Recent advances in acoustic sediment classification

Abstract

Most acoustic techniques used to classify seafloor sediments rely on operator experience and require extensive ground-truth data to calibrate system predictions. Recent developments in transducer design, inversion and deconvolution techniques, and sensor motion composition should improve the reliability and accuracy of normal-incidence, seafloor classification techniques. This presentation provides an introduction to papers that describe the application of these techniques to normal-incidence acoustic returns collected from a variety of siliciclastic and carbonate sediments in coastal waters surrounding Florida and the Bahamas. Site locations and descriptions, sediment characteristics based on in situ and laboratory measurements, and a description of the acoustic measurement techniques are given. In addition, 3-D visualization techniques that provide near real-time presentation of results are described. Profiles of sediment acoustic impedance, compressional wave attenuation, seafloor roughness, and volume heterogeneity are directly determined from acoustic returns and physical constraints. Other properties such as sediment type, grain size, porosity, and shear strength are predicted on the basis of empirical relationships relating acoustic impedance and/or attenuation to sediment physical properties. Comparisons of predicted and measured sediment properties are given and the ping-to-ping variability inherent in acoustic returns is discussed. [Work supported by ONR.]