Remote acoustic evaluation of seafloor sediment properties

A remote marine sediment classification model that can be implemented in real time is developed. The sediment classification and physical property prediction are based on attenuation and impedance estimates from normal incident seismograms. The specific issues investigated are: (1) selection and design considerations for a suitable quantitative sub-bottom profiler; (2) development of a deconvolution technique for improving the resolution of acoustic images; and (3) development of attenuation and impedance estimation techniques.; The chirp sonar is ideally suited for acquiring quantitative reflection data for remote sediment classification as it provides wideband, low-noise, digital acoustic data. An acoustic pulse with special spectral characteristics is designed to provide low temporal sidelobe levels and minimal pulse spreading due to sediment attenuation. After correlation processing, the wideband acoustic pulse yields an effective beampattern with high spatial resolution and no sidelobes.; A high resolution deconvolution technique based on autoregressive spectral extrapolation in the frequency domain is presented. The prediction is performed by a modified forward-backward linear prediction technique which can handle noisy field data. The accuracy and the robustness of the method are demonstrated by analyzing synthetic data and field data.; A relaxation time mechanism is used for modeling acoustic attenuation in marine sediments. Using available experimental data, the model is validated and a relation between sediment mean grain diameter and relaxation time is obtained. Attenuation estimates are obtained by measuring the center frequency shift of the attenuated signal. Finally, an impedance inversion method, which is robust and requires significantly less computation, is developed. The impulse response of the sediment column is evaluated by a "modified least-squares method". Relations between sediment physical properties and acoustic properties are developed by least-square fitting a physical model to sediment core measurement data. Reflection data acquired by the chirp sonar are analyzed to provide attenuation and impedance estimates and to predict the in situ sediment types. The predicted parameters are in good agreement with the available core measurements, and demonstrate the ability of the model to analyze noisy field data and characterize seafloor sediments successfully.