646
Bleil and von Dobeneck
features from about 200 to 140 ka and possibly
also around 50 ka, yet both stacks entirely diverge
for the rest of the time (Fig. 6).
Discussion
The comparison of the Ceara Rise polanty patterns with published records involves a variety of
problems. An equivalent reliable age control and
elaborate paleomagnetic analysis would be required
for every data set. Even if former paleomagnetic
methods adequately meet modern standards, older
age models often do not. Unfortunately, there are
generally insufficient means to assess appropriate
revised ages.
As exemplified by the present results, substantial discrepancies also exist between data sets
from neighbouring sites despite a common highresolution age framework and an overall homogeneous depositional environment. These findings rise
the fundamental question: where and when do
sediments acquire a stable remanent magnetization? Or, in other words, how is the age of a
sediment layer related to the age of its ChRM?
Diagenetic processes may be a serious complication in this respect (van Hoof and Langereis 1991;
van Hoof et al. 1993; Tarduno and Wilkison 1996;
see also Frederichs et al. this volume). According
to the available data, they should be of very minor
importance in the Quaternary Ceara Rise deposits, however.
Delayed Remanence Acquisition
A sediment NRM is explained by the general concept of post-depositional detrital magnetization
(PDRM) acquisition (Irving and Major 1964). This
simple physical model implies that magnetic mineral grains become gradually fixed in permanent
positions within the sediment matrix due to compaction and closure of interstitial voids. In the
locked state they are no more free as before to
accommodate for directional changes of an external field according to their magnetic momentum.
For small magnetic grains, carrying the most stable remanent magnetization, the individual lock-in
depth is hence deeper seated than for coarse
grained fractions. To some extent the basic paleomagnetic assumption identifying the stable component as the primary remanence will therefore
always be violated. Lithology and especially grainsize distribution of the sediment matrix, sedimentation rate and bioturbation are important factors
controlling variations in lock-in depth position. In
sediments with a relative to their magnetic mineral
assemblage coarse grained matrix, containing for
instance high concentrations of hollow foraminiferal
tests, it will diffuse over an expanded depth range.
For prolonged periods these systems remain open
to acquire a remanent magnetization. Short-lived
geomagnetic events may normally not be recorded
then, at least they are largely set off from their
actual age position. Due to this effect for example
in a large collection of late Quaternary biogenic
carbonate sediments from the eastern South Atlantic almost no polarity events could be detected
(ThieBen 1993). Where the matrix comprises high
amounts of fine grained components, e.g., terrigenous clays, the lock-in process will be confined
to a much narrower depth interval. These sealed
systems should document geomagnetic events
fairly accurate. Many marine sedimentary deposits at high latitutes are of this type. Interestingly,
Clement et al. (1996) noted that for ODP and
DSDP core materials optimum Pliocene -Pleistocene magnetostratigraphic data were regularly
achieved on terrigenous sediment series.
At the Ceara Rise, in a variable, climatically
controlled sedimentary environment with alternating layers of predominantly biogenic carbonates
and terrigenous components, the probability to positively register geomagnetic events should have frequently changed during the late Quaternary. Most
coherent in all cores is the short reversed episode
at 139.7 to 135.3 ka (75% reversed polarity level)
correlated to the Baffin Bay event. The best estimate for the age of its termination would be
130.2 ka, when the latest polarity transition is
observed at the shallowest lock-in depth (core
GeoB 1516-2, Fig. 7). With the same reasoning
the Baffin Bay event began at 139.3 ka (core
GeoB 1519-1) and hence lasted for 9.1 kyr. As the
lock-in depths in both cores most likely are also
>0, the reversal boundaries should actually be
somewhat younger. A calculation of apparent
lock-in depths in the other cores relative to these
Bleil and von Dobeneck
features from about 200 to 140 ka and possibly
also around 50 ka, yet both stacks entirely diverge
for the rest of the time (Fig. 6).
Discussion
The comparison of the Ceara Rise polanty patterns with published records involves a variety of
problems. An equivalent reliable age control and
elaborate paleomagnetic analysis would be required
for every data set. Even if former paleomagnetic
methods adequately meet modern standards, older
age models often do not. Unfortunately, there are
generally insufficient means to assess appropriate
revised ages.
As exemplified by the present results, substantial discrepancies also exist between data sets
from neighbouring sites despite a common highresolution age framework and an overall homogeneous depositional environment. These findings rise
the fundamental question: where and when do
sediments acquire a stable remanent magnetization? Or, in other words, how is the age of a
sediment layer related to the age of its ChRM?
Diagenetic processes may be a serious complication in this respect (van Hoof and Langereis 1991;
van Hoof et al. 1993; Tarduno and Wilkison 1996;
see also Frederichs et al. this volume). According
to the available data, they should be of very minor
importance in the Quaternary Ceara Rise deposits, however.
Delayed Remanence Acquisition
A sediment NRM is explained by the general concept of post-depositional detrital magnetization
(PDRM) acquisition (Irving and Major 1964). This
simple physical model implies that magnetic mineral grains become gradually fixed in permanent
positions within the sediment matrix due to compaction and closure of interstitial voids. In the
locked state they are no more free as before to
accommodate for directional changes of an external field according to their magnetic momentum.
For small magnetic grains, carrying the most stable remanent magnetization, the individual lock-in
depth is hence deeper seated than for coarse
grained fractions. To some extent the basic paleomagnetic assumption identifying the stable component as the primary remanence will therefore
always be violated. Lithology and especially grainsize distribution of the sediment matrix, sedimentation rate and bioturbation are important factors
controlling variations in lock-in depth position. In
sediments with a relative to their magnetic mineral
assemblage coarse grained matrix, containing for
instance high concentrations of hollow foraminiferal
tests, it will diffuse over an expanded depth range.
For prolonged periods these systems remain open
to acquire a remanent magnetization. Short-lived
geomagnetic events may normally not be recorded
then, at least they are largely set off from their
actual age position. Due to this effect for example
in a large collection of late Quaternary biogenic
carbonate sediments from the eastern South Atlantic almost no polarity events could be detected
(ThieBen 1993). Where the matrix comprises high
amounts of fine grained components, e.g., terrigenous clays, the lock-in process will be confined
to a much narrower depth interval. These sealed
systems should document geomagnetic events
fairly accurate. Many marine sedimentary deposits at high latitutes are of this type. Interestingly,
Clement et al. (1996) noted that for ODP and
DSDP core materials optimum Pliocene -Pleistocene magnetostratigraphic data were regularly
achieved on terrigenous sediment series.
At the Ceara Rise, in a variable, climatically
controlled sedimentary environment with alternating layers of predominantly biogenic carbonates
and terrigenous components, the probability to positively register geomagnetic events should have frequently changed during the late Quaternary. Most
coherent in all cores is the short reversed episode
at 139.7 to 135.3 ka (75% reversed polarity level)
correlated to the Baffin Bay event. The best estimate for the age of its termination would be
130.2 ka, when the latest polarity transition is
observed at the shallowest lock-in depth (core
GeoB 1516-2, Fig. 7). With the same reasoning
the Baffin Bay event began at 139.3 ka (core
GeoB 1519-1) and hence lasted for 9.1 kyr. As the
lock-in depths in both cores most likely are also
>0, the reversal boundaries should actually be
somewhat younger. A calculation of apparent
lock-in depths in the other cores relative to these