IODP Proceedings    Volume contents     Search

doi:10.2204/iodp.proc.340.108.2013

Physical properties

We used magnetic susceptibility to correlate Hole U1398B with Hole U1398A to 102 mbsf. High bulk densities and magnetic susceptibilities indicate the location of turbidites. Significant (>20%) variability in P-wave velocity exists over short (~5 m) depth intervals at both holes, with higher velocities often associated with more sand-rich sediment. Otherwise, there is no clear trend in P-wave velocity with depth. A linear fit to temperatures measured in the upper 74 m gives 57.4° ± 5.0°C/km.

Stratigraphic correlation between Holes U1398B and U1398A

We used magnetic susceptibility and natural gamma radiation (NGR) to correlate depths between Holes U1398B and U1398A (Figs. F7). We trimmed 10 cm off each end of the core sections in the NGR data and 5.1 cm in the magnetic susceptibility data to ensure minimization of edge effects during correlation. We then cross-referenced our correlations using NGR with magnetic susceptibility as a quality control measure. Hole U1398B was the reference hole for these correlations because it has the longest continuous record. Both holes have poor recovery at depths below 110 mbsf, making it impossible to correlate below this depth. Correlation for the uppermost 7–10 m between holes is good, with clearly matching peaks in magnetic susceptibility throughout this range; however, between 10 and 50 mbsf major discrepancies exist between the two holes. It was especially difficult to find clear correlations between 30 and 50 mbsf. In general, correlations are strongest in the uppermost 10 m and between 60 and 102 mbsf. At all other depths, correlation is poor. Ultimately, we used magnetic susceptibility data to tie all points, with NGR as an additional guide. Below 102 mbsf, correlation is not feasible. Our correlation coefficient using magnetic susceptibility data is 0.46. This value is artificially low because of meter-scale gaps and several core breaks between data sets where no data exist (Analyseries software includes these zones to calculate correlation coefficient). The poorest correlation exists between 30 and 60 mbsf. In this depth range, both physical properties and stratigraphic data sets indicate different sediment depositional histories between sites. We therefore suggest that the poor correlation between 30 and 60 mbsf results from real geologic differences between holes. We were unable to make any correlations below 102 mbsf, and no correlations were made between 10 and 50 mbsf where there was limited recovery and significant stratigraphic differences. Where we did make correlations, depth shifts for Hole U1398B never exceed 2.5 m and rarely exceed 1 m. All picked correlation depth shifts are shown in Table T4.

Gamma ray attenuation density, magnetic susceptibility, and P-wave velocity

Sharp increases in magnetic susceptibility correlate with increases in sediment bulk density. These increases generally occur where coarse-grained sediments such as turbidites exist. P-wave velocities can be broadly grouped by sediment composition: volcaniclastic (1650–1850 m/s) and hemipelagic mud (1500–1600 m/s). Gamma ray attenuation (GRA) density and P-wave velocity commonly follow the behavior of magnetic susceptibility, with values decreasing upward within thick turbidites.

Shear strength

Undrained shear strength (Su) measurements were not performed in the upper 40 and 60 m of Holes U1398A and U1398B, respectively, because of the presence of sandy and gravelly sediment. In Hole U1398A, Su measurements were possible from 40 to 110 mbsf, and in Hole U1398B they were possible from 60 to 165 mbsf. The handheld penetrometer provided higher Su measurements than those obtained with the automated vane shear (AVS) and the fall cone in both holes. Although the set of Su measurements is quite scattered in both holes, particularly those from the handheld penetrometer, we observe a general increase with increasing depth in both holes.

P-wave velocity

Discrete measurements of P-wave velocity measured on the x-axis (PW-X) match whole-core P-wave velocity measurements and are correlated with sediment composition. The lower range, consisting of hemipelagic sediment, is from 1500 to 1600 m/s, and the higher range, consisting of more sandy sediment, is from 1650 to 1890 m/s. P-wave velocities at this site show significant (>20%) scatter and no clear increase in P-wave velocity with depth.

Moisture and density

We collected 56 moisture and density (MAD) measurements (39 from Hole U1398A and 17 from Hole U1398B; Fig. F8). Despite the MAD data gap in Hole U1398A, no MAD samples were taken in Hole U1398B between 12 and 30 mbsf because those cores contained highly disturbed sand. The porosity of 42 hemipelagic samples ranges between 60% and 75%, with three samples having lower porosities between 50% and 60%. Volcanic sand samples have porosities between 39% and 67%. Calculated porosity values in high-permeability volcaniclastic sediment may be inaccurate. Values may be too low because the samples drain water during sampling, including on the core deck, during splitting, and when the MAD sample is extracted from the split core. Alternatively, where core recovery, handling, or splitting processes reorganize sand grains, sediment may become undercompacted and yield anomalously high porosities. Bulk density of hemipelagic sediment has a narrow range between 1.46 and 1.77 g/cm3. Volcaniclastic sediment has bulk densities as high as 2.2 g/cm3, and three samples have values as low as 1.52 g/cm3. Grain density for hemipelagic sediment ranges between 2.6 and 2.8 g/cm3 (with one exception: 2.3 g/cm3). Volcaniclastic sediment has similar grain density.

Thermal conductivity

Thermal conductivity was measured at 51 depths. The mean value was 1.034 W/(m·K) with a standard deviation of 0.086 W/(m·K) and a standard error on the mean of 0.012 W/(m·K).

Downhole temperature measurements

Temperature was measured with the APCT-3 at the bottom of Cores 340-U1398A-6H and 8H (45.6 and 60.4 mbsf, respectively) and Cores 340-U1398B-4H, 8H, and 10H (26.5, 58.1, and 73.8 mbsf, respectively). Measurements attempted for Core 340-U1398B-7H were compromised by fluid circulation. Attempted measurements at the bottom of Core 340-U1398B-3H were not successful because of instrument failure. Downhole temperature was monitored for 674, 670, 673, 208, and 655 s, respectively. Temperature was calculated from these time series of temperature measurements using TP-Fit (see APCT-3 user manual on the Cumulus/Techdoc database at iodp.tamu.edu/​tasapps/). We assume a thermal conductivity (k) of 1.0 W/(m·K) and ρC = 3.7 × 106 J/m3K. To calculate uncertainty, we assume k ranges from 0.9 to 1.1 W/(m·K) and ρC is between 3.2 × 106 and 4.0 × 106 J/m3K. At the base of Cores 340-U1398A-6H and 8H we obtained temperatures of 7.55° ± 0.05°C and 7.97° ± 0.05°C, respectively. At the base of Cores 340-U1398B-4H, 8H, and 10H we obtained temperatures of 6.06° ± 0.02°C, 7.65° ± 0.07°C, and 8.41° ± 0.02°C, respectively. Uncertainties are greater than the error on the best-fit solution and the probe’s measurement accuracy and are dominated by uncertainties in the thermal properties of the sediment. The temperature of ocean water at the seafloor is 4.24°C.

A best-fit linear relationship between depth and our six temperature measurements gives a temperature gradient of 57.4° ± 5.1°C/km (Fig. F9). Using measured thermal conductivity, the implied heat flow is 59 ± 5 mW/m2. This estimate does not need a correction for bathymetry, and sedimentation lowers the measured near-surface heat flow by as much as 4% (Manga et al., 2012). Of all the sites, this is the only one where a quadratic fit to the data (rather than a linear fit) produces a quadratic term that is statistically significant; however, modeling of this data indicates that a linear fit is still favored (Manga et al., 2012). As a consequence, we cannot conclude that the temperature measurements indicate fluid advection.