The results of Legs 131 and 190 and a three-dimensional (3-D) seismic survey were correlated with the Leg 196 LWD data. The correlation included physical properties measurements, wireline logs, velocity data, and two-dimensional and 3-D seismic survey images collected during 11 yr of research at Nankai Trough. To ensure accurate correlation of the data, it was important to ascertain the accuracy of the navigation of each of the associated surveys, the hole deviation, and the drill string position at the seafloor relative to the sea surface, the accuracy of the depth-converted seismic data, and the vertical and horizontal seismic resolution. Accurate correlation is critical to extend the study of the subsurface physical properties away from the direct measurements at the borehole using the 3-D seismic data.
During Legs 190 and 196 surface navigation consisted of dynamic positioning at the surface relative to an acoustic beacon at the seafloor at each drill site. Both legs took place after the Global Positioning System (GPS) signal selective availability was removed, thereby providing the accuracy available from P-code GPS. There is no navigation information for the bottom of the drill string, and therefore its exact position relative to the ship's position is unknown. We assume the hole position to be directly below the rig floor. During previous drilling, deviation of the hole from the ship's position was determined from cores recovered at shallow depths by the advanced piston corer. Little significant deviation from the vertical was noted.
The 3-D seismic reflection data (Bangs et al., 1999; Moore et al., 2001) was imaged during an 8 km x 80 km volume recorded to 12 s in time-depth. The survey used a differential GPS navigation system provided by Fugro Geodetic. Three base stations at Okinawa, Seoul, and Sapporo provided differential corrections at 1 Hz. Fixes were smoothed with a 15-s running average filter to eliminate ship's motion and determine shot locations in real time. The single 6-km streamer was navigated with differential GPS fixes transmitted back from the tail buoy and with compass readings at 300-m intervals along the streamer. Streamer configuration was constructed from compass data for each shot. Tests of navigation accuracy conducted in port show that the ships's position fell within a 2-m radius 95% of the time. During the experiment, the streamer was located better or worse by reconstructing the streamer with the compass readings. Locations are better in the direction parallel to the stream then perpendicular to it. In the parallel direction, 90% of the estimated streamer locations lay within ±10 m of the differential GPS position. Perpendicular to the streamer, 90% of the estimated locations lay within ±50 m of the tail buoy differential GPS position.
Horizontal and vertical resolution of the 3-D seismic images is dependent on the data frequency content, which becomes more bandwidth limited with depth of penetration. The Nankai Trough 3-D seismic data have an approximate bandwidth of 8-65 Hz and a dominant frequency of 33-40 Hz depending on depth below the seafloor. The vertical resolution of these data is therefore 10-15 m based on resolving distinctions of one-quarter of the dominant frequency's wavelength. The common midpoint spacing of 25 m in the inline and 50 m in the crossline direction effectively integrates the seismic acoustic impedance data over an estimated first Fresnel zone radius of ~290 m at the seafloor.
First-order depth conversions of the seismic reflection data were calculated using major seismic horizons correlated to the core and log data from Leg 190. Further refinements to the velocity model used for time-to-depth conversions have been carried out through iterative prestack depth migration (Hills et al., 2001) and through velocities obtained from the wireline and core data.
In order to correlate the 3-D seismic data with the LWD data, synthetic seismograms were constructed using the best densities and velocities for each site from the LWD density, core density, wireline P-wave velocity, core P-wave velocity, and ISONIC P-wave velocity measurements. The 256-ms long-source wavelet used at each site was calculated deterministically using 10 traces from the seafloor reflection centered at Hole 1173B.