Considerations and Pitfalls of High-Resolution
Seismic
Acquisition, Processing, and Data Analysis: Gulf of Mexico, Offshore Louisiana
Susan E. Nissen, Janet M. Combes, H. Joseph Mckinley, Anthony G. Nekut, William B. Pramik
As part of an industry consortium project studying Pleistocene sedimentation
on the outer continental shelf and upper slope of the Gulf of Mexico, 684 km of
shallow, high-resolution
, 2-D seismic data were collected offshore Louisiana.
The acquisition parameters for the seismic survey were designed to image
stratigraphic features at scales comparable to those seen in core and outcrop
(approximately 1/2 meter to 10's of meters). Digital data were acquired to 600
msec below the seafloor at a sample interval of 0.25 msec, using an anti-alias
filter of 1440 Hz. The receiver was a 100 m-long, 16-channel streamer, with a
group interval of 6.25 m, and the source was an 8 cu. in. sleeve gun array (4 ×
2 cu. in.), with a shot interval of 3.125 m.
Acquisition design parameters specified a source and receiver depth of 1 m to
attempt to optimize temporal resolution
. Subsequent analyses of arrival times
for direct waves, the primary and multiple sea floor reflections and the
associated ghost arrivals indicated that the source and receiver depths were
different than specified by the design parameters. These analyses suggest that
the streamer was tilt (near group) to 3 m (far group) and that the source depth
varied from less than 1 m to 4 m. This led to several complications in the
processing of the data, particularly with respect to determination of stacking
velocities. The incorrect acquisition geometry required anomalously slow
velocities (between 1250 and 1350 m/s) to stack the sea floor and sub-bottom
reflections for most CDP g thers. Conversely, ghosted arrivals stacked at
abnormally high velocities (up to 2000 m/s). Initially, the higher stacking
velocities were used in processing, causing the primary reflections to be
attenuated and the ghost arrivals to be enhanced. The net result of this is that
the events on the stacked sections are phase reversed and slightly delayed from
what would normally be expected. However, the quality of the processed data was
still excellent.
Frequencies of at least 400 Hz are needed to image (with a vertical
resolution
of 1.5-2.2 m) the detailed internal geometry of the shelf-edge deltas
targeted by this study. Analyses of amplitude spectra and filter panels of the
stacked seismic data show that frequencies up to 1000-1200 Hz were actually
recovered from the subsurface to approximately 150 msec (112.5 m) below the
seafloor, permitting vertical
resolution
of beds as thin as 0.4 m. Within the
deltaic clinoforms, which dip at angles up to 8°, frequencies above 850 Hz may
be spatially aliased at the 3.125 m CDP spacing of this survey. In order to
avoid spatial aliasing of the 1200 Hz events at a dip angle of 8°, a maximum CDP
spacing of 2.2 m is required.
AAPG Search and Discovery Article #91020©1995 AAPG Annual Convention, Houston, Texas, May 5-8, 1995