Water-Bottom Roll

Key Takeaways

• Water-bottom roll generation is strongest when the seismic source is coupled efficiently to the seafloor, which occurs in shallow water surveys (where the air gun is close to the seafloor) and in deep water surveys over soft, unconsolidated seafloor sediments (pelagic ooze, hemipelagic mud) with very low shear velocities (below 150 m/s): in water depths below approximately 50 meters, the seismic source (air gun array) is close enough to the seafloor that the direct wave from the source to the seafloor impinges at near-normal incidence and efficiently converts P-wave energy from the water into S-wave energy in the seafloor at the point of impact, generating a Scholte wave that then propagates laterally along the interface; in very deep water (above 1,000 m), the source-to-seafloor travel time is long enough that the angle of incidence at the seafloor is relatively steep for near-offset receivers, reducing coupling efficiency; however, in both shallow and deep water scenarios, the high-amplitude, low-velocity water-bottom roll can still be a significant noise source if the seafloor sediment has low shear velocity, because the wave's interference with reflected arrivals at near-offset traces degrades the near-trace amplitude and phase information that is critical for shallow target imaging and for AVO analysis at short offsets.
• Velocity discrimination between water-bottom roll and primary reflections provides the primary basis for attenuation in seismic data processing: water-bottom roll travels at the Scholte wave phase velocity (100 to 600 m/s, depending on seafloor sediment stiffness), while P-wave reflections from targets in the subsurface travel at the apparent velocity of the water-layer velocity (approximately 1,480 m/s at near offset) and progressively higher apparent velocities at farther offsets after NMO correction; this large velocity contrast (factor of 2 to 10 between water-bottom roll and primary reflections) allows f-k (frequency-wavenumber) filtering to separate the two events in the two-dimensional (frequency, wavenumber) domain of the shot record, since the roll occupies a low-velocity "fan" region of the f-k space (positive and negative wavenumbers at low f/k slopes) while the reflections occupy the high-velocity region near the zero-wavenumber axis; f-k fan filters that mute the energy in the low-velocity cone of the f-k domain effectively attenuate water-bottom roll by 15 to 30 dB without significantly affecting primary reflection amplitudes, which is typically sufficient to remove the roll noise from interference with shallow events; in practice, f-k filters must be designed carefully to avoid aliasing (where the spatial sampling interval of the hydrophone group spacing aliases the roll's wavenumber content into the reflection zone) and to minimize frequency-dependent edge effects that distort reflection amplitudes at the filter boundaries.
• Water-bottom roll in ocean bottom seismometer (OBS) and ocean bottom cable (OBC) surveys is more intense than in streamer surveys because the seafloor receivers record the full particle motion of the interface wave directly at the seafloor, rather than the pressure field in the water column above the seafloor: a streamer hydrophone at 8-meter depth in 200-meter water depth records primarily the pressure wave in the water column, with the water-bottom roll pressure contribution decaying exponentially with distance above the seafloor (approximately as exp(-k*z), where k is the wavenumber and z is the height above the seafloor); an OBC receiver clamped to or resting on the seafloor records the full Scholte wave particle velocity and pressure with no depth decay, making water-bottom roll the dominant noise in the near-offset OBC record at travel times from the seafloor to the first shallow reflector; the three-component (3C) geophone sensors on OBC receivers also record the horizontal components of the Scholte wave particle motion (which have a characteristic retrograde elliptical pattern at the interface), enabling hodogram analysis to identify and separate the Scholte wave contribution from P and S body wave reflections in the processing; polarization filtering based on the hodogram of the vertical and radial horizontal components is a powerful method for attenuating water-bottom roll on OBC data when the Scholte wave velocity is close to the velocity of shallow S-wave reflected or refracted arrivals (making simple velocity-based f-k filtering insufficient).
• Shallow water seismic surveys in estuaries, lakes, bays, and continental shelves face severe water-bottom roll contamination that limits shallow target resolution: in water depths below 20 to 30 meters, the seismic source (whether air guns, sparkers, or sub-bottom profiler transducers) is only a few meters above the seafloor, maximizing the coupling of acoustic energy into S-waves at the water-sediment interface and generating intense, low-frequency water-bottom roll that arrives at the near-offset receivers within milliseconds of the direct wave, overlapping the two-way reflection time from shallow targets at 10 to 100 ms depth; high-resolution surveys for environmental assessment, archaeological investigation, geotechnical hazard evaluation, and shallow-gas detection must design around this constraint by using higher-frequency sources (which generate shorter-wavelength Scholte waves that attenuate more rapidly with distance), shorter receiver spread lengths (which reduce the travel distance for roll accumulation), and smaller receiver group intervals (which improve the spatial resolution of the f-k filter for roll removal); the minimum depth resolution in shallow water seismic is ultimately limited by the ability to suppress water-bottom roll without destroying the reflection signal at the same arrival times.
• S-wave velocity profiling of near-seafloor sediments using water-bottom roll dispersion analysis is an emerging application that converts the noise event into a useful measurement: the Scholte wave is dispersive (its phase velocity varies with frequency because the wave samples the S-wave velocity structure to increasing depths as frequency decreases), providing a depth-dependent S-wave velocity profile from the surface down to approximately one wavelength of penetration; multi-channel analysis of surface waves (MASW) adapted to marine settings uses the recorded water-bottom roll on OBC or multi-channel hydrophone arrays to extract the phase velocity dispersion curve (phase velocity versus frequency) and invert it for the near-seafloor S-wave velocity profile using standard surface wave inversion algorithms; near-seafloor S-wave velocity is a key parameter for geotechnical engineering of subsea infrastructure (pipeline anchoring, foundation design for gravity-based structures, cable burial assessment), seafloor landslide stability analysis, and seismic hazard assessment in earthquake-prone offshore areas; the MASW-from-Scholte-wave technique provides S-wave velocity profiles from 0 to 50 meters below the seafloor at lateral resolution of the receiver array length, comparable to the information obtainable from laboratory testing of geotechnical cores but at much lower cost and at full seafloor coverage.