CasCast: Skillful High-resolution Precipitation Nowcasting via Cascaded Modelling

Junchao Gong, Lei Bai, Peng Ye, Wanghan Xu, Na Liu, Jianhua Dai, Xiaokang Yang, Wanli Ouyang
Proceedings of the 41st International Conference on Machine Learning, PMLR 235:15809-15822, 2024.

Abstract

Precipitation nowcasting based on radar data plays a crucial role in extreme weather prediction and has broad implications for disaster management. Despite progresses have been made based on deep learning, two key challenges of precipitation nowcasting are not well-solved: (i) the modeling of complex precipitation system evolutions with different scales, and (ii) accurate forecasts for extreme precipitation. In this work, we propose CasCast, a cascaded framework composed of a deterministic and a probabilistic part to decouple the predictions for mesoscale precipitation distributions and small-scale patterns. Then, we explore training the cascaded framework at the high resolution and conducting the probabilistic modeling in a low dimensional latent space with a frame-wise-guided diffusion transformer for enhancing the optimization of extreme events while reducing computational costs. Extensive experiments on three benchmark radar precipitation datasets show that CasCast achieves competitive performance. Especially, CasCast significantly surpasses the baseline (up to +91.8%) for regional extreme-precipitation nowcasting.

Cite this Paper

BibTeX
@InProceedings{pmlr-v235-gong24a, title = {{C}as{C}ast: Skillful High-resolution Precipitation Nowcasting via Cascaded Modelling}, author = {Gong, Junchao and Bai, Lei and Ye, Peng and Xu, Wanghan and Liu, Na and Dai, Jianhua and Yang, Xiaokang and Ouyang, Wanli}, booktitle = {Proceedings of the 41st International Conference on Machine Learning}, pages = {15809--15822}, year = {2024}, editor = {Salakhutdinov, Ruslan and Kolter, Zico and Heller, Katherine and Weller, Adrian and Oliver, Nuria and Scarlett, Jonathan and Berkenkamp, Felix}, volume = {235}, series = {Proceedings of Machine Learning Research}, month = {21--27 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v235/main/assets/gong24a/gong24a.pdf}, url = {https://proceedings.mlr.press/v235/gong24a.html}, abstract = {Precipitation nowcasting based on radar data plays a crucial role in extreme weather prediction and has broad implications for disaster management. Despite progresses have been made based on deep learning, two key challenges of precipitation nowcasting are not well-solved: (i) the modeling of complex precipitation system evolutions with different scales, and (ii) accurate forecasts for extreme precipitation. In this work, we propose CasCast, a cascaded framework composed of a deterministic and a probabilistic part to decouple the predictions for mesoscale precipitation distributions and small-scale patterns. Then, we explore training the cascaded framework at the high resolution and conducting the probabilistic modeling in a low dimensional latent space with a frame-wise-guided diffusion transformer for enhancing the optimization of extreme events while reducing computational costs. Extensive experiments on three benchmark radar precipitation datasets show that CasCast achieves competitive performance. Especially, CasCast significantly surpasses the baseline (up to +91.8%) for regional extreme-precipitation nowcasting.} }
Endnote
%0 Conference Paper %T CasCast: Skillful High-resolution Precipitation Nowcasting via Cascaded Modelling %A Junchao Gong %A Lei Bai %A Peng Ye %A Wanghan Xu %A Na Liu %A Jianhua Dai %A Xiaokang Yang %A Wanli Ouyang %B Proceedings of the 41st International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2024 %E Ruslan Salakhutdinov %E Zico Kolter %E Katherine Heller %E Adrian Weller %E Nuria Oliver %E Jonathan Scarlett %E Felix Berkenkamp %F pmlr-v235-gong24a %I PMLR %P 15809--15822 %U https://proceedings.mlr.press/v235/gong24a.html %V 235 %X Precipitation nowcasting based on radar data plays a crucial role in extreme weather prediction and has broad implications for disaster management. Despite progresses have been made based on deep learning, two key challenges of precipitation nowcasting are not well-solved: (i) the modeling of complex precipitation system evolutions with different scales, and (ii) accurate forecasts for extreme precipitation. In this work, we propose CasCast, a cascaded framework composed of a deterministic and a probabilistic part to decouple the predictions for mesoscale precipitation distributions and small-scale patterns. Then, we explore training the cascaded framework at the high resolution and conducting the probabilistic modeling in a low dimensional latent space with a frame-wise-guided diffusion transformer for enhancing the optimization of extreme events while reducing computational costs. Extensive experiments on three benchmark radar precipitation datasets show that CasCast achieves competitive performance. Especially, CasCast significantly surpasses the baseline (up to +91.8%) for regional extreme-precipitation nowcasting.
APA
Gong, J., Bai, L., Ye, P., Xu, W., Liu, N., Dai, J., Yang, X. & Ouyang, W.. (2024). CasCast: Skillful High-resolution Precipitation Nowcasting via Cascaded Modelling. Proceedings of the 41st International Conference on Machine Learning, in Proceedings of Machine Learning Research 235:15809-15822 Available from https://proceedings.mlr.press/v235/gong24a.html.

Related Material