It is commonly recognized that removal of cores from the borehole is accompanied by significant changes in the sediment physical properties. Four factors identified as being responsible for differences between laboratory measurements and in situ properties include (1) temperature change, (2) decrease in hydrostatic pressure, (3) decrease in sediment rigidity, and (4) mechanical rebound resulting in an increase in porosity (Hamilton, 1965, 1976; Mayer et al., 1985; Urmos and Wilkens, 1993).
The rebound experienced by the sediments at Sites 1218 and 1219 was estimated by subtracting the GRA density from the HLDT density. In order to minimize errors associated with the possible mismatch of the two depth profiles, the density difference values were averaged over 5-m windows. The average values for both sites are plotted together in Figure F11 against the effective overburden pressure. Pressure was calculated from the core densities according to the procedure of Busch (1989) and was used instead of depth because of the substantial differences in density among the different lithologies. In addition to the density rebound estimates for Sites 1218 and 1219, curves for predicted density rebound of pelagic clay, radiolarian ooze, and calcareous ooze, derived from the relationships of Hamilton (1976), are shown in Figure F11. The density rebound for the one interval of clay from Site 1219 plots on the predicted pelagic clay rebound line. The density difference for the radiolarian ooze from Sites 1218 and 1219 ranges between ~0.05 and 0.10 g/cm3. This difference represents a greater expansion than the 0.03–0.04 g/cm3 predicted from Hamilton (1976) for the range of pressures in which the radiolarian oozes were recovered at Sites 1218 and 1219. The estimated density rebound for the nannofossil ooze and chalk displays no consistent pattern and does not vary as a function of pressure or depth. Over much of the intervals of nannofossil ooze logged at Sites 1218 and 1219, the GRA density from the cores is greater than the HLDT density (Fig. F11). The possible influence of an enlarged borehole on the unexpectedly low HLDT densities was investigated by using the caliper logs to selectively filter the data shown on the crossplot of GRA density and HLDT density for progressively narrower borehole diameters. This filtering did not reveal a dependence of the density patterns on borehole diameter.
Comparison of laboratory and logging density measurements at Sites 1218 and 1219 does not provide a clear path to follow to use the laboratory-determined densities to estimate in situ values. For clays, radiolarian oozes, and radiolarites, the range in values on the density rebound vs. pressure crossplot is insufficient to establish a trend to predict the rebound of samples from the depth that they were extracted. As a conservative estimate, the relationships developed by Hamilton (1976) for estimating porosity rebound were used to derive the in situ density of the clays and siliceous sediments. The calcareous sediments from the logged intervals of Sites 1218 and 1219 pose a different problem in estimating the amount of mechanical rebound. There is a greater range in density rebound values, but they do not follow the prediction from the experiments of Hamilton (1976) or any other trend. In an investigation of pelagic carbonates from the Ontong Java Plateau, Urmos and Wilkens (1993) concluded that mechanical rebound did not occur in the carbonate sediments they were studying. Based on this result and the lack of a relationship between density rebound and effective overburden pressure for the nannofossil oozes and chalks at Sites 1218 and 1219, a rebound correction was not applied to estimate the in situ density of the calcareous sediment.
Estimates of in situ wet bulk density and velocity are based on the discrete sample data sets. These data were used because of (1) the unambiguous depth correspondence of the density and velocity samples, (2) the availability of data from XCB cores, and (3) the assumed better quality of the discrete sample data.
The estimated in situ wet bulk density at the Leg 199 sites is shown in Figure F12 as a composite depth profile combining data from all of the sites. The three principal lithologies, clay, radiolarian ooze, and nannofossil ooze, separate into distinct groupings on the composite depth profile. Pelagic clay, which most commonly occurs in the uppermost 50 m of the sediment columns, displays a moderately wide range in wet bulk density without a distinct downhole trend. The wet bulk density of the radiolarian ooze varies over a narrow range and increases slightly with depth. The nannofossil ooze densities are distinctly higher than those of the radiolarian ooze and more variable without a significant downhole trend.
Estimated porosity, or density, rebound has been used to estimate the effect of mechanical rebound on velocity through using an established velocity-porosity relationship for laboratory measurements (Boyce, 1976; Shipley, 1983; Mayer et al., 1985). In questioning the extent of porosity rebound in pelagic carbonates, Urmos and Wilkens (1993) questioned the appropriateness of using the laboratory measurement extrapolation to estimate the rebound effect on velocity. Because of this doubt and because there is not a well-defined relationship between density and velocity for the Leg 199 sediments (Fig. F6), a mechanical rebound correction was not applied to the velocity data.
A correction to account for the effects of the change in temperature and pressure was applied to the laboratory velocity measurements using the approach of Boyce (1976). The pressure in the borehole was assumed to be hydrostatic and was calculated using a seawater density of 1.04 g/cm3. Results from heat flow measurements at Sites 1218, 1219, and 1220 (Lyle, Wilson, Janecek, et al., 2002) were used to establish an average temperature profile that was applied to all of the Leg 199 sites. The effect of temperature and pressure on the fluid phase of the sediment was estimated using relationships published by Wilson (1960). Following Boyce (1976), it was assumed that temperature and pressure had negligible effect on the velocity of the solid phase of the sediment. Published values of the dry velocity of clay, calcite, and opal were used to calculate the velocity change of the sediment with depth.
The laboratory velocities corrected to estimate the in situ velocity are plotted as a composite depth profile of velocity for the eight Leg 199 sites (Fig. F13). The application of the temperature and pressure correction results in a velocity increase of ~1% near the seafloor to ~5% at 300 meters composite depth (mcd). The clays above 50 mcd display the greatest variability in velocity. The radiolarian ooze and radiolarites are characterized by velocities distinctly higher than those of the nannofossil ooze and chalk to a depth of 225 mcd. Below ~225 mcd, velocities of the siliceous and calcareous sediments overlap as a result of the increase in the velocity of the nannofossil ooze and the increased variability of the radiolarian ooze velocity.
The estimated in situ wet bulk density and velocity were used to calculate acoustic impedance. Crossplots of impedance and wet bulk density (Fig. F14) and impedance and velocity (Fig. F15) indicate that variation in wet bulk density is primarily responsible for differences in impedance. Greater variability in velocity and the lack of consistent velocity trends are largely responsible for the lack of a relationship between velocity and impedance. In both Figure F14 and Figure F15, the radiolarian ooze occupies distinct fields as a result of its overall low density and relatively high velocity.