DOWNHOLE MEASUREMENTS

We completed two logging runs at Site 1137 (Table T21) (see "Operations"). First, we logged the hole from 364.7 mbsf to seafloor with the triple combo logging string (resistivity, neutron porosity, and density tools), the HNGS, and the LDEO TAP tool. Interactions between the opening of the caliper and the DLL measurements degraded log quality at the bottom of the hole because both measurements used the same electrical connections in the wireline. While we opened the caliper at the bottom, the DLL current was disabled. We repeated the logs from 364.7 to 310 mbsf. In the sedimentary section (195 to 81 mbsf), borehole diameter exceeded 19 in, preventing good contact of the tools with the borehole wall. Normally, the mudline is located with the gamma-ray data to obtain the depth of the seafloor. However, in Hole 1137A, we could not detect the seafloor by natural gamma-ray gradients because of the extremely low radioactivity of the nannofossil ooze. Consequently, we used the drillers' water depth (1016 m) to shift the logging data to that seafloor reference.

For the second run, we used the spectral gamma-ray, FMS tool and the DSI combination from 367.9 to 81 mbsf. Despite difficulties in passing a series of bridges at 170-185 mbsf, well above the sediment/basement transition, we logged the hole successfully. A bridge at 165-180 mbsf, where caliper measurements for the two runs clearly show progressive closure of the borehole wall, prevented a second pass. Because the bit had already been released, drilling through the tight section was not feasible. Therefore, we abandoned logging operations, and the tool was rigged down.

Logging data of Hole 1137A range from poor to high quality (Fig. F89). The HNGS provides the highest quality data because high measurement precision and corrections for borehole diameter are possible. In the sedimentary section from 81 to 195 mbsf, the logging data are of poor quality. The density and neutron tools did not maintain contact with the borehole wall, and the DSI could not be run centralized. The maximum aperture of the caliper was 19 and 15 in for the first and second runs, respectively. In both runs, the tools lost contact with the borehole wall above 153 mbsf. In the lowermost part of the sediments, from 195 to 227 mbsf, the logging data are of good quality.

The DSI provided high quality compressional wave and shear wave data in the igneous basement. The P&S mode (providing compressional and shear waves) produced good compressional wave data, and we obtained a continuous shear wave profile in the lower dipole mode, even when the formation velocity approached mud velocity. We processed the velocity data in realtime with the Schlumberger MAXIS unit by slowness time coherence (STC) analysis. However, the analysis resulted in erroneous velocities from 177 to 192 and 242 to 244 mbsf. We deleted incorrect velocities from our data files and plots (Fig. F89). In the sedimentary section, shear wave velocities are very low; consequently small changes in compressional wave velocity have a relatively high impact on the V_p/V_s ratios and should not be used for lithologic interpretations. In the basement section the V_p/V_s ratio averages ~2 (Fig. F89B).

Logged compressional wave velocities and densities show good agreement with data obtained from discrete core samples (Fig. F90) in the basement section (below ~225 mbsf). In basement, GRAPE density data are consistently lower than both the logging densities and the discrete core data, as the core liner was not entirely filled. Despite this offset, GRAPE data may be useful for postcruise core-log integration because the high sampling rate produces measurements that generally follow the shape of the log density curve, but are offset (Fig. F90).

We observed a large mismatch between compressional wave velocities from core samples and logs between 200 and 225 mbsf (Fig. F90), corresponding to the glauconite-bearing sandy packstone above basement. Given our revised log stratigraphy (Fig. F91), we attribute the mismatch in Core 183-1137A-24R to low recovery, implying that Unit III is substantially thicker than shown in the core stratigraphy. The remaining mismatch, between 205 and 210 mbsf, is similar to other discrepancies between the two data sets, indicating that velocity peaks in core samples are commonly not sampled by the compressional wave velocity log. Densities from discrete samples and the downhole log agree. Amplitudes in the two datasets are generally comparable, including narrow density peaks.

Above 200 mbsf, the density and velocity data from discrete core measurements and logs appear to agree well. However, this agreement is misleading, as the borehole size exceeded the maximum caliper extension for large parts of the sedimentary section. This results in low-quality velocity and density logs, which show values lower than those expected for the formation. In addition, strong core disturbance of soft sediments caused low discrete core density and velocity values (see "Physical Properties"). Natural gamma-ray logs and MST natural gamma-ray data agree well in the basement section (below ~225 mbsf) (see Fig F82; "Physical Properties"). Neutron porosity and porosity measured on the core samples also agree well. In contrast, previous logging studies in basalts have a general offset of 6% and 10% between core and log measurements observed (Broglia and Moos, 1988; Lysne, 1989). The good agreement of our results is attributable to the use of the new neutron tool, the APS, which measures porosity more accurately in this geologic environment (Schlumberger, 1994).

Shipboard processing provides preliminary FMS images. Because of excellent hole conditions in basement, the FMS images show most features seen in the cores. In basalts, we can distinguish vesicles, fractures, brecciation, and massive zones. In the sedimentary beds within Unit IV, grain-size differences are apparent in the coarser clastic sediments. However, these data sets require further postcruise processing to remove effects of tool sticking.

High core recovery allows calibration of the logging results with core observations. In sections with low core recovery, the combination of standard logs and FMS images allows us to calculate true bed and flow thicknesses (Fig. F91; Table T7). Furthermore, within single flow units, we can estimate the thickness of brecciated, vesicular, or fractured parts.

From standard logging data and the FMS images, we distinguish the boundaries of lithologic Units II, III, and IV (Fig. F91). Unit II shows generally low density, resistivity, and velocity. This is mainly related to the large borehole, and, to a lesser degree, the unconsolidated character of these sediments. Nannofossil ooze shows low spectral gamma-ray values (SGR) (<50 gAPI) because of the absence of clay minerals. At 196.5 mbsf a sharp increase in the SGR marks the boundary to the underlying glauconite-bearing sandy packstone in lithologic Unit III. This unit extends to 227 mbsf, and its log-derived thickness is 30.5 m vs. the ~20 m estimated from the cores. Logging results indicate considerable lithologic heterogeneity within Unit III. It can be subdivided into an upper, middle, and lower part by the SGR. The highest SGR (150-250 gAPI) and uranium values (up to 30 ppm) are in the bottom and top of this unit. The high formation capture cross-section values () in Figure F89A correspond to high potassium contents in the lower part. That is related to glauconite.

Based on interpretation of FMS images and the standard logs, the transition to basaltic basement is at 227 mbsf (Fig. F92). At this depth, SGR and porosity decrease and density and resistivity increase. The different basaltic lava flows in basement show similar characteristics and trends in the logging data. Brecciated and vesicular, and mostly altered, flow tops and vesicular flow bottoms show the lowest densities, resistivities, and photoelectric factors within a single flow. Density ranges from 2.4 to 2.5 g/cm³, resistivity from 8 to 20 m, velocity from 5.5 to 6 km/s, and porosity from 25% to 30%. Depending on the degree of alteration, flow tops and bottoms show slightly increased natural gamma-ray values caused by potassium enrichment. Massive parts of the lava flows show the highest densities (>2.7 g/cm³), resistivities (>50 m), and velocities (3-4.5 km/s) combined with the lowest porosities (<15%). The transition from the massive to the vesicular part of basement Unit 2 is clearly visible in FMS images (Fig. F93). Differences are also apparent among lava flows. For example, basement Unit 4 does not include a homogeneous massive part but is altered and enriched in potassium corroborating the alteration observed in the cores (see "Alteration and Weathering"). Between 272 and 276.4 mbsf, we observe fractures that are mainly filled with dark green clay minerals (see "Alteration and Weathering"). The high amount of clay minerals affects physical properties because density, resistivity, and photoelectric factor are reduced, whereas the porosity and gamma-ray values are increased. In the V_p/V_s ratio (Fig. F89B), we observe only a few deviations from the average ratio of 2, notably at depths of 239-247, 287-288, 322-327, 336-338, and 345-349 mbsf, where the ratios range between 2 and 3.2. These depths correspond to zones within interbedded sedimentary rocks, brecciated basalt, vesicular basalt, and a crystal-vitric tuff, respectively (see Fig. F91). However, the basement units in which the high V_p/V_s values are found do not show elevated V_p/V_s ratios for their entire depths. Some but not all positive V_p/V_s anomalies correspond to peaks in natural gamma-ray values, suggesting a relationship with the local degree of weathering and alteration, and, thus, clay-mineral content.

The volcanic siltstone and sandstone (Unit 5) and the volcanic conglomerate (Unit 6) are clearly distinguishable in the total gamma-ray log. The clasts in Units 5 and 6 are mainly eroded volcanics (see "Lithostratigraphy"). We found the highest potassium (4.5-6 wt%) and thorium contents (>20 ppm) in these two units. The high thorium content is significant and indicates the more evolved character of most components (see "Igneous Petrology"). We distinguish the 4.4-m-thick volcanic siltstone and sandstone from the volcanic conglomerate by lower density, photoelectric factors, and resistivity and higher porosity values. The volcanic conglomerate varies widely in grain size and bed thicknesses with depth in the FMS images. Low resistivity and density correlate with finer grained beds from 302 to 303 mbsf and from 308.4 to 309.2 mbsf (Fig. F94).

Scatter plots of density vs. compressional and shear wave velocity and resistivity vs. porosity (Fig. F95) reveal various relationships among these parameters. Igneous rocks and consolidated sediments show a linear, positive correlation between velocity and density (Fig. F95A, F95B) and a linear, negative correlation of porosity with both density and resistivity (Fig. F95C, F95D). Furthermore, igneous rocks can be distinguished from the conglomerate in the density-porosity scatter plot (Fig. F95C), as both rock types have similar low porosity, but the conglomerate has a lower bulk density caused by lower grain density. In unconsolidated sediments, only porosity and density show any correlation. At densities lower than ~2.4 g/cm³, velocities appear fairly constant (Fig. F95A, F95B). This may largely be a consequence of insufficient velocity log quality where hole diameter is large. Resistivities are fairly constant at porosities >40% (Fig. F95D). However, the linear correlation of porosity and density even for porosities as high as 40%-60% and densities as low as 1.8-2.2 g/cm³ suggests that both log measurements reflect formation properties in this range, rather than reflecting drilling mud properties, as the latter is expected to result in constant porosity and density readings (Fig. F95C).

The caliper logs from the FMS-DSI logging run show remarkable hole symmetry (Fig. F96). No evidence suggests substantial hole deformation or breakouts below 230 mbsf. However, above this depth, the hole is severely enlarged in unconsolidated sediments (Fig. F89).