Rad-cGAN v1.0: Radar-based precipitation nowcasting  model with conditional generative adversarial  networks for multiple dam domains

Choi, Suyeon; Kim, Yeonjoo

doi:https://doi.org/10.5194/gmd-15-5967-2022

Articles | Volume 15, issue 15

https://doi.org/10.5194/gmd-15-5967-2022

© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/gmd-15-5967-2022

© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 15, issue 15

Model description paper

|

01 Aug 2022

Model description paper |

| 01 Aug 2022

Rad-cGAN v1.0: Radar-based precipitation nowcasting model with conditional generative adversarial networks for multiple dam domains

Suyeon Choi and Yeonjoo Kim

Model code and software

Rad-cGAN v1.0: Radar-based precipitation nowcasting model with conditional Generative Adversarial Networks for multiple dam domains S. Choi and Y. Kim https://doi.org/10.5281/zenodo.6650722

SuyeonC/Rad-cGAN: (v1.0.0) S. Choi and Y. Kim https://doi.org/10.5281/zenodo.6880997

Short summary

Here we present the cGAN-based precipitation nowcasting model, named Rad-cGAN, trained to predict a radar reflectivity map with a lead time of 10 min. Rad-cGAN showed superior performance at a lead time of up to 90 min compared with the reference models. Furthermore, we demonstrate the successful implementation of the transfer learning strategies using pre-trained Rad-cGAN to develop the models for different dam domains.