Jökull - 01.12.2006, Síða 54

Kristjánsson et al.
Eyjafjörður. In both areas the accumulation rate is
much lower than in the Tröllaskagi pile, which was
1.0 km/Ma in its lower part and 3.8 km/Ma in its up-
per part (Saemundsson et al. 1980).
According to the results presented above, we must
suspect that a major unconformity or hiatus is lurk-
ing within our profiles somewhere between the lowest
reversal (the Sólheimar/Merkidalur reverse-to-normal
transition at TB 59) and the Skati rhyolite tuff (at
the normal-to-reverse transition above PO 22). This
unconformity might indicate a suggested eastwards
jump of the active spreading zone to its current lo-
cation in North Iceland at around 7 million years ago
(Sæmundsson 1979, Hardarson et al. 1997, 1999). It
could then correlate with the major unconformities
found in the Fnjóskadalur valley (at the eastern bor-
der of Figure 1a) and near the town of Borgarnes in
West Iceland. But where is it and how can it be recog-
nized? It should be characterized by: 1) an erosional
surface 2) a sedimentary horizon 3) an accompanying
flexure zone, and preferably 4) different dips at each
side of the unconformity. Two main possibilities exist
fulfilling the first three requirements, but not the last
one:
a) The Merkidalur sedimentary layer (at TB 85-
86A and PG 43-46 in Figure 3)
b) The Tinná lignite sediments (at PO 22-23 in
Figure 3)
Both these layers are described by Hjartarson (2003)
who suggested that the Merkidalur sediment, rather
than the Tinná lignite sediment, represents this un-
conformity. Acid rock formations, suggested to be
a part of the Tinná Central Volcano, are found be-
low the Tinná lignite sediment, indicating continuity
in the volcanism. Furthermore, steep dips observed
in the outer Skagafjörður valleys, possibly represent-
ing a flexure or local loading by a central volcano, fit
better to the Merkidalur than the Tinná sediment. Ac-
cording to this we infer that an unconformity hides
between layers PG 43-46, representing a 1-2 million
year gap in the lava pile and possibly a rift jump. Time
gaps may also occur elsewhere in our composite pro-
file, for instance corresponding to the sediments at the
base of our profile PO, and/or farther up along Aust-
urdalur.
Additional studies are clearly needed to resolve
this question and various others in the strata of North
Iceland. Thus, an unconformity seems to hide in the
Eyjafjörður profiles of Kristjánsson et al. (2004); the
preliminary Ar-Ar dates of Hardarson et al. (1999) in-
dicate that it may be found in the upper part of the
VE-profile or the lower part of the GR-profile of Fig-
ure 1a, 30 km east of our profiles. It remains also to be
seen in detail how the ages of rocks immediately over-
lying and underlying the thick sedimentary rock units
described by Jóhannesson (1991) in the lower part of
the composite section of Kristjánsson et al. (2004, top
of profile AS and base of profile SG of Figure 1a) or
of the sedimentary rocks themselves, correspond to
those at the Merkidalur sediments. The accuracy of
the methods employed so far for age estimation is in-
sufficient to allow any definite correlations or conclu-
sions to this effect.
SUMMARY AND GENERAL
DISCUSSION
The paper describes the strata of six mountainside
profiles in lava flow sequences in the Norðurárdalur
and Austurdalur valleys in the Skagafjörður district.
This study has its basis mainly in unpublished strati-
graphic mapping of Ágúst Guðmundsson, while its
southern part is also based on the Ph.D. thesis of Árni
Hjartarson. It extends and supplements similar work
carried out by Saemundsson et al. (1980) to the north
and by Kristjánsson et al. (2004) in Eyjafjörður to the
east. Characteristic directions of magnetic remanence
were measured in 250 lava flows, generally yielding
very satisfactory internal agreement after appropriate
demagnetization treatment. Polarity reversals in the
lava pile turned out to be quite helpful for correlation,
when applied in conjunction with geological features,
both between some of these profiles and with profiles
from the previous studies. The directions in 2–6 suc-
cessive lavas are often clustered, which along with the
common absence of interflow sedimentary rocks indi-
cates that the extrusive volcanism in the area tended
to take place episodically. The mean virtual paleo-
52 JÖKULL No. 56