The sealed core sections from Hole 1017E were split in half lengthwise, described, and sampled at Texas A&M University approximately three months after the completion of ODP Leg 167. Curated depths (meters below seafloor [mbsf]) were corrected to account for gas expansion, cracks, and missing material at core breaks (see Kennett et al., Chap. 21, this volume; Tada et al., Chap. 25, this volume). Two textural data sets were generated from splits of the same samples comprising every 3 cm of Cores 167-1017E-1H, 2H, and 3H. Multiple methods of grain-size analysis are available to researchers. Each has its own advantages or drawbacks in what is actually measured (e.g., particle diameter, volume, settling rate, scattering effects, etc.), the accuracy and precision of analyses, size range analyzed in a single step or multiple iterations, and in the required volume of sample (Singer et al., 1988; McCave and Syvitski, 1991; McCave et al., 1995). The methods employed in our study were largely determined by the instruments available at our institutions; both are previously reviewed, standard methodologies (Singer et al., 1988; McCave and Syvitski, 1991). The first sample set (LB) was analyzed at California State University at Long Beach with a Coulter multisizer using the electrical sensing zone technique (which measures the displaced volume of individual particles). This data set includes nearly all of the 3-cm samples. The second set (TU) was analyzed at Tokyo University with a Horiba LA-920 grain size analyzer by laser diffraction (which measures the angular scattering of suspended particles). This data set is primarily restricted to the intervals 0-224 and 711-889 corrected centimeters below seafloor (cmbsf), in which every other 3-cm increment was analyzed. In both sample sets, sediments were disaggregated, rinsed of sea salt, cleaned of calcium carbonate (hydrochloric acid or acetic acid dissolution) and organic matter (hydrogen peroxide oxidation), and dispersed (sodium hexametaphosphate or sodium pyrophosphate). These steps are required to accurately measure the grain-size distribution of the siliciclastic fraction of the sediment with minimal complications from including materials with different densities, specific surface areas, or adhesive properties. Set LB was sieved to remove any sand-sized particles (>63 µm) before analysis. Set TU was initially treated with a buffered dithionite-citrate solution to remove Fe-Mn oxides (Tada et al., Chap. 25, this volume). Both sample sets retained small amounts of opaline biosiliceous debris (generally <5%).
Statistical analysis of the grain-size distribution data provided median, mode, and mean grain sizes and sorting indicators (e.g., standard deviation) for each sample. These statistics can be derived for the entire measured spectrum (2-63 µm for set LB, 0.02-2000 µm for set TU) or only from some selected subset of that spectrum. We adopt the procedure of McCave et al. (1995) to present statistics from the 10- to 63-µm-size range (the "sortable silt" fraction) in addition to those of the entire measured range. This restricted size range forms the noncohesive fraction of silt; therefore, changes in the mean size and sorting of this sediment are most likely to reflect response to hydrodynamic processes (McCave et al., 1995). Because many sediment samples analyzed display a bimodal grain-size distribution, selected samples were deconvolved by an iterative curve-fitting method to resolve the modes, sorting, and relative proportions of the separate size components that make up the entire distribution. These size components are correlated with chemical and mineralogical components via factor analysis (Tada et al., Chap. 25, this volume; Irino and Pedersen, Chap. 23, this volume).
Data are plotted with
depth (cmbsf) and also with age based on the radiocarbon and oxygen isotope age
model of Kennett et al. (Chap.
21, this volume) from 0 to 58.956 ka (0-1224.2 cmbsf). This paper also
uses a linear extension of this model to 129.84 ka (2487.3 cmbsf), based on
correlation between shipboard measurements of color reflectance (compositional
proxy) and 
values
(near surface temperature proxy; Lyle, Koizumi, Richter, et al., 1997) with the
standard marine oxygen isotope record, placing the bottom of the hole at
approximately the middle of Termination 2 (marine isotope Stage [MIS] datum 6.0;
Martinson et al., 1987).
