Reply on CEC1

Abstract

The prediction of precipitation patterns at high spatio-temporal resolution up to two hours ahead, also known as precipitation nowcasting, is of great relevance in weather-dependant decision-making and early warning systems. In this study, we are aiming to provide an efficient and easy-to-understand model - CLGAN, to improve the nowcasting skills of heavy precipitation events with deep neural networks for video prediction. The model constitutes a Generative Adversarial Network (GAN) architecture whose generator is built upon an u-shaped encoder-decoder network (U-Net) equipped with recurrent LSTM cells to capture spatio-temporal features. A comprehensive comparison among CLGAN, and baseline models optical flow model DenseRotation as well as the advanced video prediction model PredRNN-v2 is performed. We show that CLGAN outperforms in terms of scores for dichotomous events and object-based diagnostics. The ablation study indicates that the GAN-based architecture helps to capture heavy precipitation events. The results encourage future work based on the proposed CLGAN architecture to improve the precipitation nowcasting and early-warning systems.