Inferring Underwater Topography with FINN

TL;DR

This paper demonstrates that the finite volume neural network (FINN) can effectively infer underwater topography from wave dynamics modeled by shallow-water equations, showcasing its potential in geophysical and environmental applications.

Contribution

The study introduces FINN's application to reconstruct underwater topography from wave data, highlighting its efficiency and superiority over traditional and physics-aware ML models.

Findings

FINN accurately reconstructs underwater topography from wave data.

FINN outperforms conventional ML and physics-aware models in this task.

The approach enhances understanding of spatiotemporal phenomena in coastal regions.

Abstract

Spatiotemporal partial differential equations (PDEs) find extensive application across various scientific and engineering fields. While numerous models have emerged from both physics and machine learning (ML) communities, there is a growing trend towards integrating these approaches to develop hybrid architectures known as physics-aware machine learning models. Among these, the finite volume neural network (FINN) has emerged as a recent addition. FINN has proven to be particularly efficient in uncovering latent structures in data. In this study, we explore the capabilities of FINN in tackling the shallow-water equations, which simulates wave dynamics in coastal regions. Specifically, we investigate FINN's efficacy to reconstruct underwater topography based on these particular wave equations. Our findings reveal that FINN exhibits a remarkable capacity to infer topography solely from wave dynamics, distinguishing itself from both conventional ML and physics-aware ML models. Our results underscore the potential of FINN in advancing our understanding of spatiotemporal phenomena and enhancing parametrization capabilities in related domains.