用气象大模型的频谱先验，提升雷达降水预报的准确性和时长。
Extending Precipitation Nowcasting Horizons via Spectral Fusion of Radar Observations and Foundation Model Priors
- 在频域融合雷达图像与气象大模型预测，解决数据表征差异问题。
- 在SEVIR和MeteoNet上实现最先进性能，有效延长可靠预报时间。
- 适合关注天气预报、气象建模及多源数据融合的研究者。
降水短临预报对防灾减灾和航空安全至关重要。然而，仅依赖雷达的模型常因缺乏大尺度大气背景信息，在长时间预报中性能下降。虽然利用天气大模型预测的气象变量可提供补充，但现有架构难以调和雷达图像与气象数据之间的深层表征差异。为此，我们提出PW-FouCast，一种基于傅里叶结构的频域融合框架，将Pangu-Weather的预报结果作为频谱先验。该框架引入三项创新：(i) Pangu-Weather引导的频域调制，对齐谱幅值与相位；(ii) 频率记忆模块，校正相位偏差并保留时间演化；(iii) 反向频率注意力机制，重建高频细节。在SEVIR和MeteoNet基准上的大量实验表明，PW-FouCast达到当前最优表现，显著延长可靠预报时间窗，同时保持结构保真度。代码已开源：https://github.com/Onemissed/PW-FouCast。
原文摘要 · Abstract (English)
Precipitation nowcasting is critical for disaster mitigation and aviation safety. However, radar-only models frequently suffer from a lack of large-scale atmospheric context, leading to performance degradation at longer lead times. While integrating meteorological variables predicted by weather foundation models offers a potential remedy, existing architectures fail to reconcile the profound representational heterogeneities between radar imagery and meteorological data. To bridge this gap, we propose PW-FouCast, a novel frequency-domain fusion framework that leverages Pangu-Weather forecasts as spectral priors within a Fourier-based backbone. Our architecture introduces three key innovations: (i) Pangu-Weather-guided Frequency Modulation to align spectral magnitudes and phases with meteorological priors; (ii) Frequency Memory to correct phase discrepancies and preserve temporal evolution; and (iii) Inverted Frequency Attention to reconstruct high-frequency details typically lost in spectral filtering. Extensive experiments on the SEVIR and MeteoNet benchmarks demonstrate that PW-FouCast achieves state-of-the-art performance, effectively extending the reliable forecast horizon while maintaining structural fidelity. Our code is available at https://github.com/Onemissed/PW-FouCast.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。