COMPOSITE
DEPTHS
Introduction
Construction
of the composite section from Holes 1126B and 1126C
indicates that a nearly complete record of the upper
Miocene-Pleistocene sedimentary section was recovered at
Site 1126. Comparison of overlap between cores from adjacent
holes was used to establish the degree of section
continuity. Using the Splicer software, sedimentary features
and physical properties present in adjacent holes were
aligned so that they occur at approximately the same depth.
This alignment was performed downward from the mudline, and
an offset value was added to the mbsf depth of each core to
create a meters composite depth (mcd) for each core. Table T3
(also in ASCII
format) relates the mbsf depths to mcd depths
for each core and section at Holes 1126A, 1126B, and 1126C
and provides offset values for the conversion of mbsf to mcd.
Data
Input
The
primary lithologic parameters used to create the composite
section were color reflectance data (400 nm) measured on
split cores, natural gamma radiation (NGR) emissions data
collected by the multisensor track (MST) on whole-round
cores, and gamma-ray attenuation (GRA) wet bulk density data
also acquired by the MST (Fig. F11).
For specifics regarding data collection procedures and
parameters, see "Physical
Properties". Magnetic susceptibility data,
normally quite useful for correlation, were used sparingly
because of very low values (range of -3 to 1) resulting in a
very low signal to noise ratio. The MS signal was dominated
by an unidentified nonrandom noise that produced a
pronounced positive excursion at the top of each section and
then diminished downsection. The P-wave velocity
data collected by the MST were of poor quality as a result
of numerous voids in the cores and they were not used for
correlation purposes. The GRA bulk density estimates were
useful for correlation but were found to vary in response to
core disturbance. Natural gamma radiation data were useful
for correlations but were acquired at lower resolution than
other data (16 cm vs. 4 cm for GRA data). After assessing
all the data it appeared that color reflectance data,
collected at 4-cm intervals, provided the best tool for
identifying correlative features between cores.
Biostratigraphic
data aided in correlations by providing additional datums
that were compared between holes (Table T4)
(see "Biostratigraphy").
Planktonic foraminifers were particularly useful in
constraining correlative intervals of the recovered section
and corroborating ties based on physical properties. Because
most biostratigraphic samples were taken from core catchers,
the stratigraphic error is generally on the order of 10 m
(the distance between core catchers in consecutive cores).
Composite
Section Construction
Correlations
between cores were hindered by significant differential core
distortion, particularly at the very top of each core where
a much-expanded record was sometimes indicated. This core
distortion allows similar data characteristics to be
identified between holes, but it results in an
unsatisfactory statistical comparison. Insufficient overlap
was also a problem for statistical comparison of the
records. Correlations were further hindered by poor core
recovery below 100 mbsf, resulting from the presence of
multiple thick chert layers interbedded with easily
deformable hemipelagic ooze (see "Lithostratigraphy").
The
Pleistocene and upper Pliocene records are easily correlated
between holes. The Pliocene/Pleistocene boundary occurs at
~40 mcd, based upon biostratigraphic data (Fig. F12).
The Pleistocene record is characterized by high-amplitude
(40 percentage units) oscillations in color reflectance over
depth intervals of ~2-4 m. The upper Pliocene section is
characterized by lower amplitude (10-20 percentage units)
and thinner oscillations over depth intervals of ~0.5 m.
Data peaks have very similar characteristics between cores,
facilitating easy correlation down to ~60 mcd. An exception
to this occurs in Cores 182-1126B-4H and 182-1126C-4H, which
exhibit significant distortion and disruption of the record
by slumping. Below 60 mcd large data gaps and the presence
of numerous slumps made correlations difficult and highly
tentative.
Color
reflectance resolves the finer lithologic structure
corresponding to cyclic sedimentation patterns on a scale of
tens of centimeters or greater. Throughout much of the
Pleistocene and Pliocene section the low reflectance
corresponds to glauconitic wackestones at the base of a
cycle, which grade upsection into nannofossil/foraminiferal
oozes with higher reflectance. Depositional sequences are
bounded by abrupt burrowed contacts. The record of NGR
reveals features with very similar characteristics to the
color reflectance data over scales of 1 m or more. Overall,
there is an inverse correlation between color reflectance
data and NGR data (Fig. F12),
indicating higher radioactive emissions in the darker, more
clayey wackestone layers (see "Physical
Properties"). The similarity in patterns
between the records was useful for correlation purposes.
The
composite section indicates good agreement between the
lithostratigraphic records in Holes 1126B and 1126C, with
several exceptions. From the mudline to ~30 mcd
(mid-Pleistocene sediments) the records are nearly identical
(Fig. F11).
Between 30 and 40 mcd the large-scale structure is similar,
but the interval in Hole 1126C appears compressed with
respect to Hole 1126B based on color reflectance and NGR
data, resulting in offsets in the records. Correlations
using NGR provide the highest degree of confidence in this
particular interval. There is also some extreme sediment
disturbance (fluidization of sediments) in the top section
of Core 182-1126C-4H that destroyed the record from 26.9 to
28.4 mcd (25.5-27.0 mbsf in Hole 1126C). Coherence between
the data records is attained again at 40 mcd and is
maintained to 59 mcd. From 59 to ~62 mcd the record in both
holes is greatly disturbed by slumping. Intermittent slump
deposits continue downward to ~118 mcd in Holes 1126B and
1126C (~116 mbsf in both holes), making correlations between
records tenuous and intermittent. Below 110 mcd core
recovery is significantly lower and correlations continue to
be difficult and intermittent. Because a single offset value
is assigned to each core, individual features within a core
cannot be stretched or compressed using the Splicer
software. As a result, not all features in adjacent cores
are aligned during composite section construction. Further
processing of the data files is necessary to achieve an
improved alignment.
Construction
of the composite and spliced sections indicates that
recovery of the Pleistocene section was nearly complete
except for a minor gap at ~28 mcd (Table T5,
also in ASCII
format; Figs. F11,
F12).
There may in fact be several centimeters of overlap at 28
mcd; however, there is not enough overlap to provide any
confidence in a tie at this level. The Pliocene section is
nearly complete except for minor gaps in the record at ~71
and 81 mcd. Biostratigraphic data indicate that the gap at
81 mcd must be on the order of centimeters. Recovery in the
upper Miocene section and downward was not as complete, with
a minor gap occurring at 93 mcd and a major gap occurring at
138-144 mcd. The actual splice is therefore incomplete below
28 mcd. Splicing was not continued below 100 mcd because of
core distortion and poor recovery. The mcd scale expansion
relative to the mbsf scale is ~10% from 0 to 100 mcd. Below
100 mcd the offsets become more variable as a result of the
core distortion and poor recovery.
The
Splicer software allows the user to tune the core data to an
age series. An initial attempt to tune the color reflectance
data to a composite seawater 
18O
curve (Raymo et al., 1989; Shackleton et al., 1990;
Shackleton et al., 1995) produced positive results (Fig. F13).
This method assumes that facies successions within the cores
are a function of eustasy and the 18O
curve is an adequate proxy for eustasy. An initial age model
was produced using the biostratigraphic datums (Table T4).
Tuning was then accomplished by establishing ties between
the core data and the time series data using the
biostratigraphic datums as constraints. The Splicer software
provides a statistical evaluation of the proposed tie. Upon
correlation the age of the tie is applied to the age model
for the core data. The tuned record reveals a good
correspondence between color reflectance oscillations and
glacial-interglacial stages at least back to 1 Ma (Fig. F13).
The correlation suggests that the high-reflectance,
nannofossil/foraminiferal oozes are deposited during
sea-level highstands. Beyond 1 Ma the lower amplitude and
higher frequency of the 18O
record make correlations difficult. This preliminary
investigation suggests that Site 1126 may yield a nearly
complete 18O
record for at least the Pleistocene.
