Flood Extent Data for Machine Learning

本数据集名为‘Flood Extent Data for Machine Learning’，由美国国家航空航天局创建，旨在通过机器学习技术精确捕捉洪水范围。数据集覆盖美国大陆和孟加拉国约36,000平方公里的区域，包含36,000个标记的水体范围和洪水区域。创建过程中，利用了Sentinel-1 C波段合成孔径雷达（SAR）图像，并通过公民科学方法，如公开数据集和举办开放竞赛，加速洪水范围检测模型的原型开发。该数据集主要应用于洪水监测和灾害响应决策，以提高全球洪水范围模型的稳健性。

This dataset is named "Flood Extent Data for Machine Learning" and was developed by the National Aeronautics and Space Administration (NASA) to accurately capture flood extents using machine learning technologies. It covers an area of approximately 36,000 square kilometers across the continental United States and Bangladesh, and contains 36,000 labeled water body extents and flood regions. During its development, Sentinel-1 C-band synthetic aperture radar (SAR) images were utilized, and citizen science approaches including public datasets and open competitions were adopted to accelerate the prototype development of flood extent detection models. This dataset is primarily applied to flood monitoring and disaster response decision-making to enhance the robustness of global flood extent models.