Velocity and density profiles were constructed for each of the three sites by merging measurements from downhole logging and lab measurements and editing erroneous values. Preliminary processing of the logging data was carried out during Leg 180 (Taylor, Huchon, Klaus, et al., 1999), with final processing, including depth shifts and corrections for borehole diameter, carried out by the Lamont-Doherty Earth Observatory Borehole Research Group. Downhole logging was carried out at Site 1118 between 891 meters below seafloor (mbsf) and the mudline, at Site 1109 between 786 mbsf and the mudline, and at Site 1115 between 784 mbsf and the mudline. Velocity measurements were made using the array sonic tool in linear mode with the eight-receiver array used in short- and long-spacing modes to provide full waveform analysis. The logging velocity (DTCO), derived from coherence of the full eight waveforms, was used in preference to the short- and long-spacing data because of the lower levels of noise and closer agreement with shipboard velocity measurements. Downhole density (RHOM) measurements were made using the hostile environment lithodensity sonde, using a gamma ray source to determine the electron density of the formation. Quality control of the logging velocity and density data made extensive use of the caliper log (LCAL) (Shipboard Scientific Party, 1999b, 1999c, 1999d) to determine the location of washouts, where the caliper became fully extended and potentially out of contact with the borehole wall. Data from such locations were considered unreliable and were therefore discarded. Unreasonable values of density and velocity (such as those that recorded the properties of the drilling fluid) were also discarded.
Two sources of laboratory P-wave velocity measurements were used. The PWS1 and PWS2 insertion probe system was used for unconsolidated sediments (Blum, 1997). This provided measurements of the longitudinal (along the core axis) and transverse (across the core axis) velocity, respectively, based on a fixed distance and measured acoustic signal transit time. In semilithified sediments and rock cores, the PWS3 contact probe system (Boyce, 1976; Blum, 1997) was employed. The PWS3 system accommodates a variable sample thickness, which is measured in addition to transit time and can be used to measure velocity on split cores (across the core axis) or individual samples (in any orientation). When PWS1, PWS2, and PWS3 velocities were available for a single sample, we found that the PWS1 and PWS2 velocities were an average of 40 m/s slower than the PWS3 measurements. As the cores were not oriented, PWS1 and PWS3 velocity measurements should on the average be equal, thus indicating an instrumental error of ~40 m/s. As the PWS3 measurements were found to be in close agreement with the DTCO velocities, 40 m/s was added to the PWS1 and PWS2 velocities to make them approximate the PWS3 velocity (and therefore the logging velocity). For the data compilation, the more reliable PWS3 measurements were used when possible. PWS1 measurements (the longitudinal velocity better simulating logging measurements) were used in places not measured by the PWS3 probe. PWS1, PWS2, and PWS3 velocities were taken, on average, every 1.7 m, sediment type allowing. Laboratory measurements of bulk density were obtained during routine measurement of index properties. The sample interval averaged 2.25 m.
The combination of shipboard and logging data that went into making composite velocity and density profiles for each site is shown in Tables T1 and T2. For all sites, outlier values (e.g., unexplainable spikes or densities/velocities below that of water) were removed from the composite profiles before the data were resampled to 0.5 m by linear interpolation. No special treatment had to be given to the merge points between the shipboard and downhole measurements, as the data scatter in the shipboard measurements exceeded any consistent differences between the two data sets. No artifacts were seen in the synthetic seismograms resulting from these merge points. Special consideration was given to Site 1118, where the first 205 m of Hole 1118A was drilled without coring. At this site, logging data did not extend above 100 mbsf. Above 100 mbsf, shipboard measurements of density from Site 1109 were used. A 25-m Gaussian filter was applied to remove artifacts in the synthetic seismogram resulting from merging the two data sets from different sites.