Bayesian Inversion for Geoacoustic Parameters from Ocean Bottom Reflection Loss

Abstract

Geoacoustic inversion is a very important issue in underwater acoustics, and the inversion method based on bottom reflection loss is a valid technique to invert bottom parameters. This paper describes a Bayesian method for estimating bottom parameters in the deep ocean based on inversion of reflection loss versus angle data which were obtained from an experiment conducted in South China Sea in 2013. The experimental data show that bottom loss depends on frequency. The Bayesian method can be applied in nonlinear inversion problems, and it provides useful indication about the quality of the inversion and parameter sensitivities. The bottom is modeled as a two-layer model, and each layer has constant parameters. The inverted parameters of sediment show a clay feature which is consistent with the core data. Furthermore, the inversion results are used to calculate transmission losses (TLs) along the experiment track which agree well with the direct measurements. Although the inversion results are limited to reveal exact structures of bottom, they are still useful for forecasting propagation losses in this area.