Improving the Performance of Mode-Based Sound Propagation Models by Using Perturbation Formulae for Eigenvalues and Eigenfunctions

A. D. Zakharenko, Mikhail Trofimov, Pavel S. PetrovJournal of Marine Science and Engineering 2021 · DOI

TL;DR

This paper introduces a novel perturbation theory to significantly speed up underwater sound propagation simulations. By using analytical formulas to approximate changes in sound modes due to varying water depth, they reduce the computational cost by a factor of 5-10, enabling faster and more efficient acoustic field predictions.

What Problem Does This Paper Address?

Simulating how sound travels underwater is crucial for many applications, from sonar to marine research. A common approach is to represent the sound field as a sum of "normal modes," which are essentially the fundamental ways sound can propagate in a given environment. However, in real-world scenarios, the ocean floor isn't flat; it varies in depth. Calculating these normal modes for every single point in a complex, range-dependent environment is computationally very expensive, forming a bottleneck in these simulations. This paper tackles this problem by developing a way to efficiently estimate how these modes change when the water depth varies, avoiding the need to re-calculate them from scratch at every location.

Methods

The core of this work lies in developing and applying perturbation theory to the acoustic spectral problem that defines the normal modes. The authors transform the problem of water depth variations into a more manageable form, akin to potential perturbation in quantum mechanics. They derive explicit analytical formulas for the first and second-order derivatives of the eigenvalues (horizontal wavenumbers) and eigenfunctions (mode shapes) with respect to water depth. These formulas allow for the calculation of mode properties at different depths based on a single "base" calculation, significantly reducing the number of times the full spectral problem needs to be solved. They then integrate these perturbation formulas with a wide-angle mode parabolic equation to simulate sound propagation.

Datasets

Not discussed in the paper. The paper focuses on developing and validating a theoretical framework using numerical examples based on defined environmental parameters (e.g., sound speed, density, wedge geometry).

Key Insights and Ideas

  • Perturbation Theory for Acoustics: Applying established perturbation theory from quantum mechanics to the specific acoustic spectral problem of underwater waveguides offers a powerful new tool.
  • Efficiency Gain: The derived formulas can reduce the number of mode computations by a factor of 5-10, dramatically speeding up simulations. This is achieved by avoiding repeated, expensive spectral problem solves.
  • Depth Variation as the Key: The paper highlights that variations in water depth are a primary driver of changes in acoustic modes, especially for low-frequency sound.
  • Practical Applicability: The developed method is shown to be effective even in complex scenarios like a 3D coastal wedge, where depth variations are significant.
  • Trade-off Between Accuracy and Speed: Higher-order perturbation formulas provide better accuracy but require more computation. The paper demonstrates that even second-order approximations offer substantial improvements over linear ones, making them suitable for many practical applications.

Key Results

  • Explicit analytical formulas for the first and second-order derivatives of eigenvalues and eigenfunctions with respect to water depth are derived.
  • Numerical simulations in a 3D coastal wedge demonstrate that using these perturbation formulas, combined with a mode parabolic equation, can accurately predict sound propagation.
  • For a coastal wedge scenario, the 3rd-order perturbation approximation for wavenumbers was nearly perfect for moderate depth variations.
  • The 2nd-order perturbation formula provided accurate results up to 15 km in range, a significant improvement over the 1st-order approximation (accurate up to ~7 km).
  • The computational cost reduction was observed to be up to a factor of 10 in the tested wedge problem.

Limitations and Future Work

  • Accuracy Beyond Cut-off Depth: The paper notes that the approximations become invalid beyond the cut-off depth value, where the perturbation theory may not provide sufficient accuracy for wavenumbers.
  • Idealized Scenarios: While the wedge problem is considered a challenging test case, future work could explore more complex environments with additional factors like internal waves or more complex bottom structures.
  • Mode Coupling: The paper mentions that for more realistic problems, taking mode coupling effects into account could further enhance the benefits of this approach.
  • Generalization: The authors suggest their results can be generalized to higher-order perturbations and interface perturbation theories.
  • Two-Point Clamped Spline: For situations where even higher-order perturbation theory isn't sufficient, they propose a "two-point clamped spline approximation" using modes computed at maximal and minimal depths as a compromise between efficiency and accuracy.

Summary updated

Abstract

Numerous sound propagation models in underwater acoustics are based on the representation of a sound field in the form of a decomposition over normal modes. In the framework of such models, the calculation of the field in a range-dependent waveguide (as well as in the case of 3D problems) requires the computation of normal modes for every point within the area of interest (that is, for each pair of horizontal coordinates x,y). This procedure is often responsible for the lion’s share of total computational cost of the field simulation. In this study, we present formulae for perturbation of eigenvalues and eigenfunctions of normal modes under the water depth variations in a shallow-water waveguide. These formulae can reduce the total number of mode computation instances required for a field calculation by a factor of 5–10. We also discuss how these formulae can be used in a combination with a wide-angle mode parabolic equation. The accuracy of such combined model is validated in a series of numerical examples.
BibTeX
@article{zakharenko2021improving,
  author = {A. D. Zakharenko and Mikhail Trofimov and Pavel S. Petrov},
  title = {{Improving the Performance of Mode-Based Sound Propagation Models by Using Perturbation Formulae for Eigenvalues and Eigenfunctions}},
  journal = {Journal of Marine Science and Engineering},
  year = {2021},
  doi = {10.3390/jmse9090934},
}
Figure