Core recovery from a single hole is generally insufficient to obtain a complete geologic section. To maximize recovery of complete geologic sections during Leg 182, multiple holes were drilled at each site, where possible, with cores offset in depth by ~2-3 m. The offset depths and multiple holes ensured that most intervals missing within a single APC hole were recovered in an adjacent hole. The degree of continuity of the recovered section at each site was assessed by development of composite depth sections using the Splicer software, following the general methodology first used during Leg 138 (Hagelberg et al., 1992). Similar methods were used during Legs 154 (Curry, Shackleton, Richter, et al., 1995), 162 (Jansen, Raymo, Blum, et al., 1996), 167 (Lyle, Koizumi, Richter, et al., 1997), and 177 (Gersonde, Hodell, Blum, et al., 1999). This section describes the methods used to produce composite and spliced sections during Leg 182, using the Splicer software.
Cores recovered from each site were cut into 150-cm whole-round sections and allowed to equilibrate to ~20°C. Each core section was then analyzed for physical properties on a MST (see "Physical Properties"). Split cores were analyzed using a color spectral reflectance scanner (see "Lithostratigraphy"). Biostratigraphic data (first-appearance datum and last-appearance datum) from cores were also incorporated into the database for composite section development (see "Biostratigraphy").
The physical properties and biostratigraphic data described above were imported into the Splicer (version 2.0) software package. Splicer is an interactive, Unix-based software package designed by Peter DeMenocal and Ann Esmay at Lamont-Doherty Earth Observatory (LDEO) specifically for ODP stratigraphic correlation purposes. During composite section construction, data were culled to avoid the use of anomalous values resulting from voids and disturbed intervals in the cores. Natural gamma-ray emissions and color spectral reflectance data in the 400- and 700-nm bands provided the best lithologic parameters for correlation. Biostratigraphic data provided additional datums for correlation purposes, especially where correlations based on physical properties data were ambiguous.
The Splicer software allows for direct graphical and statistical comparison of data from each hole (Fig. F8). Tie lines are drawn between apparently correlative features present in the data (data excursions, peaks, troughs, and plateaus). The software provides a statistical analysis of the correlation over an adjustable depth range (typically ± 2 m). By convention, ODP sample and core depths were recorded as mbsf beginning at the top of the first core of a hole. Correlated features were then depth aligned by linearly adjusting the data and ODP coring depths downward on a core-by-core basis. Correlations and alignments were continued downward for each core in each hole, using data from one hole to fill in data gaps in another hole. No depth adjustments (stretching or squeezing) were made to the data within a core. Correlation of events, involving alignment of data present in multiple holes, provided verification of the extent of recovery of the sedimentary section. Utilization of at least two different physical properties allowed hole-to-hole correlations to be made with greater confidence than would be possible with only a single parameter. Chronostratigraphic correlations were continuously verified using biostratigraphic information. The resulting adjusted depth scale is called the mcd, and the section produced by the aligned cores is termed a composite depth section. All adjustments to the data are written to a data output file (Table T5). The offset column allows conversion of sample depths from mbsf to mcd, effectively creating a sampling strategy guide. The mcd depth for any point within a core equals the mbsf depth plus the offset. A table is presented in each site chapter summarizing core offsets for conversion from mbsf to mcd scales.
Core distortions are typically caused by the drilling process. The offsets between mbsf and mcd scales result both from random uncertainty in mbsf depths as a result of ship motion and heave and from differential distortion of cores during the coring and retrieval process. For example, core expansion typically occurs at the tops of cores, whereas the sediment at the base of a core may be compressed. The expansion causes the composite section depths (mcd) to be greater than the mbsf depths, typically by ~10%. Because cores are offset in depth between holes, distortion of any specific mbsf interval is different in different holes. As a result, it is not possible to precisely align all features between holes. Tie points were chosen in an attempt to optimize the alignment and correlation of features in multiple holes. Where possible, tie points were chosen in the middle to lower portion of cores, where the record is likely to have been least disturbed by expansion. Where correlations were uncertain based on physical properties, input from biostratigraphic data was sought. Core photographs and VCDs were also a useful reference source for identifying potentially correlative lithologic features within cores. Where overlapping data from other holes were unavailable, causing data gaps in the total section, the depth adjustment applied was the cumulative offset from the overlying aligned cores.
The Splicer software allows the user to merge, or splice, the best data from the composite section to produce a single spliced record representing the complete geologic section at each site (Fig. F8). The spliced record is constructed by patching the intervals missing in a single hole with data from adjacent holes. This process provides a single representative record of the physical properties parameters (e.g., MS, spectral reflectance, or gamma-ray attenuation [GRA] bulk density) for the entire section, which is ideally suited to guide core sampling for high-resolution paleoceanographic studies.
Splice tie points were made between adjacent holes where visually obvious features are strongly correlated. The splice operation is depth constrained so that no further core offset is possible. Because of core expansion and/or compression, the total length of the spliced record depends on which intervals of core were selected to construct it. Each splice was constructed by beginning at the mud line at the top of the composite section and working downward. Core intervals were chosen for the splice in an effort to minimize the inclusion of disturbed data. As in the composite section construction, no compression or expansion of the data are possible. Adjustments to the composite or spliced sections, such as a linear compression of the mcd scale within individual core intervals, are required to align all features exactly (e.g., Hagelberg et al., 1995).