Uncertainty-aware data assimilation through variational inference
This work addresses uncertainty calibration in data assimilation for dynamical systems, offering incremental improvements over existing methods.
The authors tackled uncertainty in data assimilation by extending a deterministic machine learning approach with variational inference, resulting in nearly perfectly calibrated predictions on the chaotic Lorenz-96 dynamics and improved performance with longer data assimilation windows.
Data assimilation, consisting in the combination of a dynamical model with a set of noisy and incomplete observations in order to infer the state of a system over time, involves uncertainty in most settings. Building upon an existing deterministic machine learning approach, we propose a variational inference-based extension in which the predicted state follows a multivariate Gaussian distribution. Using the chaotic Lorenz-96 dynamics as a testing ground, we show that our new model enables to obtain nearly perfectly calibrated predictions, and can be integrated in a wider variational data assimilation pipeline in order to achieve greater benefit from increasing lengths of data assimilation windows. Our code is available at https://github.com/anthony-frion/Stochastic_CODA.