We generated a synthetic seismogram from Site 1173 data using the most reliable density and velocity curves available and a source wavelet extracted from the seismic reflection data as described in "Core-Log Seismic Correlation and Seismic Resolution" in the "Explanatory Notes" chapter (Fig. F33). For densities, we constructed logs from various sources for three intervals. An exponential curve through the core densities from Leg 190 Hole 1173A defined density from 0 to 60 mbsf, the LWD densities (RHOB) from Hole 1173C were used from 60 to 150 mbsf, and the LWD densities from Hole 1173B were used from 150 mbsf to total depth. LWD data in the 0- to 60-mbsf interval are unreliable due to excessive hole diameter as inferred from the differential caliper. For velocities, P-wave velocity core data from Leg 190 Hole 1173A defined the velocities for 0-81 mbsf, using a linear fit through the data, and 350 mbsf to TD, using the data directly. The wireline velocities from Leg 190 Hole 1173A were used for 81-350 mbsf. The water column was set at a density of 1.1 g/cm3 and a velocity of 1500 m/s. These logs result in a reasonable reflection coefficient at the seafloor of ~0.1-0.15.
Figure F33 shows the fit of the synthetic seismogram with 10 traces on either side of Hole 1173B. In addition to the good match at the seafloor in both waveform and amplitude, the synthetic seismogram matches well with the reflectivity in the seismic data from the seafloor to 340-350 mbsf. There are events in the synthetic seismograms that match the seismic data well at ~80-100, ~175, ~265-270, and ~300-350 mbsf (Fig. F33). At ~80-100 mbsf there is a compound reflection present in the synthetic and real seismic data beneath a relatively transparent upper layer. This upper layer has been defined lithologically as the outer trench-wedge unit and is underlain by the 19-m-thick outer trench-basin transition facies (Moore, Taira, Klaus, et al., 2001). The compound reflection observed at ~80-100 mbsf may be the result of a thin-layer interference effect from this unit. The boundary between log Units 1 and 2 is at 122 mbsf; so, the trench-basin transition, based on the interpretation of the synthetic seismogram, lies within log Unit 1 at ~80 to 100 mbsf. This apparent conflict may be resolved with more accurate velocities for time-to-depth conversion, or other interpretations of these reflections. At ~175 mbsf a distinctive reflection in the seismic data is reconstructed in the synthetic section due to a small coincident decrease in density and velocity. At ~265-270 mbsf a large velocity increase causes a reflection in the synthetic seismogram that correlates to an event observed in the three-dimensional (3-D) seismic traces. The boundary between the upper and lower Shikoku Basin units (log Units 2 and 3) near 340 mbsf in the 3-D seismic data is characterized by a zone of high reflectivity that is reasonably well matched in the synthetic seismogram. In the synthetic seismogram, this zone begins at ~300 mbsf, ends abruptly at ~350 mbsf, and lies 20-30 m above the corresponding zone in the 3-D seismic data. These reflections in the synthetic seismogram are attributed to a ~0.2 g/cm3 increase in density at ~300 mbsf, a zone of higher velocities by ~50 m/s between 340 and 350 mbsf, and a zone of lower velocities by ~50 m/s between 350 and 360 mbsf.
Below ~350 mbsf, the synthetic seismogram shows a number of strong reflections not seen in the 3-D seismic data intersecting Site 1173 on line 215. The mismatch correlates with variations in velocities measured on the cores and with variations in the LWD densities. The unmatched reflections in the synthetic seismogram below ~350 mbsf illustrate potential problems with core velocity measurements, such as sampling bias or disruption from in situ conditions, as they infer seismic reflections that are not observed in the 3-D data. The diminished reflectivity from Unit II to III is consistent with diagenetic changes in and increased homogeneity of Unit III.
In summary, correlations between the synthetic seismogram and seismic reflection data at ~80-100 and ~300-350 mbsf are consistent with core-defined lithologic boundaries and are regionally significant. Additional good correlations between the log and seismic data at ~175 and 265-270 mbsf and poor correlation beneath ~350 mbsf bear further investigation.