Sediment Redistribution, 23"TIlex - Normalization and Implications for the Reconstruction of Particle Flux 411
high Late Quaternary time resolution. When the
accumulation rates of any sedimentary component
from these sediments are used for mass balance
calculations, huge overestimations may be the result. The above described processes can lead to
substantial variations of accumulation rates on very
small spatial scales, e.g. a few hundred meters, as
seen from huge differences of accumulation rates
between twin cores. No sediment-petrographic
tools are available to distinguish between these redistributed sediments and undisturbed sequences,
or to evaluate the amount of laterally supplied particles.
Current driven sediment redistribution has occurred on regional scales in the oceans, as for
example documented by large areas in the Southern Ocean where particles originating from the surface waters were not deposited in the sediments
below (Hays 1967). Synoptic maps also point to the
existence of large areas on the sea floor of the
Southern Ocean where bottom current-controlled
winnowing prevented sedimentation (Heezen and
Hollister 1971; Watkins and Kennett 1972;
Anderson 1990). The existence of contourites in
other very high bottom current enviromnents, such
as for example in the Strait of Madagascar, also
gives evidence for these transport processes. It is
evident that these laterally exported sediments must
have been deposited somewhere else in addition to
the vertical flux. Adjacent to continental slopes the
transport of redeposited particles within the
nepheloid bottom layer may play an important role
(McCave 1983). There is also evidence for sediment redistribution from sediment trap studies conducted in the Southern Ocean where particle fluxes
into deep traps exceeded the values of the shallower
ones by up to a factor of 10 (Abelmann and
Gersonde 1991; van Bennekom et al. 1988).
The application of the 230Th ex -method can serve
as a tool to identify sections of redistributed
sediments and offers the possibility to quantify the
amounts oflaterally supplied material (Bacon and
Rosholt 1982; Bacon 1984; Suman and Bacon 1989;
Francoisetal.1990, 1993; Frank etal. 1996). The
method is based on the uniform distribution of dissolved U in the ocean and thus well-known production rate of 23°Th in the water column. The high
particle reactivity of23"Th causes a very low residence time of only between 5 to 40 years in the
water column (Anderson et al. 1983). Generally,
230Th can therefore not experience a significant basin-to-basin transport within the ocean's water
column (boundary scavenging) in contrast to other,
less particle-reactive, radionuclides such as 231Pa
or lOBe. Thus, the basic idea is that the flux of
230Th ex into the sediments may be assumed to match
its local production rate of2.63 disintegrations per
minute (dpm) / cm 2 kyr· 1000 m water column.
When lateral sediment redistribution has occurred,
the flux of2 3 0Th into the sediments deviates from
ex
the expected value because together with the
redistributed particles the adsorbed 23°Th ex is also
redistributed. No significant desorption or
remobilisation of23% within the sediment has been
observed. In case of sediment focussing, the flux
of 23°Th ex exceeds the value expected from the
production rate in the water column (Fig. I). The
opposite is the case when sediment is removed from
a location (sediment winnowing). The factor of
deviation from the 230Th,x flux expected from production (II' - focussing factor) can be used to reconstruct the "true vertical" rain rates, which should
represent only the settling material not affected by
lateral sediment transport. Equation (1) gives the
rain rate (RR) of any component i in the sediment
(Bacon 1984; Suman and Bacon 1989; Francois et
aI.1993):
RR j = b z f; / initial 23°Th ex
(1)
where J3 is the production rate of 23 °Th in the water column, z is the water depth (km), ~ is the weight
fraction of component i, and initial 23°Th ex is the
concentration of 23°Th ex ' decay corrected to the
time of deposition (note that no dry bulk densities
are required for the rain rate calculation with the
above equation ). The approximately constant flux
of 23°Th also allows the calculation of a model
ex
which yields high resolution sedimentation rates,
and thus also ages, within independently fixed time
markers such as 8 18 0 isotope stage boundaries.
This method was applied, for example, to evaluate
the duration and the deposition rate of ice-rafted
material during Heinrich events HI and H2 in the
high Late Quaternary time resolution. When the
accumulation rates of any sedimentary component
from these sediments are used for mass balance
calculations, huge overestimations may be the result. The above described processes can lead to
substantial variations of accumulation rates on very
small spatial scales, e.g. a few hundred meters, as
seen from huge differences of accumulation rates
between twin cores. No sediment-petrographic
tools are available to distinguish between these redistributed sediments and undisturbed sequences,
or to evaluate the amount of laterally supplied particles.
Current driven sediment redistribution has occurred on regional scales in the oceans, as for
example documented by large areas in the Southern Ocean where particles originating from the surface waters were not deposited in the sediments
below (Hays 1967). Synoptic maps also point to the
existence of large areas on the sea floor of the
Southern Ocean where bottom current-controlled
winnowing prevented sedimentation (Heezen and
Hollister 1971; Watkins and Kennett 1972;
Anderson 1990). The existence of contourites in
other very high bottom current enviromnents, such
as for example in the Strait of Madagascar, also
gives evidence for these transport processes. It is
evident that these laterally exported sediments must
have been deposited somewhere else in addition to
the vertical flux. Adjacent to continental slopes the
transport of redeposited particles within the
nepheloid bottom layer may play an important role
(McCave 1983). There is also evidence for sediment redistribution from sediment trap studies conducted in the Southern Ocean where particle fluxes
into deep traps exceeded the values of the shallower
ones by up to a factor of 10 (Abelmann and
Gersonde 1991; van Bennekom et al. 1988).
The application of the 230Th ex -method can serve
as a tool to identify sections of redistributed
sediments and offers the possibility to quantify the
amounts oflaterally supplied material (Bacon and
Rosholt 1982; Bacon 1984; Suman and Bacon 1989;
Francoisetal.1990, 1993; Frank etal. 1996). The
method is based on the uniform distribution of dissolved U in the ocean and thus well-known production rate of 23°Th in the water column. The high
particle reactivity of23"Th causes a very low residence time of only between 5 to 40 years in the
water column (Anderson et al. 1983). Generally,
230Th can therefore not experience a significant basin-to-basin transport within the ocean's water
column (boundary scavenging) in contrast to other,
less particle-reactive, radionuclides such as 231Pa
or lOBe. Thus, the basic idea is that the flux of
230Th ex into the sediments may be assumed to match
its local production rate of2.63 disintegrations per
minute (dpm) / cm 2 kyr· 1000 m water column.
When lateral sediment redistribution has occurred,
the flux of2 3 0Th into the sediments deviates from
ex
the expected value because together with the
redistributed particles the adsorbed 23°Th ex is also
redistributed. No significant desorption or
remobilisation of23% within the sediment has been
observed. In case of sediment focussing, the flux
of 23°Th ex exceeds the value expected from the
production rate in the water column (Fig. I). The
opposite is the case when sediment is removed from
a location (sediment winnowing). The factor of
deviation from the 230Th,x flux expected from production (II' - focussing factor) can be used to reconstruct the "true vertical" rain rates, which should
represent only the settling material not affected by
lateral sediment transport. Equation (1) gives the
rain rate (RR) of any component i in the sediment
(Bacon 1984; Suman and Bacon 1989; Francois et
aI.1993):
RR j = b z f; / initial 23°Th ex
(1)
where J3 is the production rate of 23 °Th in the water column, z is the water depth (km), ~ is the weight
fraction of component i, and initial 23°Th ex is the
concentration of 23°Th ex ' decay corrected to the
time of deposition (note that no dry bulk densities
are required for the rain rate calculation with the
above equation ). The approximately constant flux
of 23°Th also allows the calculation of a model
ex
which yields high resolution sedimentation rates,
and thus also ages, within independently fixed time
markers such as 8 18 0 isotope stage boundaries.
This method was applied, for example, to evaluate
the duration and the deposition rate of ice-rafted
material during Heinrich events HI and H2 in the