Shallow water transmission loss inversion for seafloor characterization

A Bayesian inference algorithm has been developed for seafloor geoacoustic parameter determination through analysis of transmission loss measurements taken in shallow water environments. It uses a parallel tempering scheme for improved parameter space sampling and bias reduction, and evaluates ray acoustic forward models in Bellhop. The algorithm is verified by application to synthetic data generated using finite element and ray acoustic propagation models for frequencies from 0.5 to 5 kHz. Verification is assessed through the ability to recover inversion parameter values prescribed within the synthetic data, and inspection of ambiguous inversion situations. Initial inversion parameters include bulk sediment density, sound speed, and attenuation, as well as ripple height and wavelength for some seafloors. The inversion yields posterior probability distributions for each inversion parameter, which inform parameter resolvability and covariance. Algorithm accuracy, efficiency, and applicability to data from different ocean environments will be discussed. [This work has been supported by the Office of Naval Research, Task Force Ocean.]