Darryl L. DeBruin
Lester R. LeBlanc
Steven G. Schock
Ctr. for Acoust. and Vib., Dept. of Ocean Eng., Florida Atlantic Univ., 500 NW 20th St., Boca Raton, FL 33431
Using high-resolution sonar data, a rule base classifier model is developed to correlate acoustic impedance and volume scattering with sediment mean grain size and bulk density. The acoustic impedance is inverted from the impulse response of the sonar data and the scattering strength is calculated between detected layers using a scattering model. Synthetic data and field data are used to demonstrate the high-resolution layer detection method, based on a local least-squares fit, that extracts the impulse response of the acoustic data. This analysis is heuristically constrained to estimate solutions within the resolution bounds of the sonar data. The classifier uses historic measurements, empirical relations, and fuzzy set theory to build the rule base model that is used to correlate the acoustic properties with sediment mean grain size and bulk density. Comparisons between estimates of mean grain size and bulk density from the rule base classifier model analysis of actual sonar data and geotechnical measurements are made in various depositional environments. The sonar and core data were acquired in Kiel Germany in collaboration with the Coastal Benthic Boundary Layer research team.