用物理约束的扩散模型生成极端天气合成数据，解决稀有台风样本不足问题。
Physics-Informed Diffusion Model for Generating Synthetic Extreme Rare Weather Events Data
- 基于上下文UNet架构，结合风速、海况等气象参数生成物理一致的合成卫星图像。
- 在仅202个极端样本的数据集上，成功生成符合真实空间相关性的16×16风场数据。
- 适合需要高精度极端事件训练数据的气象检测与气候建模研究者使用。
数据稀缺是构建稳健机器学习模型以检测快速增强热带气旋的主要障碍。传统数据增强方法（如旋转、翻转、亮度调整）无法保持罕见4级等效事件的物理一致性及高强度梯度特征，这类事件仅占数据集的0.14%（140,514样本中202个）。本文提出一种基于上下文UNet架构的物理信息扩散模型，生成多光谱卫星图像中的合成极端天气数据。模型条件于平均风速、海况类型及发展阶段（早期、成熟期、晚期等）等关键大气参数——这些是快速增强的已知驱动因素。通过受控预生成噪声采样策略与混合精度训练，生成了从多光谱卫星图像中裁剪出的16×16风场样本，保留了真实的空间自相关性与物理一致性。结果表明，模型在十个不同上下文类别中成功学习判别特征，有效缓解数据瓶颈。特别地，面对极端类别不平衡问题——第4类（海况2，早期阶段，平均风速50节飓风）仅有202个样本，而第0类有79,768个样本——该生成框架为操作型天气检测算法提供了可扩展的数据增强方案。平均对数谱距离（LSD）为4.5dB，验证了其在提升业务化天气检测算法方面的可扩展性。
原文摘要 · Abstract (English)
Data scarcity is a primary obstacle in developing robust Machine Learning (ML) models for detecting rapidly intensifying tropical cyclones. Traditional data augmentation techniques (rotation, flipping, brightness adjustment) fail to preserve the physical consistency and high-intensity gradients characteristic of rare Category 4-equivalent events, which constitute only 0.14\% of our dataset (202 of 140,514 samples). We propose a physics-informed diffusion model based on the Context-UNet architecture to generate synthetic, multi-spectral satellite imagery of extreme weather events. Our model is conditioned on critical atmospheric parameters such as average wind speed, type of Ocean and stage of development (early, mature, late etc) -- the known drivers of rapid intensification. Using a controlled pre-generated noise sampling strategy and mixed-precision training, we generated $16\times16$ wind-field samples that are cropped from multi-spectral satellite imagery which preserve realistic spatial autocorrelation and physical consistency. Results demonstrate that our model successfully learns discriminative features across ten distinct context classes, effectively mitigating the data bottleneck. Specifically, we address the extreme class imbalance in our dataset, where Class 4 (Ocean 2, early stage with average wind speed 50kn hurricane) contains only 202 samples compared to 79,768 samples in Class 0. This generative framework provides a scalable solution for augmenting training datasets for operational weather detection algorithms. The average Results yield an average Log-Spectral Distance (LSD) of 4.5dB, demonstrating a scalable framework for enhancing operational weather detection algorithms.
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。