用概率U-Net提升气候模拟分辨率，更好捕捉极端天气与空间变化。
A Probabilistic U-Net Approach to Downscaling Climate Simulations
- 基于变分潜空间的概率U-Net架构，融合确定性网络与不确定性建模
- WMSE-MS-SSIM在极端事件下表现更优，afCRPS更擅长保留空间变异特征
- 适合需要高精度气候投影的环境影响评估与极端气候研究者
气候模型因计算成本高，常生成粗分辨率输出，而许多气候变化影响研究需更高空间精度。统计降尺度可弥补这一差距。本文采用概率U-Net方法，结合确定性U-Net主干与变分潜空间，以捕捉随机不确定性。我们评估了四种训练目标：afCRPS与WMSE-MS-SSIM在三种设置下用于从16×更粗分辨率降尺度降水与温度。主要发现为：在特定设置下，WMSE-MS-SSIM对极端事件表现更佳；而afCRPS能更好地保持跨尺度的空间变异特性。
原文摘要 · Abstract (English)
Climate models are limited by heavy computational costs, often producing outputs at coarse spatial resolutions, while many climate change impact studies require finer scales. Statistical downscaling bridges this gap, and we adapt the probabilistic U-Net for this task, combining a deterministic U-Net backbone with a variational latent space to capture aleatoric uncertainty. We evaluate four training objectives, afCRPS and WMSE-MS-SSIM with three settings for downscaling precipitation and temperature from $16\times$ coarser resolution. Our main finding is that WMSE-MS-SSIM performs well for extremes under certain settings, whereas afCRPS better captures spatial variability across scales.
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。