Jökull - 01.12.1994, Side 14

profiles, we are left with the following lava flows in a
composite section of about 2600 m total thickness:
DO 1-50, DA 5-17, DB 2-37, DC 7-15, DD 2-21, DE
3-11, DF 5-26, DG 5-23, DH 2-9, DK 10-27, DL 3-21
and DM 1-14. The profiles are generally steep and
well exposed, so that there are no major gaps in the
sampling of this sequence.
The paleomagnetic pole latitudes in the composite
section are shown as function of stratigraphic height
in Fig. 8. A simplified polarity column, constructed
from these latitudes using the criteria of McDougall et
al. (1984, p. 7041) is also shown.
The compound flow DO 18 near the base of our
ísafjarðardjúp composite section (Fig. 3) was linked
to the compound flow SB 24-29 in the profile of Mc-
Dougall et al. (1984) in Mt. Búrfell. This link which
made use of the magnetic polarity of the lava flows
and their petrography, has been confirmed in the field
by careful mapping. At both localities the compound
flow is underlain by acid tuff. We correlate the long
normally magnetised zone between DO 26 and DA 17
with that in profiles JA and JB of McDougall et al.
(1984), which is along the estimated strike direction.
The normal zone in the upper part of our profile DC
would then correlate with the normally magnetised
flows in McDougall et al. 's JA through KD, and the
normal zone from DF 12 to DH 9 may correspond to
the upper part of their JD and all JE. The correlation
between the two composite profiles is satisfactory, es-
pecially in the lower half of the section. The total
thickness is comparable but individual magnetic po-
larity chrons vary in thickness as can be expected due
to lateral and/or down-dip variations.
The sediment horizon between lavas DL 2 and 3 is
below a zone of normal magnetisation, which is also
the case with the Brjánslækur sediments at JF 46-48
in McDougall et al. (1984). The sediments are not
seen at DK 27-28 as would be expected but such sed-
imentary horizons often vary greatly in thickness and
may be missing altogether in places. Above DM 6,
four reversals occur in the eight lava flows we cored,
which does not correspond with the polarity sequence
in the uppermost part of profile JF. This may be be-
cause of a slowing up of the rate of eruptions in the
area of profile DM. Altematively, this part of the pro-
file which is on a rather flat plateau, may be affected
by faults hidden by overburden. The total number of
reversals in the composite section is at least ten, but
there is uncertainty how the episode of unstable field
directions and the reversals at the top of the section
should be counted.
Age range of the composite section
Using the correlations above, we can obtain the
age of our composite section from the K/Ar dates
published by McDougall et al. (1984). The base of the
section is close to the base of McDougall et al.'s pro-
file SB, of about 13.7 Ma age. The top of profile DM
may be expected to lie in the top reverse zone of their
profile JF, i.e. at about 12.0 Ma. This yields a mean
rate of buildup of the lava pile of 2600 m/1.7 Ma, or
1500 m/Ma which is not very different from the 1820
m/Ma found for McDougall et al.'s westem section.
McDougall et al. (1984) sampled two profiles
named SZ and SF in Kvígindisfjörður and Kolla-
fjörður, southwest of our profiles DK to DM. No ra-
diometric dates were obtained on these profiles, but
they were expected to correspond approximately to
JF. We correlate tentatively the normal zone in SZ
with that in our profiles DL and DM, but additional
mapping is required.
Our polarity column does not exhibit straightfor-
ward correspondence with previous published geo-
magnetic polarity time scales. A similar statement
was made by McDougall et al. (1984, p. 7056) on the
polarity column of their western section. The most
likely reason for this may be that in the interval before
10 or 12 Ma ago the field changed polarity more often
than is generally acknowledged, possibly exceeding 8
reversals per Ma. In oceanic anomaly lineations of
this age the magnetic signal will be attenuated, as is
evident north of Iceland (Vogt et al. 1980). However,
the observed field variations are still being interpreted
in the literature in terms of magnetic chrons of aver-
age length 0.2- 0.4 Ma. In the lava series, on the other
hand, the episodic nature of the volcanism will create
fairly thick piles during some chrons while others are
not recorded at all.
McDougall et al. (1984, Fig. 8) concluded from
their mapping and dating results on the east coast of
12
JOKULL, No. 44