DuoCast: Duo-probabilistic diffusion for precipitation nowcasting

Author Identifier (ORCID)

Kun Hu: https://orcid.org/0000-0002-6891-8059

Abstract

Accurate short-term precipitation forecasting is critical for weather-sensitive decision-making in agriculture, transportation, and disaster response. Existing deep learning approaches often struggle to balance global structural consistency with local detail preservation, especially under complex meteorological conditions. We propose DuoCast, a dual-diffusion framework that decomposes precipitation forecasting into low-and high-frequency components modeled in orthogonal latent subspaces. We theoretically prove that this frequency decomposition reduces prediction error compared to conventional single branch U-Net diffusion models. In DuoCast, the low-frequency model captures large-scale trends via convolutional encoders conditioned on weather front dynamics, while the high-frequency model refines fine-scale variability using a self-attention-based architecture. Experiments on four benchmark radar datasets show that Duo-Cast outperforms state-of-the-art baselines, achieving superior accuracy in both spatial detail and temporal evolution.

Keywords

Precipitation nowcasting, diffusion models, frequency decomposition, radar data, deep learning, weather forecasting

Document Type

Conference Proceeding

Date of Publication

1-1-2026

Volume

40

Issue

46

Publication Title

Proceedings of the AAAI Conference on Artificial Intelligence

Publisher

Association for the Advancement of Artificial Intelligence

School

School of Science

Funders

This work was supported by the Australian Research Council (ARC) Linkage Project #LP230100294 and ECU Science Early Career and New Staff Grant Scheme.

Grant Number

ARC Number : LP230100294

Comments

Wen, P., He, M., Filippi, P., Zhao, N., Zhang, F., Bishop, T. F., Wang, Z., & Hu, K. (2026). DuoCast: Duo-probabilistic diffusion for precipitation nowcasting. Proceedings of the AAAI Conference on Artificial Intelligence, 40(46), 39442–39450. https://doi.org/10.1609/aaai.v40i46.41294

Copyright

free_to_read

First Page

39442

Last Page

39450

Link to Full Text

Share

 
COinS
 

Link to publisher version (DOI)

10.1609/aaai.v40i46.41294