Physical properties were measured on core material recovered during Leg 202 in order to (1) provide data for the correlation of cores among holes at any given site for the construction of complete stratigraphic sequences; (2) detect changes in sediment properties that could be related to lithologic changes, diagenetic features, or consolidation history; (3) provide the dry density records needed for computing mass accumulation rates; (4) identify natural and/or coring-induced discontinuities (e.g., gaps and hiatuses); and (5) provide data to aid interpretation of seismic reflection and downhole geophysical logs.
Magnetic susceptibility (MS), GRA bulk density, compressional wave velocity (VP), and NGR were measured on the whole-core MST. Thermal conductivity (TC) was also measured on whole-round cores. Split-core measurements on the working half of core included VP with the ODP P-wave sensor number 3 (PWS3) and moisture and density (MAD). Color reflectance (CR), magnetic susceptibility point sensor (MSP) measurements, and digital imaging were performed on the archive-half cores.
Magnetic susceptibility was also measured with the OSUS Fast Track, installed during Leg 202 specifically for the fast logging of whole-round core sections immediately after recovery. These measurements aided real-time stratigraphic correlation without being limited by the time constraints of MST measurements.
The shipboard party decided to integrate sedimentological and physical properties observations into the "Lithostratigraphy" sections of the site chapters. This was compatible with and thought to promote the primary objective of the physical properties program, to aid lithologic characterization of sediment sections.
Rapid stratigraphic correlation of cores from adjacent holes and real-time feedback to drillers for recovering complete stratigraphic sections at all sites (see "Stratigraphic Correlation" in "Operational Innovations" in the "Leg 202 Summary" chapter and "Composite Section") were major cruise objectives. The OSUS automated, dedicated track for rapid measurement of magnetic susceptibility was therefore installed during Leg 202 to provide a proxy of sediment variability to the stratigraphic correlators without delay. It uses a Bartington MS2 susceptibility meter (108-mm loop diameter) that was zeroed before a section scan, but no drift correction was applied. The usual sampling interval was 5 cm. Many sediment intervals recovered during Leg 202 had little terrigenous material, and the 1-s integration time setting (~15-min logging time per core) provided insufficient signal resolution. We therefore measured mostly with the 10-s integration time setting, which increased the logging time to ~40 min per core. However, this was still fast enough to keep up with core recovery under most circumstances and significantly faster than the MST logging time, which was set to gather the best possible data with the four sensors as described below.
Magnetic susceptibility and GRA bulk density were measured nondestructively on all whole-round core sections with the MST. NGR was measured on most core sections, but the sensor had to be turned off to save core logging time for some intervals. Compressional wave velocity was measured on most core sections, but the sensor was turned off when acoustic coupling proved insufficient (gassy sediment) or when core disturbance was too pervasive to give a reliable signal.
Sampling intervals and periods were the same for all sensors for any one core in order to optimize MST performance. At some sites, comparison with downhole Formation MicroScanner (FMS) data shows that bulk density variability in the sediment could potentially be resolved at subcentimeter resolution. However, the limits of instrument performance combined with the time constraints from coring operations did not allow sampling at 1 cm or less. Sampling intervals were therefore set at 2.5 or 5 cm, depending on time availability, with most cores measured at 5-cm intervals. These particular intervals are common denominators of the distances between the instruments installed on the MST (30-50 cm) and allow truly simultaneous measurements and therefore optimal use of total measurement times.
Sampling periods varied from 3 to 5 s, depending on total time available, with most cores measured at 5 s per sample. Longer sampling times would have been desirable to improve the NGR and MS signal but were not possible given the rate of core recovery.
The total time availability for MST logging at a site was predicted based on operational time estimate for the site, subsequent transit time, and any other time available before core was on deck at the subsequent site. Sampling parameters were then optimized to use the total available time (e.g., 2.5 cm and 5 s [~2 hr/core]; 5 cm and 5 s [~1.2 hr/core]; 5 cm and 3 s [fast logging, ~0.9 hr/core]).
Two instruments were mounted on the AMST: the Minolta photospectrometer measuring diffuse color reflectance and an MSP. CR was measured at 2.5 cm throughout Leg 202 cores. MSP measurements were taken only at the first two sites (1232 and 1233), when it became clear that the measurement was too slow to be accommodated in the time available. The present AMST configuration requires that CR and MSP runs must be taken one after the other. A measurement system that allows simultaneous measurement of CR and MSP would be needed to make MSP measurements feasible.
Magnetic susceptibility is the degree to which a material can be magnetized in an external magnetic field. It provides information on the magnetic composition of the sediments that often can be related to mineralogical composition (e.g., terrigenous vs. biogenic materials) and/or diagenetic overprinting (e.g., Thompson and Oldfield, 1986). Magnetite and a few other iron oxides with ferromagnetic characteristics have a specific MS several orders of magnitude higher than clay, which has paramagnetic properties. Carbonate, silica, water, and plastics (core liner) have small negative values of MS. Sediments rich in biogenic carbonate and opal therefore have generally low MS, even negative values, if practically no clay or magnetite is present. In such cases, measured values approach the detection limit of MS meters.
MS was measured with the Bartington Instruments MS2C system on both the MST and the OSUS track and with the MS2f on the AMST track in some intervals at Sites 1232 and 1233 (the MS2f is not discussed here). The Bartington system applies a low field (~50 mT) with an inducing field frequency of 0.565 kHz. It provides a low-sensitivity setting of 1.0 (instrument integration time of 1 s) and 0.1 (integration time of 10 s). The MST uses an 8-cm nominal diameter loop (88-mm coil), which yields a Gaussian response function with full width at half maximum (FWHM) value of ~8 cm (Fig. F8). The OSUS system featured a 10-cm nominal diameter loop (108 mm coil) with a FWHM of 9 cm.
The output of the MS meters can be set to centimeter-gram-second (cgs) units or International System (Système International [SI]) units, and the ODP standard is the SI setting. However, to actually obtain the dimensionless SI volume-specific magnetic susceptibility values, the instrument units stored in the ODP database must be multiplied by a correction factor to compensate for instrument scaling and the geometric ratio between core and loop dimensions. For a standard APC core diameter of 66 mm and loop diameters of 88 mm (MST) and 108 mm (OSUS Fast Track), these correction factors are 1.46 × 10-6 and 0.79 × 10-6, respectively, as read from a graph in the Bartington operation manual.
The ratio (~1.86) of these two Bartington compensation factors should represent the factor necessary to convert OSUS Fast Track values to MST values. To test this, we ran a small magnetically susceptible calibration ring through both tracks and calculated the integrals under the response curves (Table T3; Fig. F9). The ratio of the integrals was 1.685. We also compared the instrument values from many core measurements taken on both tracks during Leg 202. The ratios between measured values ranged from 0.5 to 2.0 and averaged ~1.5. The variability in the ratio is thought to be related to the different shapes of the response curves resulting from different loop diameters.
A common operational problem with the Bartington meter is that 1-s measurements are rapid but not precise enough for biogenic-rich sediments, and the 10-s measurements are much more precise but take a prohibitively long time to measure at the desired sampling interval of 2.5 to 5 cm. The MST program was therefore equipped with the option to average any number of 1-s measurements, and we usually averaged five measurements. The OSUS track did not have this option and was mostly run with 10-s integration time.
Bulk density reflects the combined effect of variations in porosity, grain density (dominant mineralogy), and coring disturbance. Porosity is mainly controlled by lithology and texture (e.g., clay, biogenic silica, and carbonate content, and grain size and sorting) and compaction and cementation. In homogeneous pelagic and hemipelagic sediments drilled during Leg 202, bulk density was often a function of the relative amount of calcareous nannofossils and diatoms, which resulted in significant fabric and thus porosity variations as well as grain density variations.
The GRA densitometer consists of a 10-mCi 137Cs capsule as the gamma ray source, with the principal energy peak at 0.662 MeV, and a scintillation detector. The narrow collimated peak is attenuated as it passes through the center of the core. Incident photons are scattered by the electrons of the sediment material by Compton scattering. The attenuation of the incident intensity (I0) is directly related to the electron density in the sediment core of diameter (D), which can be related to bulk density given the average attenuation coefficient (in micrometers) of the sediment (Evans, 1965; Harms and Choquette, 1965). Since the attenuation coefficient is similar for most common minerals and aluminum for practical purposes, bulk density is obtained through direct calibration using aluminum rods of different diameters mounted in a core liner that is filled with distilled water. The GRA densitometer has a spatial resolution of <1 cm (Fig. F8).
P-wave velocity in marine sediments varies with the lithology, porosity or bulk density, state of stress such as lithostatic pressure, fabric or degree of fracturing, degree of consolidation and lithification, occurrence and abundance of free gas and gas hydrate, and other properties. P-wave velocity was measured with two systems during Leg 202, with the MST-mounted P-wave logger (PWL) on whole-round cores (Schultheiss and McPhail, 1989) and with the PWS3 on split cores. The P-wave sensors number 1 and 2 (PWS1 and PWS2), transducer pairs built into pairs of knifelike probes that are inserted into soft sediment, were not used during Leg 202. All ODP P-wave piezoelectric transducers transmit a 500-kHz compressional wave pulse through the core at a repetition rate of 1 kHz.
Traveltime is determined by the software, which automatically picks the arrival of the first wavelet to a precision of 50 ns. It is difficult for an automated routine to pick the first arrival of a potentially weak signal with significant background noise. The search method applied skips the first positive amplitude and finds the second positive amplitude using a detection threshold limit (DTL), typically set to 30% of the maximum amplitude of the signal. Then it finds the preceding zero crossing and subtracts one period to determine the first arrival. To avoid extremely weak signals, minimum signal strength (MSS) can be set (typically to 0.02 V) and weaker signals are ignored. To avoid cross-talk signals at the beginning of the record from the receiver, a delay (typically set to 0.01 ms) can be set to force the amplitude search to begin in the quiet interval preceding the first arrival. In addition, a trigger (typically 4 V) is selected to initiate the arrival search process, and the number of waveforms to be stacked (typically 5) can also be set. Length of the travel path is determined by linear voltage differential transducers.
The P-wave velocity systems require two types of calibration, one for the displacement of the transducers and one for the time offset. For the displacement calibration, five acrylic standards of different thickness are measured and the linear voltage-distance relationship determined using least-squares analyses. For the time offset calibration, room-temperature water in a plastic bag is measured multiple times with different transducer displacements. The inverse of the regression slope is equal to the velocity of sound in water, and the intercept represents the delay in the transducers.
In cases of bad acoustic coupling between the sediment and the liner, the PWL generally does not provide accurate velocity values. The system is, therefore, most useful in undisturbed APC cores, and values become highly questionable when gas is present in the sediment. A correction for the core liner, which is made of Tenite with a sonic velocity of 1987 m/s, is applied routinely. This is accomplished by subtracting the total liner thickness (2.54 mm for split cores and 5.12 mm for whole cores) from the transducer displacement measurement and subtracting the calculated transit time based on the sonic velocity in Tenite from the measured transit time.
Terrigenous sediment is often characterized by NGR from K and Th, which are present mostly in clays but can also originate from heavy minerals or lithic grains. Uranium often dominates the NGR in carbonate-rich sediments with little terrigenous input. Uranium concentration is largely controlled by organic matter flux to the seafloor and the existing redox conditions there. It is also mobile and can migrate to certain layers and diagenetic horizons. In biogenic oozes, such as those drilled during Leg 202, U records may be used as indicators of production and/or preservation of organic matter.
The natural gamma ray system consists of four shielded scintillation counters with 3 in x 3 in doped sodium iodide crystals, arranged at 90° from each other in a plane orthogonal to the core track. The FWHM of the NGR response curve is ~17 cm (Fig. F8), an interval that could be considered a reasonable sampling interval. Measurement precision is a direct function (inverse of square root) of the total counts (N), accumulated for one measurement, according to Poisson's Law of random counting error. N is the product of the intrinsic activity of the material measured and the total measurement time. Therefore, one would always want to maximize the measurement time to improve data quality. Unfortunately, time constraints on core logging during the cruise do not permit long counting times. Furthermore, NGR is measured every 5 cm (same sampling interval as other MST measurements for optimal MST efficiency) and, therefore, only for a short time (typically 5 s). It may be necessary, particularly in low-activity material, to integrate (or smooth) several adjacent measurements to reduce the counting error to an acceptable level. Five-point smoothing is a reasonable data reduction in view of the relatively wide response curve of the sensors. (See Blum, 1997, for more detailed discussions).
TC estimates were needed to calculate downhole temperature from measurements taken with the APC temperature (APCT) tool and will be required for the estimation of heat flow. TC was therefore mainly measured in intervals where downhole temperature measurements were taken, using whole-round cores and the needle probe method in full-space configuration.
We used the Teka TK04 measurement system, which employs the transient linear heat source method with a needle probe that is inserted into the soft sediment (ASTM, 1993; Blackwell, 1954; DeVries and Peck, 1958; Von Herzen and Maxwell, 1959; Vacquier, 1985). The TK04 uses an automated routine to find the conductivity by least-squares fitting to the measured temperature time series. The probes were calibrated with a Macor ceramic standard with a value of 1.61 W/(m·K). Measured values were low in the high-porosity sediments recovered during Leg 202, ranging between 0.8 and 1.2 W/(m·K).
Water (moisture) content, bulk density, grain density, dry density, porosity, and void ratio are solid to interstitial water phase relationships in sediments calculated routinely in the ODP Janus database from the measurement of wet and/or dry mass and wet and/or dry volume (ASTM, 1990), according to equations given in Blum (1997). These phase relationships are useful for general lithologic characterization, calculation of mass accumulation rates, and calibration of GRA bulk density records measured at much higher depth resolution.
Samples of ~10 cm3 were taken from the working-half sections with a piston minicorer and transferred into 10-mL glass vials. Usually, one sample per section was taken at a fixed location (75 cm) in each section coincident with the location of the PWS3 sonic velocity measurements, carbonate and total organic content samples, and smear slide samples. The sampling program was usually completed in the most complete hole at a site, and gaps were filled with measurements from other holes from that site if warranted. Wet and dry masses were determined with twin balances that allow for compensation of the ship's heave gravitational effect on the balance and give a precision better than 1%. The samples were dried in a convection oven at a temperature of 105°-110°C for a period of 24 hr. Dry volume was measured with a helium-displacement Quantachrome Penta-Pycnometer with an uncertainty of <0.02 cm3. Five measurements were averaged for each sample.
Sediment color primarily reflects the composition of the sediment and is affected by some other properties such as moisture, sediment grain size, and split-core surface texture (Balsam et al., 1997). The plastic film put over the sediment to protect the instrument also affects the reflectance spectra collected (Balsam et al., 1997). Quantitative, high-resolution measurement of color has the advantage over visual color description of providing measurements compatible with other physical properties measurements and allowing rigorous data analyses and sediment classification based on color. Some sediment components significantly influence sediment color, such as calcium carbonate representing the relative amount of major biogenic components, iron oxides that may indicate eolian origin, organic matter that may be used as a proxy for paleocean productivity, or chlorite as an indicator of diagenetic alteration. Such components can be (semi)quantitatively recognized in the color spectrum in its derivative values (Deaton and Balsam, 1991; Balsam et al., 1995, 1997, 1998; Balsam and Deaton, 1996; Balsam and Wolhart, 1993; Balsam and Damuth, 2000; Giosan, 2001) or in color parameters calculated from the spectrum, such as L*a*b*, which define a lightness-chroma color space.
During Leg 202, raw reflectance spectra were used to identify absorption features typical for organic pigments and first derivatives of the spectra were used to detect goethite and hematite. Semiquantitative estimates of chlorins and hematite contents were obtained from their characteristic features expressed in the raw spectra and their first derivative, respectively. Both spectra and derivatives were employed in a multiple regression between direct geochemical measurements and reflectance spectroscopy data for estimating carbonate and total organic carbon (TOC) contents. Reflectance input data included 31 raw reflectance values over the visual spectrum (400-700 nm) at a 10-nm interval and their first derivatives with respect to wavelength. Regression equation terms were selected with a stepwise procedure, and they were retained in the equations only if significant above a 95% level. Using this technique in real time during the cruise promotes understanding of the relationship between sediment composition and physical properties and allows detection of variability in carbonate and TOC that cannot be measured rapidly at high resolution.
Diffuse reflectance of visible light was routinely measured using the Minolta CM-2000 spectrophotometer mounted on the AMST. This instrument measures light at visible wavelengths (400-700 nm) in 31 intervals of 10 nm (ASTM, 1985). A black-and-white calibration was performed every day.
Split cores were covered with clear plastic wrap and placed on the AMST. A laser detector mounted on the AMST creates a log of the split-core surface height and ensures gentle and complete contact of the integration sphere with the sediment surface and skipping of empty intervals. Small cracks and surface irregularities may go undetected, however, and the data may contain spurious measurements that should be eliminated from the data set before use.