A SIOSEIS seismic processing software script (sio_geom) was used as follows:
1. The raw SEG-D shot files were read with the process SEGDDIN specifying the geometrics format; the process HEADER was used to insert time of day into the header based on the start and end times recorded in HYPACK files and from the cruise survey log entries. The process GEOM was used to describe the shot and streamer geometries and to calculate the reflection point (RP) numbers used to gather (sort traces) the seismic line. The process GEOM also set the shot-receiver distance in the trace header of every trace. GEOM type 9 was used for all lines (except for lines 2 to 6 and l30f1) to calculate distances between shots from shot point locations in the SEG-D headers and to bin traces as common RP's (reflection points). GEOM type 6 was specified for lines 2 to 6 (because positions were absent or bad in the header), which computes a distance from last shot (DFLS) for each shot based on the shot time in the header and the navigation from an ASCII file containing time and position. The ASCII navigation file was derived from the HYPACK data and merged with RP files with the Unix join command. The process GATHER was used to sort the shot order traces by the RP numbers (computed by GEOM). A normal move out (NMO) applied a travel time correction to each trace based on time offset and a velocity of 1530 m/s. Lastly, the shot ordered trace gathers were written with the process DISKOX in SEG-Y rev. 1, IEEE floating point format. This step and all subsequent steps were completed by Dave Foster.
2. A SIOSEIS script (sio_gather) was used to sort traces into RP order trace gathers using the RP values calculated by the SIOSEIS process GEOM. RP ordered navigation coordinates to be used in process step 3.
3. A SIOSEIS script (sio_nmo) was used to apply a normal moveout to the RP gathers from step 2. Moveout velocities of 1510 to 1540 m/s were used.
4. A shell script (layback_stack) read the navigation output created in step 2 and extracted the navigation coordinates for channel 1 within each RP. The result was read by a Python script (layback_rp.py) which applied a layback correction (offset from the GPS antenna source and to the midpoint from the source to the first active channel). The script calculated easting and northing differential values between trace positions. Headings between consecutive traces were calculated using the arctangent function (arctan2(dy,dx)), and reciprocal back bearings were determined using a lookup table. Back bearings were smoothed along track using a moving median function. Layback easting and northing offsets were calculated by multiplying the layback distance by the sine and cosine of the smoothed back bearing, respectively. Offset values were then added to the original coordinates to produce layback positions. The script (layback_stack) then called a SIOSEIS script (siostackpop) which stacked the RP gathers and apply a bandpass filter. The process DISKIN read the RP sorted SEG-Y file. The process STACK was used to sum traces, compute the average amplitude for each trace sample, and write the computed samples to one trace. The trace header values of the first trace in the gather were used for the stacked trace. The process FILTER applied a zero-phase bandpass frequency domain filter between 350 and 1800 Hz with a slope of 48 decibel per octave slope. The process HEADER was used to populate the SEG-Y headers with the layback corrected RP coordinates. Lastly, the processed stacked traces were written to disk with the process DISKOX in SEG-Y rev. 1 , IEEE floating point format. The script layback_stack used AWK to format navigation files with unique RP positions and positions for every 100 reflection points in CSV format, which contains attributes for layback corrected UTM Zone 18 coordinates, layback corrected geographic coordinates, line (file) name, year, day, RP number, and fold. Finally, a Python script (GBrptoSQL15001.py) written by Wayne Baldwin, imported the CSV files to the Spatialite (version 3.0.1) enabled SQLite (version 3.7.9) database. Esri shape polyline and shot point files were exported directly from the SQLite database.
5. A Seismic Unix (version 4.0) script (plot_geomet) was used to read the stacked, and filtered SEG-Y files and to plot the data as 8-bit gray scale Postscript files using the Seismic Unix 'psimage' algorithm. All images were created with a horizontal scale of 60 traces per inch. Images were plotted within a constant 14-inch vertical window (Two-Way Travel Time) duration. The Postscript images were then converted with ImageMagick (version 6.8.7-2) to PNG format.