We built a continuous meters composite depth (mcd) scale and a splice (as defined in "Composite Section" in the "Explanatory Notes" chapter) that range from the top of Core 184-1143B-1H to the bottom of Section 184-1143B-20X-5. The splice and the continuous portion of the mcd scale extend from 0 to 190.85 mcd. Splice construction below this interval was precluded by incomplete core recovery, decreased core quality (e.g., biscuiting), and alignment of core gaps. As described below, however, we were able to construct a discontinuous, or "floating," mcd scale for the cores below 191 mcd (see "Composite Section" in the "Explanatory Notes" chapter).
The mcd scale and the splice are based on the stratigraphic correlation of whole-core multisensor track (MST) and split-core color spectral reflectance (CSR) data (lightness, L*) collected at 2- to 4-cm intervals (see "Physical Properties" for details). From the MST, we used magnetic susceptibility (MS), gamma-ray attenuation (GRA) bulk density, and natural gamma radiation (NGR) data; the P-wave data were of no stratigraphic use. These data, and the splice constructed from them, are presented on the mcd scale in Figures F1, F2, F3, and F4 (also as Synergy Software KaleidaGraph plots and Microsoft Excel data files [see the "Supplementary Materials" contents list]; the spliced records are also available in ASCII format). The depth offsets that comprise the mcd scale are given in Table T3 (also in ASCII format). The splice tie points (Table T4, also in ASCII format) should be used as a guide for detailed postcruise sampling.
Magnetic susceptibility was the most useful stratigraphic tool for correlation. CSR and NGR were helpful in intervals where the MS structure was ambiguous. Gamma-ray attenuation was the least useful because intracore squeezing and stretching produced a signal that was often larger than the natural variability. We constructed the mcd scale by assuming that the uppermost sediment (the "mudline") in Core 184-1143B-1H was the sediment/water interface. This core, the "anchor" in the composite depth scale, is the only one with depths that are the same on the mbsf and mcd scales. From this anchor we worked downhole, correlating the stratigraphy core by core. Based on the correlations, we added a constant to the mbsf depth for each core in each hole in order to align the stratigraphic structure of each core log. We did not squeeze or stretch data within the depth domain. Because we did not make intracore adjustments to depth, we were not able to align every feature in the stratigraphic records. A good example of this occurs between 80 and 90 mcd: MS spikes representing ash layers are perfectly aligned, yet other correlative features may be misaligned (by as much as 50 cm).
Correlation of Cores 184-1143A-13H, 184-1143B-13H, and 184-1143C-13H (~115-130 mcd) was particularly challenging. We chose to match structure within the MS and CSR data. Alternative correlations are possible. In the splice, we chose to append successive cores from Hole 1143C at two locations (49.84 and 89.94 mcd) instead of incorporating short intervals from Hole 1143B (Table T4). Correlative structure in Holes 1143A and 1143B demonstrates that no significant amount of sedimentary section is missing between the appended cores from Hole 1143C.
Below 191 mcd, core quality and recovery decrease and turbidite frequency increases (see "Lithostratigraphy"). Although the cores in this interval could not be tied directly to the composite depth scale (and thus the splice), they could be correlated with each other. Here we placed the cores in a relative, or discontinuous, composite depth scale, which we refer to as the "floating" mcd (as defined in "Composite Section" in the "Explanatory Notes" chapter). This scale is not tied to the overlying mcd scale (and thus back to the mudline). Instead, the positions of the cores in Holes 1143A and 1143B below 191 mcd are adjusted such that correlative features match those in Hole 1143C, whose depths retain the offset accumulated at the bottom of the splice. Correlation below 191 mcd is less certain than in the overlying interval, except for the ties made between MS events associated with turbidite layers (Fig. F1).
In several cases, we noted stratigraphic distortion among correlative cores from adjacent holes. We attribute much of this to the normal coring process, although we cannot discount the possibility that some may reflect true differences among the stratigraphic sections. An example of the stratigraphic distortion is the low GRA values at the tops of cores, indicating stretching relative to the bottoms (Fig. F3).