IODP Proceedings    Volume contents     Search

doi:10.2204/iodp.proc.339.105.2013

Physical properties

Physical properties at Site U1387 were determined in Holes U1387A, U1387B, and U1387C. High-resolution scanning at 2.5 cm intervals on whole-round segments was performed on the WRMSL. The Special Task Multisensor Logger (STMSL) was only used for Hole U1387B for stratigraphic correlation purposes. NGR scanning was performed on all cores of Site U1387 at low resolution (~20 cm intervals) with one position of the detector array. Thermal conductivity probes were applied on Section 2 or 3 in Cores 339-U1387A-1H through 6H because of the switch to XCB coring at 47.7 mbsf. Reasonable P-wave velocities were obtained on the WRMSL only in Hole U1387A until Core 6H and on split-core segments (working half) in Hole U1387A for every section in the upper 50 mbsf and in some lithified parts in Hole U1387C (Cores 339-U1387C-19R and 43R through 50R). Moisture and density were determined for every second section of each core on Hole U1387A and downhole below Core 339-U1387C-8R. Color reflectance and split-core point-magnetic measurements were made for every segment at 5 cm intervals.

Based on these physical properties, four main intervals were identified. The first change occurs at Site U1387 at 50 mbsf, above which high values of magnetic susceptibility, gamma ray attenuation (GRA) density, NGR, and the spectral reflectance component a* are recorded (physical properties Unit I; Figs. F29, F30). Below this depth downhole to 220 mbsf, a second interval is characterized by a persistent positive correlation between GRA density, NGR, magnetic susceptibility, and spectral reflectance components a* and L*. In particular between ~50 and 100 mbsf, trends as well as superimposed fluctuations agree very well between these parameters. Below 100 mbsf, the positive correlation between all studied physical properties remains, except that L* trends start to correlate negatively with the rest of the studied parameters (physical properties Unit II; Figs. F29, F30). A third interval is defined between 220 and ~460 mbsf, characterized by variable correlations between the studied parameters. Particularly notable is the occurrence of more distinct decimeter- to meter-scale cycles, high fluctuations in L*, and mostly low magnetic susceptibility values (physical properties Unit III; Figs. F29A, F30A). The discontinuity previously defined by Llave et al. (2001, 2007a, 2011) and Hernández-Molina et al. (2006) as the mid-Pleistocene revolution discontinuity (MPR) and as Heinrich Event H4 by Roque et al. (2012) at this site appears to be approximately between 198 and 230 mbsf. Therefore, there seems to be a good correlation between major unconformities and cyclical changes in the lithology reflected by physical properties, but further detailed work is required for confirmation. The fourth interval was differentiated downhole from ~460 mbsf to the base of Hole U1387C, where in spite of poor recovery the most remarkable feature is that NGR is more positively correlated to GRA density and negatively correlated to L* and, albeit less pronounced, a* (Figs. F29, F30). The identified hiatus at ~460 mbsf (see “Lithostratigraphy”), coincident with this third change in physical properties and lithology, indicates a transition from a mixed turbidite-contourite system with the presence of slumps below 460 mbsf to a more contouritic system above. Elevated GRA density, NGR, and magnetic susceptibility values are observed below 800 mbsf, where they are often correlated, although not consistently, to the presence of sandstones in the sediment (Figs. F29B, F30B).

Whole-Round Multisensor Logger and Special Task Multisensor Logger measurements

The STMSL was not used for Hole U1387A because no immediate acquisition of data for stratigraphic correlation was necessary, but the STMSL was used for Hole U1387B. Temperature equilibration before starting with the WRMSL was at least 3 h.

Gamma ray attenuation density

WRMSL GRA density data at Site U1387 is highly variable between 1.6 and 2.2 g/cm3 (Fig. F29). Some of the observed scatter is due to incomplete filling of liners, especially in sandy sections and cores to which the more destructive XCB and RCB drilling techniques were applied. Cores taken using the RCB (Hole U1387C), furthermore, show low GRA densities compared to those taken with the APC and XCB (Holes U1387A and U1387B), caused by lesser volume of the actual sediment recovered. Thus, GRA density in Hole U1387 is particularly biased by core recovery. This is also evident from the continuously underestimated GRA density with respect to bulk density determined on discrete samples, especially in parts of Hole U1387C with poor recovery (Fig. F29).

In general, the GRA density record reflects variations in sand content of the sediment and its texture. An uppermost interval from 0 to 50 mbsf can be distinguished by two distinct GRA density peaks ranging from 2.1 to 1.6 g/cm3. The high densities in the upper interval of Site U1387 mainly correspond to intervals rich in sandy beds and silty mud. Below a minimum in GRA density of 1.6 g/cm3 at ~50 mbsf, density increases until ~250 mbsf with no general increase below this depth. GRA densities oscillate between 1.7 and ~2.1 g/cm3 throughout the entire record. The highest GRA densities in this interval correspond to a dolostone (Core 339-U1387C-19R), debrites, slumps, and well-cemented sandstones. Otherwise, a complex relation to grain size is observed. Sandy silty layers show high values in GRA density, NGR, and magnetic susceptibility (Section 339-U1387C-2R-5); however, sandy layers often have low values of magnetic susceptibility and NGR but high GRA density. In other cases, low values of magnetic susceptibility, NGR, and GRA density are identified (Section 339-U1387C-5R-5). This indicates that the compositions are a variable mixture of contourites and turbidites, particularly with regard to the abundance of calcareous material, with no clear distinction possible between the two sedimentary facies.

Magnetic susceptibility

Based on WRMSL data, magnetic susceptibility increases in the upper 20 mbsf at Site U1387, with maximum values at ~50 × 10–5 SI (Fig. F29A). At 30–50 mbsf, a divergence between magnetic susceptibility and GRA density becomes apparent in that GRA density has a second maximum, whereas magnetic susceptibility remains at an intermediate level. Magnetic susceptibility is rather low (10 × 10–5 to 20 × 10–5 SI) below ~100 mbsf with some peaks that become less abundant below ~320 mbsf, except between 800 and 820 mbsf where there are slightly higher values. The cyclic fluctuations in magnetic susceptibility presumably record alternating beds that are richer in either biogenic or siliciclastic components. The downhole decrease in signal strength is possibly due to diagenetic degradation of magnetic properties. The lowermost part of Hole U1387C is dominated by turbidites, slumps, and 20 m of lithified sandstone, correlating with the observed high values in magnetic susceptibility in this interval.

WRMSL and split-core magnetic susceptibility data agree relatively well for APC cores. For sediments retrieved using the XCB or RCB, WRMSL values are consistently lower and the split-core magnetic susceptibility values have considerable scatter. We assume that this can be attributed to the incomplete filling of the liner for those cores, providing less volume for sensor loop integration, as well as the fracturing and biscuiting of the material that particularly affects the split-core measurements.

P-wave velocity

Sonic velocities were measured with the WRMSL for Hole U1387A, and an attempt was made to determine P-wave velocities on split cores in each section of Holes U1387A and U1387C (Fig. F29). Because of poor sediment-to-liner coupling, reasonable results from the WRMSL could only be obtained for the upper ~45 mbsf. Although the sediment surface appeared to be smooth and should have provided adequate coupling to the transducers on split cores, no clear acoustic signal could be obtained for P-wave velocity determinations at greater depth. An explanation for this might be the relatively stiff and brittle nature of the sediment, which promotes the formation of small cracks that negatively affect signal propagation.

P-wave velocity measured with the WRMSL follows the trend of increasing GRA density in the upper 30 mbsf, with values of 1450–1500 m/s in the uppermost intervals, almost 1700 m/s between 10 and 20 mbsf, and a decrease to 1400–1570 m/s at 20–30 mbsf (Fig. F29A). The increase in GRA density between 30 and ~50 mbsf is not well covered by the WRMSL record but can be traced with the split-core data (only accounting for automatically processed data). Both types of measurement agree well, especially when only considering split-core data with high signal quality.

In Hole U1387C, some discrete specimens of lithified rock were analyzed. In the dolomite-cemented interval in Core 339-U1387C-19R, a P-wave velocity of 2590 m/s was determined, whereas the indurated sandstones found in Cores 44R through 50R have values between 3500 and 5600 m/s (Fig. F29). Sonic velocities >5000 m/s correspond to a sandstone layer, mainly composed of carbonate cement with bioclasts.

Natural gamma radiation

The upper part of Hole U1387A reveals high NGR values (35–50 cps) followed by minimum values of ~30 cps at ~50 mbsf. Below this depth, NGR gradually increases to peak values of 60 cps with superimposed fluctuations. A notable positive correlation of NGR with GRA density, magnetic susceptibility, and a* exists above 220 mbsf (Fig. F30A). From 220 to 650 mbsf, NGR values correlate inversely to GRA density and a* but positively to magnetic susceptibility. Below 650 mbsf, the cyclicity of NGR values shows a positive correlation to magnetic susceptibility, GRA density, and a*. The varying nature of the correlation between the different physical properties indicate that changes in the depositional style and sediment source are affecting the mineralogical composition of the sediments.

Moisture and density

Determination of moisture and density on discrete sediment samples was performed on every second section of each core in Hole U1387A (Fig. F31). Sampling was preferably carried out on the dominant lithology of a section, avoiding small interlayers of differing grain size (e.g., centimeter-scale sand layers). Porosity and moisture content generally run in parallel, both decreasing downhole until the sandstone-bearing interval (Cores 339-U1387C-43R through 47R) is reached. This decrease is more obvious in Hole U1387A, with values of 23%–32% (moisture content) and 44%–56% (porosity) at the top and 18%–24% and 37%–47%,respectively, at the base of the hole. In Hole U1387C, moisture content and porosity only slightly decrease to 16%–25% and 35%–47% (Fig. F31). Interestingly, the dolomite-cemented interval in Core 339-U1387C-19R, itself characterized by low moisture content and porosity, is not related to any change in the general trends. The sandstone beds in Cores 339-U1387C-43R through 47R are low in porosity and moisture content, as expected for well-cemented sandstone. They also seem to represent a boundary below which no decreasing trends in porosity or moisture content are discernible. Slightly above the sandstones, in Cores 339-U1387C-38R through 40R, lower porosities are found in sandy intervals, indicating that some degree of cementation occurred in these intervals as well. The somewhat cyclic pattern of moisture and porosity fluctuations continues farther downhole without clear evidence of a relation to grain size variation described in the lithostratigraphy.

Grain density varies mostly between 2.7 and 2.85 g/cm3 at Site U1387. The upper ~140 mbsf is somewhat exceptional because grain densities are commonly <2.81 g/cm3 (Fig. F31A). Also, occasional samples from 340–390 and 740–790 mbsf had high grain densities. The admixture of calcareous material might explain the high density in these cases.

Thermal conductivity

Thermal conductivity was measured once per core using the full-space probe, usually close to the middle of Section 3, in Cores 339-U1387A-1H through 6H (see “Downhole measurements”). Cores retrieved using the XCB and RCB were affected by cracks and voids, producing unreliable thermal conductivity measurements; therefore, no measurements are reported from these cores. The results show a decrease from values close to 1.7 W/(m·K) at the top of Hole U1387A to 1.3 W/(m·K) at ~45 mbsf. Thermal conductivity often follows reduced porosity and water content of the sediment, as observed for Site U1386. However, in the case of Site U1387 the direct comparison of conductivity to both moisture content and porosity does not convincingly confirm this pattern. The coarse spatial resolution of the measurements and the variable lithology might explain this lack of correlation to a certain extent; however, other as yet unknown factors might influence thermal conductivity as well.

Summary of main results

Based on physical property data, four main units were identified, commonly related to boundaries between the defined lithologic units and subunits. The upper part of Site U1387 to 50 mbsf (physical properties Unit I) is characterized by high magnetic susceptibility, GRA density, NGR, and a* values.

Between 50 and 220 mbsf (physical properties Unit II), physical properties show pronounced cyclicity and a positive correlation between physical properties, except L*, which is negatively correlated with the rest of the studied parameters. A third unit (physical properties Unit III) is defined between 220 and ~460 mbsf, characterized by variable correlations between the studied parameters. Physical properties Unit IV is differentiated downhole from ~460 mbsf to the base of the Hole U1387C. Although affected by poor recovery, the most notable feature of the latter interval is that NGR is positively correlated to GRA density and negatively to L* and, albeit less pronounced, a*. The upper two intervals reflect a good correlation between physical properties, lithology, and cyclical climatic changes.