用流匹配模型将降水数据从8公里分辨率提升至2公里，效果优于传统方法。
Flow Matching for Convective-Scale Precipitation Downscaling
- 采用流匹配框架，将8公里降水数据生成2公里高分辨率结果。
- 在多个降水阈值和尺度下，空间结构精度更高，位置误差相当。
- 适合关注降水空间分布细节的研究者，尤其适用于对流尺度模拟。
生成式机器学习正日益成为动态降尺度的重要补充，用于生成高分辨率降水预测。目前扩散模型是主流方法，而流匹配是一种新兴的生成框架，在图像、视频等领域已取得优异表现，并展现出降尺度应用的潜力。本文训练了一个流匹配模型，将新加坡为中心的对流尺度区域的日降水数据从8公里分辨率提升至2公里，并与基于得分的扩散模型CPMGEM进行对比。结果表明，流匹配在空间技能上表现更优：在所有降水阈值和邻域尺度下均获得更高的分数技能得分，且结构和振幅分量的SAL评分更紧致，位置技能相当；但流匹配低估了降水分布的上尾，导致气候平均值存在干偏差。这些结果表明，流匹配是具备竞争力的生成框架，尤其擅长捕捉降水的空间结构。
原文摘要 · Abstract (English)
Generative machine learning is an increasingly important complement to dynamical downscaling for producing high-resolution precipitation projections, with diffusion models currently the leading approach. Flow matching is a related generative framework that has recently achieved strong results across image, video and other domains, and shown early promise for downscaling. We train a flow matching model to map daily precipitation from 8 km to 2 km over a convective-scale domain centred on Singapore, and benchmark it against CPMGEM, a score-based diffusion model. Flow matching achieves consistently better spatial skill: higher fractions skill score at every precipitation threshold and neighbourhood scale tested, and tighter structure and amplitude components of the SAL score with comparable location skill. However, flow matching underestimates the upper tail of the precipitation distribution, resulting in a dry bias in the climatological mean. These results suggest that flow matching is a competitive generative framework for convective-scale precipitation downscaling, particularly well suited to capturing spatial structure.
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。