from different vents, and landscapes formed by con-
temporaneous erosion. The few radiometric age deter-
minations so far available from the Pleistocene in Ice-
land do not have sufficient resolution to allow definite
correlations between them, especially as the number
of suggested short intervals of normal polarity during
the Matuyama chron in the literature has been increas-
ing in recent years. Therefore, the present study and
those listed above represent only very small steps to-
wards stratigraphic mapping of the Pleistocene vol-
canism in Iceland. Further efforts of a similar nature
should be undertaken in close collaboration with spe-
cialists in petrology, sedimentology and glacial geol-
ogy. Eventually, correlations with an improved inter-
national geomagnetic polarity time scale might also
be attained.
Among the results presented in this paper is the
confirmation of a boundary between thick reverse
and normal polarity zones in the Botnssúlur moun-
tain (Figure 3). The paper also contains the first de-
scription of a similar boundary which can probably
JÖKULL No. 60 159