Inversion of bottom parameters using a backscattering model based on the effective density fluid approximation

• Published in: Applied acoustics Vol. 182; p. 108187
• Main Authors: Yu, Shengqi, Liu, Baohua, Yu, Kaiben, Yang, Zhiguo, Kan, Guangming, Zong, Le
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.11.2021
• Subjects: Acoustic scattering; Bayesian inversion; Bottom parameters; Hybrid optimization algorithm; Bottom parameters; Hybrid optimization algorithm; Bayesian inversion; Acoustic scattering
• ISSN: 0003-682X; 1872-910X
• DOI: 10.1016/j.apacoust.2021.108187

•The backscattering model is most sensitive to γ2 and w2.•Estimation results of γ3, β, and ρr, are satisfied as well.•Data errors lead to larger uncertainty of inversion parameters.•This method performs robustness in indirect inversion of geoacoustic parameters.•Estimation of attenuation is much better than traditional matched field inversion. An inversion method using backscattering strength is used to obtain bottom parameters. A backscattering model based on the effective density fluid approximation was established as the forward model. The objective function used to scale the mismatch between the model predictions and the measured data (obtained by employing two monostatic transducers with narrow directivity) was established on the assumption that the data errors follow a Gaussian distribution. The maximum a posteriori estimations of inversion parameters were obtained by using a two-step hybrid optimization algorithm combining the differential evolution algorithm and the particle swarm optimization algorithm. The uncertainty and the correlation of inversion parameters were presented using marginal probability distributions and a covariance matrix through Bayesian inversion. Using this method, we can acquire not only geoacoustic parameters (including sound speed and attenuation) but also partial physical parameters of marine sediments and statistical character parameters of seafloor roughness and sediment heterogeneity. Finally, the efficiency of this inversion method was verified through comparison between inversion results and sampling data from sand sediment at the bottom of a water tank.