Jökull - 01.01.2009, Page 58

J. T. Andrews and J. Harðardóttir
The Iceland sites show a strong negative asso-
ciation between the NRM Median Destructive Field
(MDF) (r2 = 0.86) and the ARM(J0)/ARM(J20) ra-
tio (Figure 3D) indicating the importance of magnetic
grain-size in controlling the impact of NRM demag-
netization steps. The MDF has a range between 8.8
and 32.6 mT (Appendix 2). The East Greenland sites
show little variation in the MDF (Figure 2D) whereas
there is considerably more variability along the Ice-
land margin in both parameters (Figures 2C and 2D).
This difference in ARM and NRM demagne-
tization response between the two areas is also
seen in a strong, positive correlation between
ARM(J0)/ARM(J20) and NRM(J0)/NRM(J60) for
the Iceland sites (Figure 3C), but the East Greenland
sites show, if anything, a negative correlation. The
Iceland data group into three distinct entities with
#1232 being an outlier with high values for both ra-
tios. The middle group consists of sites from Djúp-
áll (#335 and #336), Ísafjarðardjúp (#342), and inner
Húnaflói (#330) (Figure 1), whereas the group with
the smallest response to AF demagnetization consists
of samples from North Iceland (#317, #321, #324),
inner Ísafjarðardjúp (#339), and Southwest Iceland
(#347). The East Greenland data grade from #1205
across the shelf to #1216 with ARM(J0)/ARM(J20)
values increasing and NRM(J0)/NRM(J60) ratios de-
creasing although only from !10 to !7 (Figure 3C).
Thus the East Greenland samples portray a different
response to progressive demagnetization than the sed-
iments from the Iceland margin.
4. Paleomagnetic Inclination
Assuming that the cores penetrated vertically, a first-
order measure of the expected paleomagnetic inclina-
tion for the field area (Figure 1) can be calculated on
the basis of a simple geocentric axial dipole model
(Butler, 1992). This gives values of!77 to 79!. Vari-
ations from this can be attributed to a non-dipole ge-
omagnetic field and to secular variations (Stoner and
St-Onge, 2007).
The median MAD values indicate that a substan-
tial fraction of the measured paleomagnetic secular
variations lie well within accepted limits (i.e. >10!)
(Stoner et al., 2007). However, low MAD values
characterized the upper sections of some cores where,
because of the typically disturbed nature of the sed-
iment, inclinations were clearly incorrect (i.e. very
low or negative inclinations). The median MAD val-
ues from the East Greenland sites (Figure 2E) are
low (<4!) but they are more variable in the Iceland
sites. The median inclination (Figure 2F) correlates
extremely well with the median inclination after the
30-mT-demagnetization step (r2 = 0.87). Only site
#1216 has inclinations that lie well below the ex-
pected range; it is noteworthy that this site has by far
the lowest rate of sediment accumulation but it might
also be due to the core not being taken vertically. The
median characteristic inclination values for the last
10,000 14C years lie between 65! and 85!. There is
no significant difference between median inclination
values from East Greenland versus Iceland, although
four of the East Greenland sites have the lowest me-
dian inclinations – this might be associated with the
differences in sediment disturbance between a gravity
corer (JM96) and a piston corer (B997) (Skinner and
McCave, 2003).
5. Normalized magnetic intensity
The strength of the remanent magnetism in sediment
is a function of the Earth’s magnetic field strength at
the time the signal gets locked into the sediments as
well as the magnetic character of the sediment itself.
Because of the strong correlation between magnetic
susceptibility and kARM we present magnetic inten-
sity only as NRM(J30)/ARM(J30) = NI (30). There is
an increased interest in the use of this and similar mea-
sures as potential chronological tools in deep sea and
lacustrine settings (Stoner et al., 2002). The NI(30)
data has no clear geographic pattern (not shown).
Median NI(30) range between 0.1 and 0.4 with a
distinct bimodal distribution, but there is no system-
atic difference between the Iceland and East Green-
land sites, nor is there any pattern to the CV% val-
ues. This observation strengthens the possibility that
NI(30) might have utility in the correlation between
marine sediment records in this area of the North At-
lantic (Stoner et al., 2007).
58 JÖKULL No. 59