arXiv：cs.LG· Shunya Nagashima, Takumi Bannai·· 3 小时前AI 评分26
尺度递归整流流：少步降水集合预报新方法
Scale-Recursive Rectified Flows for Few-Step Precipitation Ensembles
AI 导读
研究者提出尺度递归整流流（scale-recursive rectified flow），先生成大尺度降水格局再补局部细节，并按空间尺度比较集合变异性与预测误差来分配采样步数。在美国本土卫星到雷达降尺度任务中，该方法识别出大尺度降水格局是减少采样预算下集合变异性不足的主因，将更多步数分配给粗尺度流后，在固定架构与计算成本下提升了概率精度和降雨检测能力，且以更短采样时间优于使用更多步数的非递归流。
正文
Abstract:Fine-resolution precipitation estimates support flood risk assessment and water management, but coarse satellite products cannot resolve rainfall within each grid cell. Generative models address this ambiguity by producing ensembles of plausible high-resolution rainfall fields. Among these models, rectified flows generate samples by iteratively transforming random noise into rainfall fields. Reducing the number of sampling steps accelerates generation but can make ensemble members too similar, understating uncertainty. We propose a scale-recursive rectified flow that generates broad patterns before local details and guides sampling-step allocation by comparing ensemble variability with prediction error across spatial scales. Validation scores and rainfall power spectra constrain the allocation to avoid excessive amplification. In satellite-to-radar downscaling over the contiguous United States, our analysis identified broad rainfall patterns as the main source of insufficient ensemble variability under reduced sampling budgets. Allocating more steps to the coarse flow improved probabilistic accuracy and rain detection across training seeds at fixed architecture and computational cost. The proposed model also achieved better probabilistic accuracy with shorter sampling time than a nonrecursive flow using more steps.
| Subjects: | Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV) |
| Cite as: | arXiv:2610.02611 [cs.LG] |
| (or arXiv:2610.02611v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.02611 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Shunya Nagashima [view email]
[v1]
Fri, 2 Oct 2026 00:11:53 UTC (381 KB)
来源：arXiv：cs.LG · arxiv.org