用单张消费级显卡实现全球14天高精度天气预报，速度比传统方法快十万倍以上。
WeatherMesh-3: Fast and accurate operational global weather forecasting
- 采用潜在空间滚动预测与模块化架构，支持任意长度预报且无需中间编码解码。
- 在0.25度分辨率下12秒完成14天全球预报，相比现有模型误差降低37.7%。
- 仅需一张消费级显卡即可部署，适合快速、低成本的实时气象业务应用。
我们提出WeatherMesh-3（WM-3），一种基于Transformer的全球气象预报系统，在准确率和计算效率上均超越当前最优水平。其创新包括：1）潜在空间滚动机制，可在潜在空间中实现任意长度预测，无需中间编码或解码；2）模块化架构，灵活使用多时间尺度处理器，并融合多个实时分析数据生成混合初始条件。WM-3在单张RTX 4090显卡上，以0.25度分辨率完成14天全球预报仅需12秒，相较传统数值预报方法提速超10万倍，同时在均方根误差（RMSE）上最高提升37.7%，显著优于现有业务模型，且部署仅需单张消费级GPU。我们旨在通过该模型推动气象预报的普惠化，为实际业务提供轻量高效、高性能的机器学习解决方案。
原文摘要 · Abstract (English)
We present WeatherMesh-3 (WM-3), an operational transformer-based global weather forecasting system that improves the state of the art in both accuracy and computational efficiency. We introduce the following advances: 1) a latent rollout that enables arbitrary-length predictions in latent space without intermediate encoding or decoding; and 2) a modular architecture that flexibly utilizes mixed-horizon processors and encodes multiple real-time analyses to create blended initial conditions. WM-3 generates 14-day global forecasts at 0.25-degree resolution in 12 seconds on a single RTX 4090. This represents a >100,000-fold speedup over traditional NWP approaches while achieving superior accuracy with up to 37.7% improvement in RMSE over operational models, requiring only a single consumer-grade GPU for deployment. We aim for WM-3 to democratize weather forecasting by providing an accessible, lightweight model for operational use while pushing the performance boundaries of machine learning-based weather prediction.
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。