Tropical cyclone (TC) forecasting is critical for disaster warning and emergency response. Deep learning methods address computational challenges but often neglect physical relationships between TC attributes, resulting in predictions lacking physical consistency. To address this, we propose Phys-Diff, a physics-inspired latent diffusion model that disentangles latent features into task-specific components (trajectory, pressure, wind speed) and employs cross-task attention to introduce prior physics-inspired inductive biases, thereby embedding physically consistent dependencies among TC attributes. Phys-Diff integrates multimodal data including historical cyclone attributes, ERA5 reanalysis data, and FengWu forecast fields via a Transformer encoder-decoder architecture, further enhancing forecasting performance. Experiments demonstrate state-of-the-art performance on global and regional datasets.
@article{arxiv.2603.00521,
title = {Phys-Diff: A Physics-Inspired Latent Diffusion Model for Tropical Cyclone Forecasting},
author = {Lei Liu and Xiaoning Yu and Kang Chen and Jiahui Huang and Tengyuan Liu and Hongwei Zhao and Bin Li},
journal= {arXiv preprint arXiv:2603.00521},
year = {2026}
}