Convergence analysis of a variational data assimilation scheme for bathymetry detection from surface wave observations

TL;DR

This paper presents a variational data assimilation method to improve ocean bathymetry estimates from surface wave data, demonstrating that even with some errors, surface wave predictions remain highly accurate, aiding tsunami modeling.

Contribution

It introduces a convergence analysis and regularization approach for a variational data assimilation scheme applied to shallow water equations for bathymetry detection.

Findings

Convergence conditions depend on system parameters.

Low-pass filtering enhances bathymetry regularity.

High surface wave accuracy achieved despite bathymetry errors.

Abstract

Accurate mapping of ocean bathymetry is a multi-faceted process, needed for safe and efficient navigation on shipping routes and for predicting tsunami waves. Currently available bathymetry data does not always provide the resolution to capture dynamics of such nonlinear waves accurately. However collection of accurate mapping data is difficult, costly, and often a dangerous affair. As an alternative, in this study we implement a variational data assimilation scheme on the one-dimensional shallow water equations to improve estimates of bathymetry, using a finite set of observations of surface wave height to optimise predictions. We show necessary conditions on system parameters for convergence, and implement a low-pass filter for increased regularity of our reconstructed bathymetry. If our objective is to use this to model tsunami propagation, we observe that a relatively higher error in the optimal reconstruction of the bathymetry still yields a highly accurate prediction of the surface wave, suggesting low sensitivity of surface waves to small scale effects in the bathymetry. These conclusions are based on numerical experiments for both Gaussian and sandbar profile bathymetry, and with different observation operators.These extension of these results to realistic models can potentially have a significant impact, as computational cost can be minimised through a priori knowledge of sufficient error tolerances needed for accurate tsunami prediction.