The recovery of continuous sections from APC-cored intervals of the sedimentary sequence was crucial to the high-resolution, paleoceanographic objectives of Leg 178. Drilling multiple APC holes offset in depth at each site has traditionally helped to ensure that intervals of no recovery in a single APC hole were recovered in an adjacent hole. During Leg 178, as on several previous ODP legs, the continuity of recovery was tested by the development of composite depth sections and splices at the multiple-cored sites (Sites 1095, 1096, 1098, and 1099). The method used during Leg 178, and briefly summarized below, was similar to that used to construct composite depth sections during Leg 138 (Hagelberg et al., 1992), Leg 154 (Shipboard Scientific Party, 1995), Leg 162 (Shipboard Scientific Party, 1996), and Leg 167 (Shipboard Scientific Party, 1997b). At each site, high-resolution (2 cm) measurements of magnetic susceptibility, GRAPE wet bulk density, P-wave velocity, and standard measurements (15 cm) of natural gamma ray were made on whole-core sections using the MST soon after the core had equilibrated to laboratory temperature. These measurements were entered into the shipboard database. Additionally, measurements of color reflectance in the 0- to 700-nm band were made at 5- and 2-cm resolution on the split cores (see "Lithostratigraphy") using a Minolta color scanner. These parameters were used to correlate between the different holes at each site and to construct a composite section and spliced record.

Adjustments to the shipboard mbsf depth scale are required for several reasons (cf. Ruddiman et al., 1987; Farrell and Janecek, 1991; and Hagelberg et al., 1995). Elastic response of the sediment to reduced pressure accompanying core recovery causes the cored sediment sequence to be expanded relative to the drilled interval. In addition, other factors, including tides, random variations in ship motion, and heave, can affect the true in situ depth of each core by introducing errors. Even between successive cores having nominally 100% recovery or greater, portions of the sediment sequence are usually missing. As a result, the composite depth scale increases downhole relative to the mbsf scale, typically on the order of 10%. To span the gaps between cores, multiple holes are required, and the core breaks must be staggered so that material missing at one core break is available in another hole.

Hole-to-hole correlation was made using interactive Unix platform software (Splicer) developed by the Borehole Research Group of Lamont-Doherty Earth Observatory (LDEO) and patterned after the Leg 138 correlation software. Initially, all data sets listed above for each hole were visually examined to check the consistency of data. One near-continuous data set, with visually distinctive and high-amplitude variations, was chosen for primary use in correlation. This data set, typically magnetic susceptibility or reflectance, was plotted on the screen for all holes at the site simultaneously. Subsequently, the core that appeared to have the best recovery at the sediment/water interface was chosen as the top core of the composite section.

By visually comparing the data of the working core to data of the uppermost fixed core in one of the other holes, distinctive patterns could be observed in both cores. Recognized features were tied and correlated using a cross-correlation function. The depth of the working core was then adjusted upward or downward relative to the depth of the fixed core. This process was repeated for the uppermost core from other holes. Moving back and forth between all holes (commonly three), the sedimentary sections were correlated between holes and the depth scale was adjusted. The resulting composite depth scale (meters composite depth [mcd]) was then added as an additional column to the original data files. Depths are not adjusted within each core, thus no stretching or compression is allowed by the available software in any individual cores.

After creating a preliminary composite depth, this composite depth scale was applied to other parameters. The process as described above was then repeated. The position of each core was examined, and minor adjustments were made where necessary.

The composite depths for each site are presented in tabular form in the composite depth section of each site chapter (see, for example, Table T36, in the "Site 1095" chapter). The mcd tables include the original (mbsf) depth of the top of each core, the composite depths (mcd), and the offset between the two depth scales, which had to be used to convert from the mbsf to the mcd scale. The current version of the scale is considered to be preliminary owing to time limitations on the ship and to the evolving nature of age data (magnetostratigraphy and biostratigraphy), which were not incorporated into the current scale. Therefore, the mcd scale has not been incorporated into the JANUS database. Instead, we intend to produce a revised mcd scale, incorporating also the use of the proposed "Sagan" software, for publication in the Leg 178 Scientific Results volume, with the goal of incorporating a revised scale into JANUS. The offset tables are presented within the "Composite Depths" sections of the site chapters, should the current scale need to be used.

Finally, a single spliced record of each parameter from all holes was assembled using the best-correlated tie points, to select segments of the composite section to be inserted into the final spliced section. The splice was extended downward by moving from hole to hole and bridging the gaps in the sedimentary sequence at the core breaks. Each splice was designed to provide a continuous record of sedimentary sequence for a site. The tie points used to splice cores are always given as a table in each site chapter (e.g., Table T37, in the "Site 1095" chapter). An example of a composite section is illustrated in Figure F14. In column A, the magnetic susceptibility from two holes at Site 1096 is shown on the mbsf depth scale. In column B, the same record is shown after composition to mcd scale, so that correlative features are aligned. Where the amount of offset necessary to align features was ambiguous or imprecise for all lithologic parameters, or where multiple-hole data were unavailable, no composite depth adjustments were made. In these cases, the total amount of offset between mbsf depth and mcd equals the cumulative offset from the overlying core parts. The composite depth section extends only to the base of multiple-cored intervals. Below the multiple-cored intervals, core depths were appended using the offset of the last core in the composite domain.