L. Dale Bibee

The Effects of Shear Velocity Structure on Seafloor Noise

The “microseism peak” and adjacent “noise notch” are prominent features of the earth’s background noise spectrum between 0.08 Hz and about 5 Hz. They are caused by the combination of a peak in the source (ocean wave) spectrum and and by the presence of peaks in the spatial mode Q of the 4 gravest surface (and interface) wave modes. The peaks in the mode Q occur at a phase velocity of 1.5 km s-1and are due to the high Q of the water layer. The depressed mode Q at very low frequencies is caused by the low Q of the asthenosphere. The frequency of the low sides of the peaks in mode Q is strongly influenced by the ocean depth and the frequency of the high frequency side is controlled by the shear velocity structure of the shallow sediments.

Implications of Deep-Water Seismometer Array Measurements for Scholte Wave Propagation

A field exercise was conducted in March 1990 to make measurements of Scholte wave propagation characteristics in a deep ocean environment. Signals from a series of bottom explosive shots were recorded on an array of ocean bottom seismometers. Clear Scholte phases were observed on the vertical seismometers to ranges of 1.25 km, but were attenuated to noise levels by 2 km range. Collocated hydrophones did not detect the Scholte wavetrain even at the closest ranges. The ratio of pressure (µPa) to vertical ground velocity (nm/s) was 68 dB in time windows dominated by body waves, but only 20 dB in windows dominated by the Scholte wave. Group velocities were low (30-100 m/s) and showed considerable variability despite the expected uniformity of the seafloor in this abyssal environment.

Full Waveform Inversion of Seismic Interface Wave Data

A technique for inverting marine seismic interface wave data for shear wave attenuation parameters of ocean sediments is presented. Because of the difficulty in estimating spectral ratios of seismograms in the case when multiple modes are present, a linear perturbation technique that uses the recorded time series seismogram or its envelope as input data is proposed. The technique is stable for tests using synthetic data. Application of the technique to data collected in the Gulf of Mexico yields a model producing a good fit to the data. However, a simple source model is insufficient to correctly predict modal excitations, and the source must be parameterized as part of the inversion process.

Shear Q estimates from interface waves in marine sediments

Scholte wave arrivals at two diverse test sites were analyzed with the spectral ratio method to obtain estimates of average shear Q of deep sea bottom sediments. The Scholte waves were generated by bottom‐located explosive charges and recorded on ocean bottom seismometers. The results obtained are compared to Q estimates reported in the literature.

Observations of the Relative Contributions of Waterborne and Sediment Paths to the Received Acoustic Signal

Examples of the relative contributions of waterborne and sediment paths to acoustic propagation are presented using results obtained in a recent experiment conducted off the Oregon Coast by the Naval Oceanographic and Atmospheric Research Laboratory. The spectral energy contributions of the ground and waterborne paths to the received hydrophone signal are compared for several source-receiver separations, water depths, and frequency bands. The ground path is shown to increase in significance with decreasing frequency and, in some situations, to be as efficient a propagation path as the high-frequency waterborne path.

Modeling the effects of sediments on the acoustic response of buried elastic cylinders to low‐frequency broadband signals.

Low‐frequency broadband (LFBB) sonars have the advantage of penetration into the sea‐floor, thus allowing ensonification of buried targets, but the disadvantage of inferior spatial resolution. Hence, this technology heavily relies on reducing false alarms through recognition of details of the acoustic signature. This reports on an effort to characterize the changes in the acoustic signatures that occur for elastic cylinders buried in various sediments. Numerical modeling of both target and reverberation highlight several difficulties in detection and classification of buried targets, most notably reduced signal excess for targets in sands and larger grain size sediments. Moreover, practical limits on sources and receivers limit the usable bandwidth of LFBB systems, consequently restricting the observable information used for characterizing the acoustic response of targets. Time‐frequency analysis methods of modeled signals reveal little information that distinguishes a signal of interest from a rigid or p...

Acoustic velocity measurements in seafloor sands at frequencies from 1 to 400 kHz

We measured the acoustic velocity and attenuation at frequencies from 1 kHz to 400 kHz in shallow seafloor sands off Fort Walton Beach, FL, USA. We used three separate systems with overlapping frequency ranges in order to cover this entire frequency range. For frequencies from 1 to 20 kHz, we implanted a seafloor array of 35 hydrophones and 5 three‐component accelerometers at depths from 0 to 1 m over a 4 m by 4 m area, and recorded signals from two acoustic sources positioned at offsets from 1 m to 20 m. Measurements from 15 to 120 kHz were made at 30‐cm sediment depths with the In Situ Sediment Acoustic Measurement System (ISSAMS), which consists of a linear array of 4 piezoelectric probes; the outer probes transmit a single‐frequency burst while the inner probes act as receivers. Additionally, velocity measurements were made on diver‐collected cores (5–20 cm sediment depths) at frequencies from 50 to 400 kHz. We present comparisons of the measured frequency dependence of the acoustic velocity and atten...

Deconvolution of Chirp Sidescan Sonar Data

The first return of a sidescan sonar chirp contains information about the bottom and immediate subbottom character and roughness. Measured real data from a chirp sidescan sonar are basebanded, producing complex signals. These signals are deconvolved using an ideal transmitted source signal, a measured source signal, and, for cross‐correlated data, the autocorrelation of the source. Four deconvolution techniques are used: (1) Fourier division in the frequency domain; (2) a least squares technique in the time domain; (3) the reblurring iterative deconvolution method of Kawata and Ichioka (which by definition uses the autocorrelation of the source); and (4) an alternative always‐convergent modification of van Cittert iterative deconvolution. The deconvolved signals are compared to the cross‐correlation of the source with the received signal (matched filter) without deconvolution. Results are discussed in relation to the known bottom characteristics. [Research supported in part by an NRL/ASEE Summer Faculty Fellowship and ONR.]

The effect of transient signal vertical directionality on hydrophone array performance

Signals from explosive sources were measured by a near‐bottom vertical array of hydrophones during an experiment conducted by the Naval Research Laboratory. Using a beamformer, estimates of the source energy spectrum were made with the array steered toward the direct‐path arrival; estimates of the directional noise energy spectrum are made using the same look angle. Signal‐to‐noise (SNR) estimates were then computed for various ranges and steering angles. As expected, the SNR decreases rapidly at longer ranges and lower grazing angles. This decrease is due in part to lower signal levels, but is dominated by large increases in the ambient noise field for grazing angles less than the critical angle of the water–sediment reflection. The SNR for a vertical acoustic dipole formed by differencing two elements of the array also is computed. Despite the decreased response of the dipole beam at low grazing angles, the SNR remained unexpectedly high at longer range and exceeded that of the beam‐steered array. As gr...

The influence of bottom geoacoustics on the dispersive behavior of Scholte interface waves

Scholte seismic interface waves can be an important component of very low‐frequency (VLF) acoustic propagation, particularly in shallow‐water environments. The propagation characteristics of these seismic waves, including their presence in the water column, are dependent on bottom geoacoustics (particularly shear speeds) and the types of layers overlying the basement level. This work examines the results of recent NRL/Stennis Space Center measurements and compares them with numerical predictions based on the SAFARI model. Scholte waves measured off the Oregon Margin (1991) are shown to have strikingly different characteristics from those obtained off the San Diego coast (April 1992). In particular, the latter results (San Diego) are characterized by strongly dispersed Scholte waves in the 1‐ to 5‐Hz band, whereas those measured off the Oregon Margin contain mostly lower frequencies (1–2 Hz) and display less clearly defined dispersive behavior. The differences in the two data sets are attributed to the sig...