用预训练编码器DINOv3+轻量头实现高效概率降雨预报，性能提升26%。
DINOv3 as a Frozen Encoder for CRPS-Oriented Probabilistic Rainfall Nowcasting
- 将DINOv3-SAT493M与视频投影头结合，输出4小时累积降雨的离散经验累积分布函数
- 在Weather4Cast 2025上取得CRPS 3.5102，比最优3D-UNET提升约26%
- 适合需要快速部署、高精度概率降水预测的气象应用
本文提出一种高效且具有竞争力的概率降雨临近预报方法。将一个视频投影器（V-JEPA Vision Transformer）与轻量级概率头连接至预训练的卫星视觉编码器（DINOv3-SAT493M），将编码器令牌映射为4小时累积降雨的离散经验累积分布函数（eCDF）。投影头在排序概率得分（RPS）下端到端优化。作为对比，采用基于总排名概率得分和逐像素伽马-截断目标函数训练的3D-UNET基线模型。在Weather4Cast 2025基准测试中，所提方法表现优异，达到CRPS 3.5102，相较于最优3D-UNET实现约26%的有效性提升。
原文摘要 · Abstract (English)
This paper proposes a competitive and computationally efficient approach to probabilistic rainfall nowcasting. A video projector (V-JEPA Vision Transformer) associated to a lightweight probabilistic head is attached to a pre-trained satellite vision encoder (DINOv3-SAT493M) to map encoder tokens into a discrete empirical CDF (eCDF) over 4-hour accumulated rainfall. The projector-head is optimized end-to-end over the Ranked Probability Score (RPS). As an alternative, 3D-UNET baselines trained with an aggregate Rank Probability Score and a per-pixel Gamma-Hurdle objective are used. On the Weather4Cast 2025 benchmark, the proposed method achieved a promising performance, with a CRPS of 3.5102, which represents $\approx$ 26% in effectiveness gain against the best 3D-UNET.
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。