An Ensemble Information Filter: Retrieving Markov-information from the SPDE discretisation

TL;DR

This paper introduces the Ensemble Information Filter, a novel data assimilation method that encodes Markov properties into the precision matrix to improve scalability and consistency in high-dimensional systems.

Contribution

The Ensemble Information Filter directly incorporates Markov properties into the statistical model, eliminating ad-hoc localisation and enhancing performance in large-scale data assimilation.

Findings

Improves filtering, smoothing, and parameter estimation accuracy.

Reduces issues of spurious correlations and ensemble collapse.

Enhances scalability and statistical consistency.

Abstract

Ensemble-based Data Assimilation faces significant challenges in high-dimensional systems due to spurious correlations and ensemble collapse. These issues arise from estimating dense dependencies with limited ensemble sizes. This paper introduces the Ensemble Information Filter, which encodes Markov properties directly into the statistical model's precision matrix, leveraging structure from SPDE dynamics to constrain information to propagate locally. EnIF eliminates the need for ad-hoc localisation, improving statistical consistency and scalability. Numerical experiments demonstrate its advantages in filtering, smoothing, and parameter estimation, making EnIF a robust and efficient solution for large-scale data assimilation problems.