用单步推理替代多步采样，高效生成精准天气集合预报。
Diffusion Distillation for Efficient Weather Ensembles
- 通过能量距离对齐师生模型输出与真实观测，实现压缩蒸馏。
- 单步推理性能媲美甚至超过多步教师模型，极端天气预报更准确。
- 适合需要快速生成高精度天气集合的气象业务与实时系统。
扩散模型能生成高质量的天气集合预报，但需耗费大量迭代采样。本文提出一种监督式能量距离蒸馏方法，将多步扩散教师模型压缩为单步学生模型，通过对齐学生预测、教师样本与真实观测来提升性能。在全局天气预报和台风路径预测任务上，该学生模型优于现有蒸馏方法，且对极端天气事件仍保持高技能水平。其仅需每自回归步骤一次神经网络前向计算，即可在关键指标上达到或超越教师模型表现。
原文摘要 · Abstract (English)
Diffusion models generate skillful weather ensembles but require costly iterative sampling. We introduce a supervised energy-distance distillation method that compresses a multi-step diffusion teacher into a single-step student by aligning student forecasts with teacher samples and ground-truth observations. Experiments on global forecasting and typhoon-track prediction show that our student outperforms existing distillation methods and preserves skill for extreme events. It matches or surpasses the teacher across key metrics using only one neural function evaluation per autoregressive step.
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。