Single-hydrophone Bayesian matched-field geoacoustic inversion

This paper presents Bayesian matched-field geoacoustic inversion results based on measurements from a single hydrophone. To efficiently compute the posterior distribution, we employ an adaptive Metropolis-Hastings sampling strategy, which dynamically adjusts the proposal distribution. Our method is applied to experimental data from the Shallow Water 2006 experiment, considering two broadband acoustic sources. The inversion targets four key parameters: source range, source depth, water depth, and sediment compressional sound speed. We present results indicating that Bayesian single-hydrophone inversions can provide accurate parameter estimates, underscoring their utility in resource-limited or rapid-deployment scenarios.