Using Rock Magnetic Proxy Records for Orbital Tuning and Extended Time Series Analyses
615
bioturbation (Berger and Heath 1968) and an increasing error of astronomical calculations with age
(Berger and Loutre 1992).
Specific Aspects of Tuning Rock Magnetic
Records
Orbital tuning of proxy records implies the assumption that a parameter is almost exclusively responding to primary orbital forcing and that the response
function is not misleading the correlation of proxy
and target signals. Saturation effects, competing
anticorrelated orbital response mechanisms and
secondary overprinting will bias the tuning approach and eventually make it pointless. It may be
possible to master such problems with specifically
designed target curves (e.g., Berger and Jansen
1994a), but this strategy requires a deep understanding of all processes involved. Obviously there
is a need for simpler criteria defining which
sediments are appropriate for orbital tuning. A few
general rules for rock magnetic age modelling in
various depositional environments can be outlined.
Of all rock magnetic records those from subor anoxic sediments are the most precarious as they
are usually affected by magnetite reduction. The
second example in the paper by Frederichs et al.
(this volume) illustrates, how multi-parametric
magnetic methods can be used to reconstruct a
susceptibility record which has been diagenetically
overprinted during temporary suboxic conditions.
An abrupt coarsening or fining ofthe magnetic
mineral fraction as indicated by magnetogranulometric proxies such as M,/M;r (ibid.) reveals
changes in the sedimentation modus, that probably
also mask paleoceanographic informations of other
proxy parameters. Whatever may cause this disturbance - erosion, winnowing, reductive diagenesis,
slumping or intercalated turbidite or tephra layers -
the respective sections should be carefully studied
and eventually dissected. The remaining record can
still be reassembled and tuned, if a pattern correlation with undisturbed records permits a convincing quantification of resulting age gaps.
We found the simultaneous orbital tuning of 520 regionally related cores much more conclusive
than a treatment of individual records, because it
gives a better impression of pattern variability in
problematic sections and clues for correlation uncertainties. Parallel core processing is supported by
the speed of experimental data acquisition in environmental magnetism.
Partial or total carbonate dissolution does not
affect the stability of the magnetic mineral fraction.
However, without carbonate or opal sedimentation
(pelagic clay) the most important mechanism of
susceptibility modulation, the mutual dilution of
terrigenous and biogenic sediment fractions (Robinson 1990; Diester-Haass 1991; Mienert and Chi
1995) is missing. Unless orbital variations induce
compositional changes within the lithogenic fraction, concentration-dependent magnetic parameters
will merely reflect consolidation with depth. In this
as well as other cases of purely terrigenous sedimentation, it is advisable to base orbital tuning upon
magnetomineralogic (e.g., hematite-magnetite ratio) or -granulometric parameters (e.g., S03T or
M,/M), These records often carry a climatic signal related to varying source regions or transport
(oscillating eolian or glaciomarine fluxes, bottom
current activity).
In oceanic regions with very low, irregular or
high sedimentation rates, rock magnetic records deviate more or less severely from standard patterns
(as do other types of records). A promising age
modelling strategy is to start out with establishing
susceptibility-based core-to-core correlation frameworks. When all inconsistencies in this multiple
correlation scheme are resolved, the most convincing and complete record is dated by the best available method and may then serve as initial age model
for subsequent tuning of the entire collection.
In principle, magnetostratigraphy provides ideal
initial age models for orbital tuning of rock magnetic parameters. Paleo- and rock magnetic investigations can share the same sample set which is
advantageous for cross-validation. At least one
core reaching Matuyama age within a set of
magnetically paralleled cores is needed to assign a
rough age frame to all others. Although numerous
short inverse polarity events ofthe Brunhes Chron
« 0.78 Ma) have been detected (Nowaczyk et al.
1994) their chronostratigraphic use is severely hampered by lithology-dependent lock-in effects (Bleil
and von Dobeneck, this volume).
615
bioturbation (Berger and Heath 1968) and an increasing error of astronomical calculations with age
(Berger and Loutre 1992).
Specific Aspects of Tuning Rock Magnetic
Records
Orbital tuning of proxy records implies the assumption that a parameter is almost exclusively responding to primary orbital forcing and that the response
function is not misleading the correlation of proxy
and target signals. Saturation effects, competing
anticorrelated orbital response mechanisms and
secondary overprinting will bias the tuning approach and eventually make it pointless. It may be
possible to master such problems with specifically
designed target curves (e.g., Berger and Jansen
1994a), but this strategy requires a deep understanding of all processes involved. Obviously there
is a need for simpler criteria defining which
sediments are appropriate for orbital tuning. A few
general rules for rock magnetic age modelling in
various depositional environments can be outlined.
Of all rock magnetic records those from subor anoxic sediments are the most precarious as they
are usually affected by magnetite reduction. The
second example in the paper by Frederichs et al.
(this volume) illustrates, how multi-parametric
magnetic methods can be used to reconstruct a
susceptibility record which has been diagenetically
overprinted during temporary suboxic conditions.
An abrupt coarsening or fining ofthe magnetic
mineral fraction as indicated by magnetogranulometric proxies such as M,/M;r (ibid.) reveals
changes in the sedimentation modus, that probably
also mask paleoceanographic informations of other
proxy parameters. Whatever may cause this disturbance - erosion, winnowing, reductive diagenesis,
slumping or intercalated turbidite or tephra layers -
the respective sections should be carefully studied
and eventually dissected. The remaining record can
still be reassembled and tuned, if a pattern correlation with undisturbed records permits a convincing quantification of resulting age gaps.
We found the simultaneous orbital tuning of 520 regionally related cores much more conclusive
than a treatment of individual records, because it
gives a better impression of pattern variability in
problematic sections and clues for correlation uncertainties. Parallel core processing is supported by
the speed of experimental data acquisition in environmental magnetism.
Partial or total carbonate dissolution does not
affect the stability of the magnetic mineral fraction.
However, without carbonate or opal sedimentation
(pelagic clay) the most important mechanism of
susceptibility modulation, the mutual dilution of
terrigenous and biogenic sediment fractions (Robinson 1990; Diester-Haass 1991; Mienert and Chi
1995) is missing. Unless orbital variations induce
compositional changes within the lithogenic fraction, concentration-dependent magnetic parameters
will merely reflect consolidation with depth. In this
as well as other cases of purely terrigenous sedimentation, it is advisable to base orbital tuning upon
magnetomineralogic (e.g., hematite-magnetite ratio) or -granulometric parameters (e.g., S03T or
M,/M), These records often carry a climatic signal related to varying source regions or transport
(oscillating eolian or glaciomarine fluxes, bottom
current activity).
In oceanic regions with very low, irregular or
high sedimentation rates, rock magnetic records deviate more or less severely from standard patterns
(as do other types of records). A promising age
modelling strategy is to start out with establishing
susceptibility-based core-to-core correlation frameworks. When all inconsistencies in this multiple
correlation scheme are resolved, the most convincing and complete record is dated by the best available method and may then serve as initial age model
for subsequent tuning of the entire collection.
In principle, magnetostratigraphy provides ideal
initial age models for orbital tuning of rock magnetic parameters. Paleo- and rock magnetic investigations can share the same sample set which is
advantageous for cross-validation. At least one
core reaching Matuyama age within a set of
magnetically paralleled cores is needed to assign a
rough age frame to all others. Although numerous
short inverse polarity events ofthe Brunhes Chron
« 0.78 Ma) have been detected (Nowaczyk et al.
1994) their chronostratigraphic use is severely hampered by lithology-dependent lock-in effects (Bleil
and von Dobeneck, this volume).