Model-Based Acoustic Remote Sensing of Seafloor Characteristics

Abstract

The characterization of seafloor sediments by acoustic means is carried out using normal-incidence, dual-frequency, single-beam echo sounder backscatter data at 20 locations in the western continental shelf of India. A temporal backscatter model for a finite duration of transmitted signal is used to estimate the values of four sediment parameters through inversion-mean grain size, roughness spectrum strength, roughness spectrum exponent, and sediment volume scattering parameter. Ground truth information on sediment mean grain sizes at these locations is obtained from the laboratory analysis of grab samples. Applicability of the temporal backscatter model over different sediment compositions at 33 and 210 kHz is investigated here for characterizing the seafloor through inversion. Moreover, a combined two-frequency inversion approach has been attempted for improved characterization by estimating a single set of values for the aforementioned four sediment parameters, which is applicable to the backscatter data at both of the acoustic frequencies. A linear regression analysis demonstrates that the estimated values of sediment mean grain size obtained from these three inversion cases statistically agree with the ground truth data. In addition, correlations of the estimated values of seafloor roughness spectrum parameters with the values of sediment mean grain size are compared with published information available in the literature.