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 DuoCast consistently outperforms state-of-the-art baselines, achieving superior accuracy in both spatial detail and temporal evolution.
@article{arxiv.2412.01091,
title = {DuoCast: Duo-Probabilistic Diffusion for Precipitation Nowcasting},
author = {Penghui Wen and Mengwei He and Patrick Filippi and Na Zhao and Feng Zhang and Thomas Francis Bishop and Zhiyong Wang and Kun Hu},
journal= {arXiv preprint arXiv:2412.01091},
year = {2026}
}