Sediment Redistribution, 23°Th" - Nonnalization and Implications for the Reconstruction of Particle Flux 413
Diester-Haass 1983) it may occur despite the very
short residence time of230Th. Francois et al. (1990)
were able to determine a maximum contribution of
boundary scavenging-derived 230Th" of 50 % in
excess of the expected 230Thex flux into sediment
traps which received particle fluxes in the order of
10 g/cm2 kyr coinciding with strong coastal
upwelling. Similar excesses 0[230Th ex fluxes into
sediment traps caused by an advectional 230Thex
supply into the high biogenic particle flux area of
the Bransfield Strait and the King George Basin
were reported by Rutgers van der Loeff and Berger
(1991 ). Such an advective water column transport
of 230Th can also occur when the residence time
of water masses in extremely low particle flux areas, such as the Weddell Sea, is in the same order
of magnitude as the water column residence time
0[230Th (Rutgers van der Loeff and Berger 1993).
A certain build-up of 230 The x in the water column
may then occur which is exported from the Weddell
Sea (Walter et al. 1997) and not scavenged until
this enriched water mass reaches an area ofhigher
particle flux rates. There, 23 0 Th,x is scavenged to
the sediments in excess of its local production rate.
Similarly, with a given deep-water residence time
in the Atlantic Ocean of about 80 years a maximum
export of about 15 % of all the 23 0 Th produced in
the Atlantic Ocean to the Southern Ocean by
advective transport with the North Atlantic Deep
Water was modelled recently by Yu et al. (1996).
A related phenomenon leading to a decrease in the
flux of 230Th" was recently described. The contribution of 230Th" -depleted, young former surface
water masses to the deep water formed in the NE
Atlantic was identified as the reason why a number
of sediment trap profiles showed a deficiency of
230Thcx flux into the deeper traps (Vogler 1997;
Moran et al. 1997). Correction for this effect increased the calculated efficiency of the deep traps
from about 50 % to 80 %. A detailed discussion of
the water column processes affecting 230Th and
231 Pa in the Southern Ocean can be found in the
contribution by Walter in this volume.
In summary, the flux of 230Th" into the
sediments may be changed by other factors than
sediment redistribution in areas of strong water
mass advection and upwelling and spatial contrasts
of particle fluxes. In these areas only fluxes of
230Th" which differ from their locally expected values by more than at least 50 % should be interpreted
in terms of sediment redistribution.
Results and Discussion
Sediment Mass Accumulation Rates in Low
Resolution Sediments of the Ceara Rise
Mass accumulation rates of Late Quaternary marine sediments are usually calculated by measuring the 8180 profiles of single foraminiferal species,
and by ascribing certain features and events to the
SPECMAP standard 8 18 0 stacked record (Imbrie
et al. 1984; Martinson et al. 1987). This record was
itself dated by comparison to the well-known variations of the orbital parameters of the earth that
vary with a periodicity ofl9, 23, 41, and 100 kyr.
The error in dating single events of this oxygen isotope stratigraphy amounts to ±2.5 kyr on average,
and thus gives very reliable sediment accumulation
rates for the main isotope stages and substages. The
pursuit of an increased time resolution in sediments
with low sedimentation rates may, however, lead
to problems when this method is applied. The reason is that the age difference between subevents
of the stacked 8 18 0 record may come close to the
error of2.5 kyr which may result in large offsets
when the sedimentation rates and sample resolution are low. These uncertainties can result in an
artificial variability of the sedimentation rates (easily
in the range of 50-1 00 %) which then may become
the factor dominating in accumulation rate variability.
The application 0[2 30 Th" to calculate rain rates
was shown to be a valid approach in the area of
the Ceara Rise (western Equatorial Atlantic
Ocean). Six cores yielded fluxes of 23CYfh" very
close to the expected values, giving evidence for
an either low or even negligible sediment redistribution intensity (Francois et al. 1990). The calculation of the rain rates is not significantly affected
by the above described errors, although the decay
correction of the 230Th" concentration is carried out
using the 8180 stratigraphy. However, the half-life
of 23 0Th amounts to 75,200 years and therefore the
effect of a 2.5 kyr error is negligible.
Diester-Haass 1983) it may occur despite the very
short residence time of230Th. Francois et al. (1990)
were able to determine a maximum contribution of
boundary scavenging-derived 230Th" of 50 % in
excess of the expected 230Thex flux into sediment
traps which received particle fluxes in the order of
10 g/cm2 kyr coinciding with strong coastal
upwelling. Similar excesses 0[230Th ex fluxes into
sediment traps caused by an advectional 230Thex
supply into the high biogenic particle flux area of
the Bransfield Strait and the King George Basin
were reported by Rutgers van der Loeff and Berger
(1991 ). Such an advective water column transport
of 230Th can also occur when the residence time
of water masses in extremely low particle flux areas, such as the Weddell Sea, is in the same order
of magnitude as the water column residence time
0[230Th (Rutgers van der Loeff and Berger 1993).
A certain build-up of 230 The x in the water column
may then occur which is exported from the Weddell
Sea (Walter et al. 1997) and not scavenged until
this enriched water mass reaches an area ofhigher
particle flux rates. There, 23 0 Th,x is scavenged to
the sediments in excess of its local production rate.
Similarly, with a given deep-water residence time
in the Atlantic Ocean of about 80 years a maximum
export of about 15 % of all the 23 0 Th produced in
the Atlantic Ocean to the Southern Ocean by
advective transport with the North Atlantic Deep
Water was modelled recently by Yu et al. (1996).
A related phenomenon leading to a decrease in the
flux of 230Th" was recently described. The contribution of 230Th" -depleted, young former surface
water masses to the deep water formed in the NE
Atlantic was identified as the reason why a number
of sediment trap profiles showed a deficiency of
230Thcx flux into the deeper traps (Vogler 1997;
Moran et al. 1997). Correction for this effect increased the calculated efficiency of the deep traps
from about 50 % to 80 %. A detailed discussion of
the water column processes affecting 230Th and
231 Pa in the Southern Ocean can be found in the
contribution by Walter in this volume.
In summary, the flux of 230Th" into the
sediments may be changed by other factors than
sediment redistribution in areas of strong water
mass advection and upwelling and spatial contrasts
of particle fluxes. In these areas only fluxes of
230Th" which differ from their locally expected values by more than at least 50 % should be interpreted
in terms of sediment redistribution.
Results and Discussion
Sediment Mass Accumulation Rates in Low
Resolution Sediments of the Ceara Rise
Mass accumulation rates of Late Quaternary marine sediments are usually calculated by measuring the 8180 profiles of single foraminiferal species,
and by ascribing certain features and events to the
SPECMAP standard 8 18 0 stacked record (Imbrie
et al. 1984; Martinson et al. 1987). This record was
itself dated by comparison to the well-known variations of the orbital parameters of the earth that
vary with a periodicity ofl9, 23, 41, and 100 kyr.
The error in dating single events of this oxygen isotope stratigraphy amounts to ±2.5 kyr on average,
and thus gives very reliable sediment accumulation
rates for the main isotope stages and substages. The
pursuit of an increased time resolution in sediments
with low sedimentation rates may, however, lead
to problems when this method is applied. The reason is that the age difference between subevents
of the stacked 8 18 0 record may come close to the
error of2.5 kyr which may result in large offsets
when the sedimentation rates and sample resolution are low. These uncertainties can result in an
artificial variability of the sedimentation rates (easily
in the range of 50-1 00 %) which then may become
the factor dominating in accumulation rate variability.
The application 0[2 30 Th" to calculate rain rates
was shown to be a valid approach in the area of
the Ceara Rise (western Equatorial Atlantic
Ocean). Six cores yielded fluxes of 23CYfh" very
close to the expected values, giving evidence for
an either low or even negligible sediment redistribution intensity (Francois et al. 1990). The calculation of the rain rates is not significantly affected
by the above described errors, although the decay
correction of the 230Th" concentration is carried out
using the 8180 stratigraphy. However, the half-life
of 23 0Th amounts to 75,200 years and therefore the
effect of a 2.5 kyr error is negligible.