Ocean-depth measurement using shallow-water wave models

TL;DR

This paper presents an algorithm to accurately reconstruct ocean bottom topography from surface wave measurements using shallow-water models, addressing ill-posed inverse problems with model-based techniques.

Contribution

It introduces a novel two-step inverse problem approach combining velocity and boundary recovery, improving accuracy in shallow-water wave topography estimation.

Findings

Successful reconstruction of bottom boundary from surface data

Effective separation of inverse problems for velocities and boundary

Highlighting the importance of model selection in accuracy

Abstract

In this paper, we consider a problem inspired by the real-world need to identify the topographical features of ocean basins. Specifically we consider the problem of estimating the bottom impermeable boundary to an inviscid, incompressible, irrotational fluid from measurements of the free-surface deviation alone, within the context of dispersive shallow-water wave models. The need to consider the shallow-water regime arises from the ill-posed nature of the problem and is motivated by prior work. We design an algorithm using which, both fluid velocities and the bottom-boundary profile, may be accurately recovered assuming an a priori relatively inaccurate guess for the bottom boundary. We achieve this by considering two separate inverse problems: one to deduce the bottom-boundary from velocities and the surface deviation, and another to recover the velocities from the surface deviation and an approximate bottom-boundary. The former is a classic inverse problem that requires the inversion of an ill-conditioned matrix while the latter employs the observer framework. Combining the two inverse problems leads to our reconstruction algorithm. We emphasise the role played by model selection and its impact on algorithm design and the accuracy of the reconstruction.