Locations of seismic lines are shown in Figure 1 and Figure 2. Five long transects cross the ridge flank out to a distance of roughly 150 km from the ridge axis, where the crustal age is 5 Ma. The direction (typically N 107°E) was chosen to be perpendicular to local magnetically defined isochrons, which in most of the area included by the seismic lines contain no significant offsets associated with fracture zones or propagating rift traces (Currie et al., 1982; Elvers et al., 1972; Wilson 1993; Wilson et al., 1984). The length of the lines was chosen to provide the regional context for detailed heat-flow, coring, and seismic refraction studies in several local areas, and for the Leg 168 drilling transects that extend from roughly 20 to 100 km from the ridge axis. Suites of closely spaced lines and lines parallel to the strike of the ridge axis were collected to provide local structural and hydrologic/lithologic control for the detailed studies.
The Hydrocell-95 seismic survey was carried out on board the Canadian RV John P. Tully in August 1995. Lines were shot using a single generator-injector (GI) gun, with a reduced volume of 0.74 L (45 in³) for both the generator and the injector chambers. The air compressor sustained 12 MPa (1700 psi) pressure for nominally 10-s shot intervals. The injector firing delay was set to 35 ms throughout the survey.
Seismic lines in 1996 were shot from the German research vessel Sonne using two GI airguns with a generator chamber volume of 2.9 L (175 in³) and an injector chamber volume of 1.7 L (105 in³). The injector was fired with a delay of 40-48 ms to reduce the bubble oscillations. Operating pressure was 15 MPa (2250 psi), and guns were towed at a depth of 5-6 m.
A Teledyne array was used in both survey years, with 16 hydrophone groups evenly distributed over 100 m of active length. The tow lead was 120 m long, and the active section was directly coupled with no transformer section. All 16 channels were electrically summed into a single channel.
The custom-built acquisition unit digitized data using two 12-bit analog-to-digital converter channels in combination. One channel was offset in amplification by a factor of 512, then both channels were combined into a single 21-bit data value, yielding a nominal dynamic range of 120 dB (20 bits plus sign). Data were sampled at an initial rate of 2-4 kHz and subsequently decimated. A trigger delay (SEG-Y header-word delrt) was applied according to the two-way traveltime of the seismic signal in the water column.
Navigation control was derived from global positioning system data acquired in differential mode (DGPS). In 1996 this was augmented by auxiliary sensors including a Doppler velocity log and gyro compass. Shipboard GPS antenna positions (mounted roughly 15 m forward of the airgun towing point in 1995 and 35 m forward in 1996) were recorded for each shot at the time of the trigger. Trigger signals were derived from the DGPS-determined positions to yield a shot spacing of 25 m along all lines in 1995 and most in 1996. In the instances of Lines 960902a, 05a, and 27a, the airguns were triggered by GPS time once per minute. Ship speed was maintained as steady as possible at about 3 kt to yield a shot spacing of ~100 m along these three lines. Approximate positions for each shot are included in SEG-Y trace headers fields sx, sy, gx, gy, but because SEG-Y navigation keywords do not accommodate the full DGPS precision, separate navigation files are also provided.
Full-precision DGPS navigation data are provided in ASCII files included on CD-ROM Disc 2 in the back pocket of this volume. The DGPS data contains shot times and positions, as well as manually picked two-way traveltimes to the seafloor and to the top of seismic basement. The format for the records of these files (found in the XYZ directory on CD-ROM Disc 2) is shot-#, time-hh, time-mm, time-ss, latitude, longitude, dist-from-ridge, twt-sea-floor, and twt-basement. The day field in the headers of the seismic files is set to the day of the month, not to the Julian day as required by SEG-Y. The date on which the profile was started can be read from the filename. For example, file 950804A.XYZ represents data collected from the line initiated on August 4, 1995. Examples showing how reflection times were picked are discussed below. Calculation of distance from ridge is also described below.
Each line is provided at two different levels of processing. Data in files with name-extension SGY were processed with a bandpass filter and resampled at typically 2-ms sampling intervals. Data in files with name-extension SGM are additionally migrated on a smooth generic velocity-depth function. This velocity-depth function was designed to carry velocities from 1500 m/s at the upper sediments to about 2500-3000 m/s in the basement and is by no means accurate. Because of the great water depth and limited aperture of the seismic system, traveltime migration results are not very sensitive to velocity errors; however, and the quality of the migrated sections is generally very good. Some profiles are additionally dip-filtered after migration to reduce migration noise.
Seismic data are stored in SEG-Y format with an EBCDIC header in individual files with either filename extension sgy (unmigrated data) or sgm (migrated data). These files can be found in the SGY and SGM subdirectories, respectively, on CD-ROM Disc 2 in the back pocket of this volume. File names reflect the date when the line was started. For example, file 950804A.SGM represents data collected from the line initiated on August 4, 1995. Naming convention for the *.SGY, *.SGM, and *.XYZ files has been kept consistent to facilitate matching seismic data and DGPS navigation.