Online machine-learning forecast uncertainty estimation for sequential data assimilation
This work addresses the high computational cost of uncertainty quantification in numerical weather prediction, offering a more efficient alternative for meteorologists and data assimilation practitioners.
The authors tackled the computational burden of ensemble-based forecast uncertainty estimation in data assimilation by proposing a machine learning method using convolutional neural networks to predict the forecast error covariance matrix from a single model integration. Their hybrid data assimilation approach, tested on the Lorenz'96 model, achieved similar performance to the ensemble Kalman filter and outperformed it with small ensembles.
Quantifying forecast uncertainty is a key aspect of state-of-the-art numerical weather prediction and data assimilation systems. Ensemble-based data assimilation systems incorporate state-dependent uncertainty quantification based on multiple model integrations. However, this approach is demanding in terms of computations and development. In this work a machine learning method is presented based on convolutional neural networks that estimates the state-dependent forecast uncertainty represented by the forecast error covariance matrix using a single dynamical model integration. This is achieved by the use of a loss function that takes into account the fact that the forecast errors are heterodastic. The performance of this approach is examined within a hybrid data assimilation method that combines a Kalman-like analysis update and the machine learning based estimation of a state-dependent forecast error covariance matrix. Observing system simulation experiments are conducted using the Lorenz'96 model as a proof-of-concept. The promising results show that the machine learning method is able to predict precise values of the forecast covariance matrix in relatively high-dimensional states. Moreover, the hybrid data assimilation method shows similar performance to the ensemble Kalman filter outperforming it when the ensembles are relatively small.