Answers are grounded in this paper and cite their sources.
In ocean acoustics, vertical array depth shifts caused by currents lead to missing shallow-water acoustic field data. This study applies a Gaussian process (GP) regression framework with a Normal-Mode-Based Kernel to address the extrapolation problem in transmission loss reconstruction. Based on simulations of the 2024 South China Sea experiment environment and validation with measured data from the SWellEx-96 experiment, the GP approach successfully extrapolates the acoustic field to missing depths. Results demonstrate that, compared to using the transmission loss value at the closest available depth as a reference, the reconstruction error is significantly reduced. Furthermore, by incorporating dominant modal prior information, the Normal-Mode-Based Kernel-GP method maintains reliable performance even for complex sound fields, improving reconstruction stability under array deployment uncertainties.
No takes yet. Share an insight, caveat, or question.
Yang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: