Data Assimilation Networks

TL;DR

This paper introduces a deep learning architecture that generalizes data assimilation methods, capable of handling nonlinear dynamics and non-Gaussian densities, and performs comparably to traditional ensemble Kalman filters in experiments.

Contribution

It proposes a fully data-driven neural network approach for data assimilation that extends existing methods to nonlinear and non-Gaussian settings.

Findings

Achieves performance comparable to EnKF in Lorenz-95 system experiments.

Handles general nonlinear dynamics and non-Gaussian densities.

Does not require explicit regularization techniques.

Abstract

Data assimilation (DA) aims at forecasting the state of a dynamical system by combining a mathematical representation of the system with noisy observations taking into account their uncertainties. State of the art methods are based on the Gaussian error statistics and the linearization of the non-linear dynamics which may lead to sub-optimal methods. In this respect, there are still open questions how to improve these methods. In this paper, we propose a fully data driven deep learning architecture generalizing recurrent Elman networks and data assimilation algorithms which approximate a sequence of prior and posterior densities conditioned on noisy observations. By construction our approach can be used for general nonlinear dynamics and non-Gaussian densities. On numerical experiments based on the well-known Lorenz-95 system and with Gaussian error statistics, our architecture achieves comparable performance to EnKF on both the analysis and the propagation of probability density functions of the system state at a given time without using any explicit regularization technique.