Multichannel 3-D seismic data were collected during summer 2000 on the Thomas Thompson (Tréhu and Bangs, 2001). The survey covers a 4 km x 9 km region that includes the southern summit of Hydrate Ridge and an adjacent slope basin to its east (Fig. F1C). The ship was navigated by Global Positioning System (GPS) provided by Racal (including differential GPS [dGPS] corrections from a weighted average of four Racal base stations). These raw fixes were recorded at 1 Hz. They were smoothed to remove the rolling and pitching of the ship for derivation of the location and the velocity of the ship at each update and were corrected to compensate for the offset of the source from the GPS antenna. The streamer's geometry was determined by four bird compasses positioned at 150-m intervals. These records were used to reposition each single receiver channel. There was no tail buoy for independent determination of the location of the end of the streamer. The depths and the wing angles of the birds were monitored by the watch-stander using a graphical display. Shots were fired at 15-m intervals based on the smoothed velocity estimate from the dGPS fixes. The source was two generator-injector (GI) guns, each set at 45/45 in3, towed at a depth of 2.5 m, and fired simultaneously. They proved to be a good broadband source, producing usable energy from 10 to 240 Hz. The signal was recorded on a portable, 600-m-long, 48-channel streamer, which was towed at a depth of 2.5 m.
The bridge used the Nav99 program written by Mark Weidenspahn (University of Texas Institute for Geophysics) to steer the source location down the selected line. Nav99 also provided a minimum radius turn path between lines, which were always shot heading east in the northern half of the grid and heading west in the southern half. Recorded trace length was 6 s. The 3-D survey grid consisted of 81, 11-km-long east-west sections that were shot 50 m apart in a racetrack pattern. During the cruise, 3-D fold was monitored based on streamer reconstruction to identify locations where additional data were needed in order to obtain adequate fold in each bin. This resulted in reshooting 18 lines because of variable degrees of streamer feathering or navigational problems resulting from strong currents. Onboard the Thomas Thompson, the raw SEGD field data were copied from 3480 tapes to SEGY disk files using Sioseis. The SEGY disk files were then converted to Paradigm Geophysical's FOCUS internal format. Seven bad channels were deleted at this stage. Data were processed at sea as two-dimensional (2-D) lines for quality control and a first look at the data. Source navigation and streamer location data for the 3-D grid were converted to universal transverse Mercator (UTM) coordinates and written in UKOOA90 format. These navigation data were converted into FOCUS format 3-D navigation traces and binned using the FOCUS 3-D QC/binning package. Fold maps were generated and optimized for 50 m x 25 m and 25 m x 12.5 m bin sizes. In both cases, the optimum grid azimuth was 348.5°.
On land, data were sorted into 12.5 m x 25 m bins for the 3-D processing and renumbered from 200 to 360 to accommodate interpolated lines. A velocity analysis was done for every 10th line and every 100th common midpoint (CMP). A high-pass filter with a ramp from 15 to 25 Hz was applied to the traces to remove noise resulting from choppy seas and a large swell. The data were then corrected for normal moveout (NMO), the inner trace was muted (to attenuate the seafloor multiple), and the data were stacked. A 3-D poststack Kirchhoff migration using stacking velocities was applied. This yields the 3-D volume, which is 4 km wide x 9 km long and contains a frequency range of 20 to 180 Hz with a dominant frequency of 125 Hz, resulting in a nominal resolution of 3 m (1/4 wavelength). The continuity of north-south and time slices through the data volume, however, is degraded by static effects due to tidal changes in water depth, swell, and changes in the towing depth of the streamer and/or GI guns.
Six regional 2-D lines were recorded on the same multichannel streamer (Tréhu and Bangs, 2001). The tracks of the three north-south lines used during this study are shown on Figure F1C. The shot interval for these lines was 37.5 m, and data were sorted into 12.5-m bins, resulting in 16- to 24-fold data. A high-pass filter was applied with a ramp from 25 to 35 Hz. The data were corrected for NMO, stretch mute was done with a 35% maximum stretch, and the data were stacked assuming a constant velocity of 1500 m/s. A velocity of 1500 m/s was also used for frequency-wave-number migration of the data.
Nine sites were drilled within the boundaries of the 3-D survey during ODP Leg 204 (Fig. F1C). The maximum coring depth ranged from 90 to 540 meters below seafloor (mbsf). Sediment classification was recorded based on visual observation and core description, smear slide analysis, and correlation with physical property data (Shipboard Scientific Party, 2003a). Mudstones and siltstones of turbidite origin, debris-flow deposits, and intervening hemipelagic clays dominate the lithologies observed in the Leg 204 cores. Classification of sediments into lithologic units was based on sometimes subtle changes in major and minor lithologies, changes in biogenic content, changes in magnetic susceptibility, and changes in the frequency and thickness of turbidite layers. Absolute ages of strata were determined based on the bioevents distinguished in the cores and presented in the Leg 204 Initial Reports volume (Tréhu, Bohrmann, Rack, Torres, et al., 2003). The biostratigraphic zones were inferred from the first and last occurrences of diatoms and calcareous nannofossils. Biostratigraphic ages and lithologic units determined by the Leg 204 shipboard sedimentologists are summarized in Figure F2. Although this lithologic and biostratigraphic information was used to aid in the interpretation of the seismic data, seismic units (also shown in Fig. F2) were determined independently from the lithologic units based on seismic stratigraphic packages, as outlined in "Use of Sequence Stratigraphy in a Tectonically Active Environment." The boundaries between lithologic and seismic units generally coincide.