Jökull - 01.12.2006, Blaðsíða 46

Kristjánsson et al.
Figure 1c. A cross-section of the landscape along C-D of Figure 1b, indicating some of the stratigraphic cor-
relations between our profiles. – Þverskurður af landslagi, með sýnasöfnunarsniðum og nokkrum tengingum
milli þeirra.
magnetization that is commonly present in the rock.
Additionally, we applied 25 mT demagnetization as a
routine in 58 lavas, and samples from 8 of these were
further demagnetized at 30, 40 and 50 mT. In gen-
eral, the 20 mT treatment yielded the lowest value for
the 95% confidence radius (α95) of the flow-mean di-
rections. The within-unit agreement of directions is
excellent: the 95% confidence radii are generally well
below 10◦ as is seen from Table 1, and their r.m.s.
value is 5◦. Only 13 individual samples had to be dis-
carded due to instability or within-flow inconsisten-
cies. Treatment to 30 mT and beyond resulted in mi-
nor or no improvement in directional agreement (< 1◦
reduction in α95) and small random changes in mean
directions (of order 1◦ of arc). Hence, there is nothing
to gain in quality here by applying extensive demag-
netization or sophisticated statistical data processing
techniques.
The tectonic tilt in the Norðurárdalur-Austurdalur
area is mostly between southwest and southeast (Fig-
ure 1b). It reaches 10◦ or more at the level of the
Austari Jökulsá river and decreases with altitude, to a
few degrees at mountain tops. However, the magni-
tude and direction of dips exhibit considerable lateral
variations. It is not always possible to measure the
dip vectors in the profiles themselves, and therefore
we have used our estimates of regional values (see
the Appendix) for correcting the paleomagnetic direc-
tions. The angular uncertainty of the dip corrections
may reach 4-5◦ in the lowermost parts of the profiles;
this is the major source of possible errors in our results
regarding the geomagnetic field direction and virtual
pole positions at the time of emplacement.
RESULTS OF THE PALEOMAGNETIC
MEASUREMENTS
Intensities. The arithmetic average remanence inten-
sity in the lavas after 10 mT alternating field treatment
is about 3.2 A/m, which is similar to that in other stud-
ies from Tertiary Icelandic lavas (Kristjansson 2002,
Figure 3). Low-field susceptibility at room tempera-
ture was measured in one sample from each flow with
a Bartington MS2 instrument. The arithmetic aver-
age value of this parameter was about 0.023 SI units,
which is also similar to averages from other areas in
Iceland. High susceptibility values (> 0.04) tend to be
found in samples with weak and unstable remanence.
No thermal demagnetization or thermomagnetic ex-
periments were carried out.
Grouping of lava directions. Inspection of the data
set of Table 1 reveals that very commonly two or more
consecutive lavas have similar remanence directions
(within 10◦ or so), as previously noted also in Eyja-
fjörður (Kristjánsson et al. 2004).
In some cases even five or six lavas are involved,
such as TB 44-48 and 66-70, TD 6-11, PH 27-32 and
44 JÖKULL No. 56