CLGAN: A GAN-based video prediction model for precipitation nowcasting
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.
Yan Ji et al.
Status: final response (author comments only)
CEC1: 'Comment on egusphere-2022-859', Juan Antonio Añel, 12 Dec 2022
- AC1: 'Reply on CEC1', Yan Ji, 12 Dec 2022
RC1: 'Comment on egusphere-2022-859', Anonymous Referee #1, 13 Dec 2022
- AC2: 'Reply on RC1', Yan Ji, 17 Jan 2023
RC2: 'Comment on egusphere-2022-859', Anonymous Referee #2, 16 Dec 2022
- AC3: 'Reply on RC2', Yan Ji, 17 Jan 2023
CC1: 'Comment on egusphere-2022-859', Qiuming Kuang, 20 Dec 2022
- AC4: 'Reply on CC1', Yan Ji, 17 Jan 2023
Yan Ji et al.
Yan Ji et al.
Viewed (geographical distribution)
Unfortunately, after checking your manuscript, it has come to our attention that it does not comply with our "Code and Data Policy".
Your manuscript states that you have archived the code and data necessary to replicate your study in Zenodo. However, the text does not indicate how to reach such a repository; therefore, this is the same that does not publish it.
Your manuscript should have never been published in Discussions with such a shortcoming. However, now that this oversight by the Topical Editor has happened, we are offering you the possibility to solve it. In this way, you must publish a reply to this comment as soon as possible with the link and DOi of the suitable repository, according to our policy, where the code and data are stored.
Otherwise, we will have to end the Discussion stage and reject your manuscript for publication.
Juan A. Añel
Geosci. Model Dev. Executive Editor