用视频变压器模型提升短时降雨预测精度，胜出国际竞赛。
A Space-Time Transformer for Precipitation Nowcasting
- 将降雨预测转为分类任务，缓解长尾分布问题
- 采用频率加权损失函数，显著提升预测准确率
- 在2025年天气挑战赛中排名第一，适合气象与AI交叉研究者
长期以来，基于物理的数值天气预报（NWP）模型在业务预报中占据主导地位，但存在计算成本高、短时预测能力弱等局限。为此，近年兴起的AI-天气预测（AI-WP）通过神经网络模拟分析数据，成为替代方案。然而，现有方法在降水及短时预报（nowcasting）应用方面仍不充分。本文提出SaTformer：一种专为短时降雨预测设计的时空变换器模型。针对本质为长尾回归的任务特性，将预测问题转化为分类任务，并引入频率加权损失函数。该方法在NeurIPS Weather4Cast 2025「累计降雨量」挑战赛中取得第一名。代码与模型权重已开源：github.com/leharris3/satformer。
原文摘要 · Abstract (English)
Until recently, numerical weather prediction (NWP) models have stood rivalless in operational forecasting despite a few limitations. Namely, physically-based models are computationally demanding and struggle at short lead times, reducing their utility for nowcasting. Motivated by these shortcomings, recent work proposes AI-weather prediction (AI-WP) alternatives that emulate analysis data with neural networks. While these data-driven approaches have achieved high skill for medium-range forecasting-applications of AI-WP to precipitation and to nowcasting are less explored. To these ends, this paper discusses \textit{SaTformer}: a video transformer adapted for precipitation nowcasting. To ameliorate some problems related to what is essentially a fat-tailed regression task, we find it prudent to formulate nowcasting as a classification problem and employ a frequency-weighted loss. This straightforward approach scored first on the NeurIPS Weather4Cast 2025 ``Cumulative Rainfall'' challenge. Code and model weights are available: \texttt{\href{github.com/leharris3/w4c-25}{github.com/leharris3/satformer}}.
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。