Geomagnetic Events and Relative Paleointensity Records
645
event which has previously been observed in different distant parts of the world. No unequivocal
trace could be detected of the Biwa I (Jamaica)
event with a postulated age of around 175 ka and
also repeatedly reported from various locations.
Between about 250 and 185 ka several horizons
of south oriented ChRM declinations were encountered in the different cores. Because of their poor
agreement in age, the stack rises just over 25%
reversed polarities (2 of 7 cores) only at around
199 and 248 ka. These findings could correspond
to the Biwa I and a short event at 235 ka identified
in marine sediments ofthe Fram Strait (Nowaczyk
1991). As there are convincing arguments that the
Blake event is younger than oxygen isotope stage
boundary 5/6 (e.g., Tucholka et al. 1987), the exceptionally well confined reversed interval at 137.5
± 2.2 ka should best be correlated to the Baffin
Bay event for which ages ranging from 161 to
129 ka have been determined at several northern
high latitude sites.
Relative Paleo intensities
According to the AF demagnetization analysis
(Fig. 3) at least 35 mT are necessary to isolate
single-component remanences in the Ceara Rise
sediments. Relative paleointensities were therefore
determined from remanent magnetization intensities
at this level which is somewhat higher than in most
other studies (typically 15 mT). To account for
variable magnetic mineral concentrations, they
have been normalized to susceptibility (anhysteretic
remanent magnetization data are not yet available
for this purpose at present).
Relative paleointensities fluctuate by about one
order of magnitude at each site (Fig. 5) in accord
with previous reports. The large majority of prominent variations is directly associated with changes
in polarity. Most, though not all, noticeable minima
coincide either with singular short reversed episodes amid opposite polarities or with reversal
boundaries, where geomagnetic events last for
longer periods oftime. In various intervals of the
records there is some evidence for trends resembling the so-called' saw-tooth' pattern, e.g., in oxygen isotope stage 9 prior to the Biwa II event and
in the predominantly reversed section of oxygen
isotope stage 5. Such a gradual decay of relative
paleointensities towards reversal boundaries followed by an abrupt increase immediately after the
polarity transition were most clearly documented
for the Gauss and early Matuyama Chrons at time
constants;::: 0.5 Myr. An analogous phenomenon
could not be identified in the Brunhes Chron (Valet
and Meynadier 1993) and is neither recognized in
the present data sets. Saw-tooth variations have
been interpreted as systematic changes in Earth's
magnetic field strength reflecting intrinsic characteristics of the geodynamo system. This view
was recently challenged, however. Mazaud (1996)
proposed very long delays in remance acquisition,
Tarduno and Wilkison (1996) discussed diagenetic
overprints as their potential primary causes.
Correlations (Pearson' s coefficients) of relative
paleointensity time series to sediment parameters
which are explicitly controlled by climate variations
range fromr = 0.06 (core GeoB 1517-1) to r = 0.38
(GeoB 1520-2) for concentrations of terrigenous
material and from r = 0.10 (GeoB 1517-1) to
r=0.22 (GeoB 1515-1) for bulk sedimentation
rates. In agreement with conclusions ofTauxe and
Shackleton (1994), a systematic correspondence
to orbital frequencies in the precession or obliquity
band and hence a link to climatic changes appears
highly unlikely, therefore (the records are too short
to be conclusive for eccentricity). Relative paleointensities should consequently relate to variations
in geomagnetic field, provided reasonably constant
adequate rock magnetic properties prevailed (the
otherwise entirely different situation will not be
discussed here).
Combining relative paleointensity records of
the different sites to a stack for the Ceara Rise
preserves most of their major characteristics and
thus a fairly straightforward parallelism to the
polarity stack. As in similar regional compilations
(e.g., Meynadier et al. 1992; Tric et al. 1992) the
scatter is continuously very high (Fig. 6) and
individual sites deviate significantly from predominant trends at various levels (Fig. 5). An inspection
of the Ceara Rise data together with the 'Sint200' synthesis ofGuyodo and Valet (1996), which
accumulates results of pelagic sediment series
spread over the globe at mid latitudes and covers
the last 200.000 years, reveals certain common
645
event which has previously been observed in different distant parts of the world. No unequivocal
trace could be detected of the Biwa I (Jamaica)
event with a postulated age of around 175 ka and
also repeatedly reported from various locations.
Between about 250 and 185 ka several horizons
of south oriented ChRM declinations were encountered in the different cores. Because of their poor
agreement in age, the stack rises just over 25%
reversed polarities (2 of 7 cores) only at around
199 and 248 ka. These findings could correspond
to the Biwa I and a short event at 235 ka identified
in marine sediments ofthe Fram Strait (Nowaczyk
1991). As there are convincing arguments that the
Blake event is younger than oxygen isotope stage
boundary 5/6 (e.g., Tucholka et al. 1987), the exceptionally well confined reversed interval at 137.5
± 2.2 ka should best be correlated to the Baffin
Bay event for which ages ranging from 161 to
129 ka have been determined at several northern
high latitude sites.
Relative Paleo intensities
According to the AF demagnetization analysis
(Fig. 3) at least 35 mT are necessary to isolate
single-component remanences in the Ceara Rise
sediments. Relative paleointensities were therefore
determined from remanent magnetization intensities
at this level which is somewhat higher than in most
other studies (typically 15 mT). To account for
variable magnetic mineral concentrations, they
have been normalized to susceptibility (anhysteretic
remanent magnetization data are not yet available
for this purpose at present).
Relative paleointensities fluctuate by about one
order of magnitude at each site (Fig. 5) in accord
with previous reports. The large majority of prominent variations is directly associated with changes
in polarity. Most, though not all, noticeable minima
coincide either with singular short reversed episodes amid opposite polarities or with reversal
boundaries, where geomagnetic events last for
longer periods oftime. In various intervals of the
records there is some evidence for trends resembling the so-called' saw-tooth' pattern, e.g., in oxygen isotope stage 9 prior to the Biwa II event and
in the predominantly reversed section of oxygen
isotope stage 5. Such a gradual decay of relative
paleointensities towards reversal boundaries followed by an abrupt increase immediately after the
polarity transition were most clearly documented
for the Gauss and early Matuyama Chrons at time
constants;::: 0.5 Myr. An analogous phenomenon
could not be identified in the Brunhes Chron (Valet
and Meynadier 1993) and is neither recognized in
the present data sets. Saw-tooth variations have
been interpreted as systematic changes in Earth's
magnetic field strength reflecting intrinsic characteristics of the geodynamo system. This view
was recently challenged, however. Mazaud (1996)
proposed very long delays in remance acquisition,
Tarduno and Wilkison (1996) discussed diagenetic
overprints as their potential primary causes.
Correlations (Pearson' s coefficients) of relative
paleointensity time series to sediment parameters
which are explicitly controlled by climate variations
range fromr = 0.06 (core GeoB 1517-1) to r = 0.38
(GeoB 1520-2) for concentrations of terrigenous
material and from r = 0.10 (GeoB 1517-1) to
r=0.22 (GeoB 1515-1) for bulk sedimentation
rates. In agreement with conclusions ofTauxe and
Shackleton (1994), a systematic correspondence
to orbital frequencies in the precession or obliquity
band and hence a link to climatic changes appears
highly unlikely, therefore (the records are too short
to be conclusive for eccentricity). Relative paleointensities should consequently relate to variations
in geomagnetic field, provided reasonably constant
adequate rock magnetic properties prevailed (the
otherwise entirely different situation will not be
discussed here).
Combining relative paleointensity records of
the different sites to a stack for the Ceara Rise
preserves most of their major characteristics and
thus a fairly straightforward parallelism to the
polarity stack. As in similar regional compilations
(e.g., Meynadier et al. 1992; Tric et al. 1992) the
scatter is continuously very high (Fig. 6) and
individual sites deviate significantly from predominant trends at various levels (Fig. 5). An inspection
of the Ceara Rise data together with the 'Sint200' synthesis ofGuyodo and Valet (1996), which
accumulates results of pelagic sediment series
spread over the globe at mid latitudes and covers
the last 200.000 years, reveals certain common