Downhole measurements in Hole 1165C were made after completion of RCB coring to a total depth of 999.10 mbsf. Prior to logging, Hole 1165C was filled with sepiolite mud, reamed, and flushed of debris. In addition, five round trips of the BHA had been made in the borehole while we waited on icebergs. During each round trip, the condition of the borehole was assessed for collapse and deterioration. Apart from a number of ledges, the borehole was in fairly good shape for the logging runs. Two tool strings were run: the triple combo and the FMS-Sonic (see "Downhole Measurements" in the "Explanatory Notes" chapter; Fig. F76; Table T17).
The triple-combo tool string (porosity, density, resistivity, and natural gamma-ray emissions) was successfully lowered to 995.5 mbsf, within 5 m of the total hole depth, and logged up to the base of the pipe at 174 mbsf. This logging run had initially encountered sticking problems caused by a ledge or bridge at 119 mbsf, just beneath the base of the pipe. To overcome this problem, the tools were partially rigged down and three stands of drill pipe were added to push through the obstruction, putting the base of the drill pipe at 174 mbsf, where it remained throughout the logging operations. During logging, the tool string was trapped under a ledge at 303 mbsf for a short time, affecting the logs at this point. The tools provided continuous and high-quality log data. The borehole diameter averaged ~15 in, increasing to a maximum of 18 in toward the top of the logged section. Occasional ledges were encountered, although on the whole the borehole walls remained fairly smooth.
The FMS-Sonic tool string (microresistivity [FMS], seismic velocity, and natural gamma-ray emissions) was lowered without difficulty to 584.8 mbsf, where a bridge that had been encountered during the previous run prevented access to the lower reaches of the hole. A failure of the z-axis accelerometer on the general purpose inclinometer tool just before logging was followed by a general software communication failure to the FMS tool after only 50 m of FMS data were recorded. The DSI sonic tool was run in the P and S and dipole shear data acquisition modes. The shear wave slowness log was good, but the compressional wave slowness log contained several intervals of anomalously slow values. (The DSI results are not as good where formation velocities are generally slow).
The wireline depth to seafloor was determined from the step increase in gamma-ray values at the sediment water interface to be at 3546.5 mbrf; the driller's mudline depth used for establishing core depth was 3548.7 mbrf.
The sedimentary sequence was divided into three log units on the basis of changes in the character of the downhole logs (Figs. F77, F78).
Logging Unit 1 is characterized by its large-scale uniformity, showing only small variations about the compaction trend. Porosity (density and core) ranges from 65% at the top to 50% at the bottom, and compressional P-wave velocity ranges from 1550 to 1775 m/s. Isolated highly resistive and lower-porosity beds are present, having a greater amplitude in Subunit 1c than in 1a or 1b.
This subunit has, on average, slightly lower natural gamma-ray (HSGR) and photoelectric effect (PEF) values than deeper subunits. In this unit, the HSGR baseline value increases slightly at 210 mbsf. Similar abrupt changes are not observed in the other logs, suggesting that the HSGR shift may reflect a mineralogy change.
This subunit is distinguished from Subunit 1a by initially lower values of the natural gamma-ray emission followed by a rapid increase. This feature, however, may be due partially to the tool string becoming stuck at this location. Correlation between core (MST) and log gamma-ray values is difficult, and MST gamma-ray values do not decrease at this depth. Tentative correlation between the two records indicates that the decrease may occur in a break between Cores 188-1165C-35X and 36X at a log depth of 308 mbsf (Fig. F79). The ledge at this depth, apparent in the caliper log, is possibly a result of swelling clay. Isolated small-amplitude spikes in resistivity (e.g., at 452 mbsf) indicate that harder, likely calcareous, beds are present within the claystones of lithologic Unit III (see "Lithostratigraphy"). Pervasive small-scale lows in density (porosity highs) correlate with lows in natural gamma. These are interpreted to correspond to intervals richer in diatoms: the microfossil framework provides the extra porosity and the reduced clay fraction lowers the natural gamma.
A very resistive bed between 480 and 482 mbsf is the first of several that distinguish the subunit. The interval between 550 and 585 mbsf contains many resistive beds and has a higher background resistivity than the rest of the subunit.
The top of logging Unit 2 is marked by a step decrease in porosity from 55% to 45% and step increases in density and resistivity. The coincident step increase in natural gamma-ray emission can be accounted for by the decreased porosity; therefore, the change from Unit 1 to Unit 2 is more likely to be petrophysical (e.g., cementation) than lithologic. The isolated harder beds with resistivity spikes seen in Unit 1 continue into Units 2 and 3, and they can be correlated with thin calcareous intervals observed in the cores (Table T18). The PEF log also has spikes at the calcareous intervals because calcium carbonate has a higher PEF than most other common rock-forming minerals.
Unit 3 has a somewhat "sawtooth" pattern to the resistivity log—it consists of a series of intervals in each of which the resistivity increases downhole, but the bases have abrupt decreases in resistivity. The overall trend is a porosity decrease downhole, from 40% at the top to around 30% near the base of the hole. The isolated resistivity spikes continue in this unit and sometimes mark the boundaries of the porosity intervals.
There is no strong similarity between the natural gamma-ray values measured on the MST track and the gamma-ray values (HSGR) measured downhole (Fig. F77). Differences are due in part to the variable diameter of the core, especially the XCB cores. Greater porosity will decrease the natural gamma-ray values; however, porosity will be similar in core and log values, except in the drilling slurry between the biscuits. The effect of variance in borehole diameter is corrected for in the HNGS tool during logging; however, even given the above sources of error, the mismatches between core and log values are still larger than may be expected from measuring the same property on the same formation.
Another difference between the core and log measurement lies in the width of the gamma radiation energy spectrum windowed in the measurement: the MST gives the total gamma counts in all parts of the spectrum, whereas the HSGR is more discriminating, including only the gamma-ray energies above 1100 eV—the region covering the main potassium, uranium, and thorium peaks. The MST records data in 250 separate energy windows, so in an effort to compare like parts of the energy spectrum between the MST and the log, the MST range above 1100 eV was used for the comparison (Fig. F79). However, the match remains poor both in terms of the shapes of the cores and in the amplitude of the variations.
The thin, resistive, dense, low-porosity beds observed in the logs correlate well with calcified beds, or intervals containing carbonate nodules, observed in the core (Table T18). As an example, the 24-cm-thick bed at 731 mbsf gives responses in all the logs (Fig. F77, F80). High porosity, density, and resistivity reflect the porosity decrease that accompanies carbonate cementation. The PEF peak reflects an increase in the proportion of calcium carbonate (PEF ~5 barn/e-), and the concurrent drop in natural gamma likely reflects a decrease in the proportion of clays.
In addition to these clear large log spikes, the sediment between them contains smaller, regularly spaced log peaks, or cycles (Fig. F80). To determine the average cycle thickness, resistivity peaks were counted in two intervals, one in logging Unit 1 and the other in Unit 2. Resistivity appears to be the log most sensitive to cementation and is relatively insensitive to borehole effects, so it was chosen for this exercise. Counting gamma-ray or density peaks would lead to slightly different results, but those results would be in the same ballpark as that derived from resistivity. In logging Unit 1, between 307 and 470 mbsf, there are about 90 resistivity peaks with an average cycle thickness of 1.81 m. The dating of this part of the hole is poor, but the sedimentation rate is roughly 8 cm/k.y. At this rate, the cycle duration is roughly 22.6 k.y., close to the precessional periodicity. In logging Unit 2, between 612 and 867 mbsf, there are about 140 resistivity peaks with an average cycle thickness of 1.82 m. There is only one date in this part of the section, but it is thought that the Oligocene was not reached. A reasonable estimate for the sedimentation rate is 11 cm/k.y., giving it an average cycle duration of 16.5 k.y. The smaller log peaks may be caused by minor calcification (or silicification), though at a level too subtle to be observed in the cores. They may also result from cyclic sedimentation processes similar to those described in "Color Alternations in Cores".
A composite synthetic seismogram was created from a combination of core-based P-wave velocities and log-based porosities. Sonic log data were not used to calculate the synthetic trace because only a short log was recorded as the result of an obstruction in the hole at ~580 mbsf and because sections of the data were unreliable. The bulk-density log was used to calculate a density-based porosity, and that porosity was in turn used to create a pseudo-sonic log for the logged interval. The synthetic seismogram was calculated from the pseudo-sonic and bulk-density logs, with the core-based P-wave velocities and bulk densities filling in the unlogged interval above 178 mbsf.
The bulk-density log was converted to density-porosity using the grain densities from the core (see "Physical Properties") and a seawater density of 1.03 g/cm3. Rather than using a single average grain density for the entire log, grain densities were matched depth for depth with the log data because of variations caused by diatom abundance.
The density-porosity was converted to velocity using the global porosity-velocity model for siliciclastic sediments of Erickson and Jarrard (1998). This model proposes an empirical relationship covering the entire range of possible porosities for both normally compacted and highly compacted sediments, derived from an analysis of 23 wide ranging, mostly marine data sets. The bulk-density log was compared to the MAD bulk-density data (see "Physical Properties") for compatibility (Fig. F81) and the core P-wave velocities were likewise compared to the pseudo-velocities derived from the model (Fig. F82) as well as the sonic log data. The pseudo-sonic and real-sonic logs indicated that a 7% in situ correction of the core velocities was needed. Seven percent seemed a bit too much of a correction for the MAD bulk-density data, but for compatibility, both velocity and density were corrected by the same amount.
Porosity is the dominant control on velocity for sediments, including clays, with >40% porosity; velocity does not become lithology dependent until below this threshold (Erickson and Jarrard, 1998). A highly consolidated sediment tends to have a slightly higher velocity at a given porosity than a normally consolidated sediment because of an overall stiffening of the sediment (Erickson and Jarrard, 1998). For the upper half of the hole, the empirical relationship for highly consolidated sediments gave a better match to the discrete core-based P-wave velocities (see "Physical Properties") than the relationship for normally consolidated sediments. Consequently, the empirical relationship relating porosity to velocity in highly consolidated sediments was used throughout.
For the lower half of the hole, core velocities were measured on three orthogonal axes. The x- and y-directions are across the core, with the x-direction value taken through the radius of the split core and the y-direction value across the diameter. The z-direction is down the length of the core (see "Physical Properties"). It was originally thought that the x-direction would be the most representative for in situ conditions because core expansion caused by degassing and lithostatic unloading would part horizontal bedding planes resulting in anomalously low z-direction velocities. The velocities derived from the model, however, greatly underestimated the x-direction core measurements in the lower part of the hole. On the other hand, porosity-velocity estimates are very close to the z-direction measurements, indicating that perhaps the z-direction core measurements are more representative of the true seismic velocities in the core than originally thought and that post-recovery fracturing of the contourite beds and fissile texture of the cores caused less velocity change than expected.
For the lower 100 mbsf of the hole, the porosities are <40%, the z-direction core velocities are lithology dependent, with a shale fraction of 0.9, giving the best estimate to the porosity-velocity model values (Fig. F83).
The source wavelet of the water gun was extracted by digitizing the seafloor reflection of trace 1191 (i.e., shotpoint 1191) from the site survey, near the position of the hole. The wavelet is ~85 ms long (170 ms two-way traveltime), with a 15 ms (30 ms two-way traveltime) precursor arrival before a two-lobed main impulse (Fig. F84). Convolution of this wavelet with a reflection coefficient log derived from the pseudo-sonic and bulk-density data produced the resultant synthetic trace that matches the seismic section quite well in the upper part, but that overshoots the basal reflection (5-83 s) by ~0.2 s (synthesis not shown).
Changing the shale fraction in the model of Erickson and Jarrard (1998) from 0.9 to 0.5 shifts this reflection up to match the reflection without adversely shifting the data above 5.6 s (Fig. F85), although some change in the character of the synthetic is observed. This change to a shale fraction of 0.5 overestimates the z-direction core velocities by ~400 m/s, but improves the match between model velocities and the x- and y-direction core velocities (Fig. F86).
The synthetic seismogram and depth-time models are preliminary, and postcruise work is expected to improve them. Several factors are responsible for mismatches between the synthetic and the seismic. First, the bulk-density data were not edited for any but the most egregious outlying data points even though there are places where hole conditions were rough. Further editing of the log should improve the character match. Second, the synthetic seismogram does not include the loss of signal caused by spherical divergence of the original seismic input and does not contain any automatic gain control. This might explain differences in the visual importance of certain reflections at depth, such as the packet of strong reflectors ending at 800 ms and their equivalents in the synthetic seismogram. Third, the wavelet used to generate the synthetic section is not a completely accurate representation of the original water-gun signature because it was manually extracted from the seismogram based on the seafloor reflection. The presence of any reflections close to the seafloor will have distorted the original wavelet. For example, the base of lithostratigraphic Unit I is at 63 mbsf (see "Lithostratigraphy"), within a wavelet's length of the seafloor, and the reflection from the transition of Unit I to Unit II, if any, is included in the wavelet. Fourth, alternative velocity logs can be derived from the porosity.
The synthetic seismogram was created to put a depth scale on the seismic data and determine the likely position of lithostratigraphic units on the seismic section (Fig. F87). Since the wavelet response to a distinct geologic horizon is a series of impulses over an 85 ms interval, 15 ms of which is a precursor to the main impulse, it is difficult—even on the synthetic seismogram—to identify the precise depth of the horizon that leads to a reflector. The best way to pick depths is to refer to the reflection coefficient log (Fig. F88) that was created to generate the synthetic seismogram.
Six prominent reflections other than the seafloor reflection were identified and their likely origins were identified from the cores.
Reflector 1 (~0.25 s two-way traveltime below seafloor) was identified as a >40-cm-thick nannofossil chalk at 210 mbsf in Cores 188-1165B-24X-3 and 24X-CC. Reflector 2 (~0.44 s) is at the top of a zone of poor recovery at 330 mbsf, around Core 188-1165B-40X, at a depth containing chalk nodules. Reflector 3 (~0.6 s) is the uppermost strongly cherty/calcified bed seen in the hole at ~490 mbsf. Reflector 4 (~0.65 s) is the top of another zone of poor recovery in Core 188-1165B-63X at ~550 mbsf, in an area with much harder calcified beds and claystones. The central zone of reflectors between Reflectors 4 and 5 (~0.82 s) appears to be caused by multiple closely spaced hard calcified beds. Reflector 5 is at a depth of 710 mbsf on the depth-time model. Because of the many closely spaced reflectors in the overlying reflector zone, the wavelet becomes quite distorted, and in fact the impedance contrast that represents the end of this zone is at 650 mbsf (the opal-A/opal-CT transition), with the wavelet trailing out another 60 m. Furthermore, there is a change in Core 188-1165B-74X at the same depth, from a cemented siliceous bed >25 cm thick to a softer claystone. The basal target Reflector 6 (~1.06 s) is found at ~960 mbsf in Core 188-1165C-32R and is a thick calcite-cemented horizon. In general, most strong reflectors are caused by relatively thin single or multiple intervals of calcified or silicified beds with high acoustic impedance.
The Lamont-Doherty temperature-acceleration-pressure tool recorded the temperature of the fluid in Hole 1065C during the first pass of the triple-combo tool string (Fig. F89). These measurements underestimate the formation temperature, as the fluid temperature does not have time to equilibrate to the formation temperature. A temperature of 36°C was recorded at the bottom of the hole (993 m), so the temperature gradient is at least 36°C/km. The downgoing and upgoing curves have an offset of ~5°C, owing to borehole reequilibration during acquisition.