Latent assimilation with implicit neural representations for unknown dynamics

TL;DR

This paper introduces LAINR, a novel data assimilation framework utilizing implicit neural representations and uncertainty estimation to improve efficiency and accuracy in modeling unknown dynamics.

Contribution

The paper proposes LAINR, combining SINR and neural network uncertainty estimation, offering a new approach to data assimilation for complex, high-dimensional systems.

Findings

LAINR outperforms AutoEncoder-based methods in accuracy

LAINR demonstrates improved efficiency in data assimilation tasks

Experimental results validate the effectiveness of SINR and uncertainty estimation

Abstract

Data assimilation is crucial in a wide range of applications, but it often faces challenges such as high computational costs due to data dimensionality and incomplete understanding of underlying mechanisms. To address these challenges, this study presents a novel assimilation framework, termed Latent Assimilation with Implicit Neural Representations (LAINR). By introducing Spherical Implicit Neural Representations (SINR) along with a data-driven uncertainty estimator of the trained neural networks, LAINR enhances efficiency in assimilation process. Experimental results indicate that LAINR holds certain advantage over existing methods based on AutoEncoders, both in terms of accuracy and efficiency.