Towards Physically Consistent Data-driven Weather Forecasting: Integrating Data Assimilation with an Equivariance Preserving Spatial Transformer

NeurIPS 2020